2016-07-01T00:01:43Z eschatologist quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-01T00:02:30Z eschatologist joined #ccl 2016-07-01T00:02:34Z eschatologist quit (Client Quit) 2016-07-01T00:04:13Z eschatologist joined #ccl 2016-07-01T00:22:08Z eschatologist quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-01T00:22:25Z eschatologist joined #ccl 2016-07-01T01:02:07Z ekinmur joined #ccl 2016-07-01T01:05:47Z ekinmur quit (Client Quit) 2016-07-01T01:55:00Z ekinmur joined #ccl 2016-07-01T02:37:00Z pierpa quit (Ping timeout: 276 seconds) 2016-07-01T03:05:32Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-01T03:13:44Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-01T03:39:00Z eschatologist joined #ccl 2016-07-01T04:42:57Z eschatologist quit (Read error: Connection reset by peer) 2016-07-01T04:54:28Z eschatologist joined #ccl 2016-07-01T05:09:28Z logrus quit (Quit: Leaving) 2016-07-01T05:37:22Z rme quit (Quit: rme) 2016-07-01T05:37:22Z rme quit (Quit: rme) 2016-07-01T06:50:40Z DavidGu joined #ccl 2016-07-01T06:53:57Z DavidGu1 joined #ccl 2016-07-01T06:57:30Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-01T06:57:30Z DavidGu1 is now known as DavidGu 2016-07-01T08:05:23Z hhdave joined #ccl 2016-07-01T08:27:49Z edgar-rft quit (Quit: edgar-rft) 2016-07-01T08:32:10Z edgar-rft joined #ccl 2016-07-01T08:49:51Z rumbler31 joined #ccl 2016-07-01T09:39:11Z DavidGu quit (Ping timeout: 258 seconds) 2016-07-01T10:31:06Z DavidGu joined #ccl 2016-07-01T11:53:41Z DavidGu1 joined #ccl 2016-07-01T11:56:26Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-01T11:56:26Z DavidGu1 is now known as DavidGu 2016-07-01T12:16:42Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-01T12:36:06Z logrus joined #ccl 2016-07-01T12:57:54Z rme joined #ccl 2016-07-01T13:14:25Z ekinmur joined #ccl 2016-07-01T13:49:40Z pierpa joined #ccl 2016-07-01T13:55:00Z pierpa` joined #ccl 2016-07-01T14:22:48Z pierpa` quit (Quit: brb) 2016-07-01T14:23:47Z pierpa quit (Quit: brb) 2016-07-01T14:24:35Z pierpa joined #ccl 2016-07-01T14:25:22Z rumbler31 quit (Ping timeout: 250 seconds) 2016-07-01T15:14:42Z rme quit (Ping timeout: 186 seconds) 2016-07-01T15:14:51Z rme quit (Ping timeout: 240 seconds) 2016-07-01T15:41:18Z edgar-rft quit (Quit: edgar-rft) 2016-07-01T15:43:46Z rme joined #ccl 2016-07-01T16:28:01Z rumbler31 joined #ccl 2016-07-01T16:42:29Z hhdave quit (Ping timeout: 260 seconds) 2016-07-01T16:49:09Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-01T17:32:08Z rumbler31 joined #ccl 2016-07-01T18:55:20Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-01T19:09:12Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-01T19:40:43Z sigjuice quit (Ping timeout: 252 seconds) 2016-07-01T19:42:29Z akkad quit (Ping timeout: 252 seconds) 2016-07-01T19:42:29Z kini quit (Ping timeout: 252 seconds) 2016-07-01T19:47:29Z sigjuice joined #ccl 2016-07-01T19:51:13Z kini joined #ccl 2016-07-01T19:51:56Z akkad joined #ccl 2016-07-01T19:56:28Z eschatologist joined #ccl 2016-07-01T21:00:07Z edgar-rft joined #ccl 2016-07-01T21:22:00Z rumbler31 quit (Ping timeout: 272 seconds) 2016-07-01T21:31:10Z rumbler31 joined #ccl 2016-07-01T21:53:11Z fe[nl]ix quit (Quit: No Ping reply in 180 seconds.) 2016-07-01T21:53:12Z Blkt quit (Quit: No Ping reply in 180 seconds.) 2016-07-01T21:53:17Z Blkt joined #ccl 2016-07-01T21:53:18Z fe[nl]ix joined #ccl 2016-07-01T22:36:47Z taij33n quit (Quit: ZNC - http://znc.in) 2016-07-01T22:54:24Z rumbler31 quit (Remote host closed the connection) 2016-07-01T23:50:25Z rumbler31 joined #ccl 2016-07-02T00:58:46Z eschatologist quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-02T01:30:51Z ekinmur joined #ccl 2016-07-02T01:34:05Z rumbler3_ joined #ccl 2016-07-02T01:35:00Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-02T02:45:27Z rumbler3_ quit (Remote host closed the connection) 2016-07-02T03:06:08Z logrus quit (Quit: Leaving) 2016-07-02T03:14:11Z eschatologist joined #ccl 2016-07-02T03:19:43Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-02T03:43:16Z eschatologist joined #ccl 2016-07-02T03:47:43Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-02T04:02:52Z pierpa quit (Ping timeout: 260 seconds) 2016-07-02T04:24:39Z jdz quit (Ping timeout: 258 seconds) 2016-07-02T04:30:15Z jdz joined #ccl 2016-07-02T05:04:50Z fe[nl]ix quit (*.net *.split) 2016-07-02T05:04:50Z Blkt quit (*.net *.split) 2016-07-02T05:04:50Z edgar-rft quit (*.net *.split) 2016-07-02T06:17:41Z DavidGu joined #ccl 2016-07-02T07:09:44Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-02T07:15:31Z DavidGu joined #ccl 2016-07-02T07:24:56Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-02T07:26:02Z DavidGu joined #ccl 2016-07-02T07:34:47Z DavidGu quit (Ping timeout: 258 seconds) 2016-07-02T07:40:11Z edgar-rft joined #ccl 2016-07-02T07:53:21Z rme quit (Quit: rme) 2016-07-02T07:53:22Z rme quit (Quit: rme) 2016-07-02T07:53:42Z DavidGu joined #ccl 2016-07-02T07:58:06Z DavidGu quit (Read error: Connection reset by peer) 2016-07-02T07:58:10Z DavidGu1 joined #ccl 2016-07-02T08:02:11Z DavidGu1 quit (Ping timeout: 240 seconds) 2016-07-02T08:18:10Z DavidGu joined #ccl 2016-07-02T08:23:12Z DavidGu quit (Ping timeout: 276 seconds) 2016-07-02T08:26:20Z DavidGu joined #ccl 2016-07-02T08:30:35Z DavidGu quit (Ping timeout: 244 seconds) 2016-07-02T09:12:51Z DavidGu joined #ccl 2016-07-02T09:15:11Z schjetne quit (Ping timeout: 240 seconds) 2016-07-02T09:34:50Z schjetne joined #ccl 2016-07-02T10:49:14Z DavidGu quit (Quit: DavidGu) 2016-07-02T11:49:53Z fe[nl]ix joined #ccl 2016-07-02T12:20:04Z rme joined #ccl 2016-07-02T13:15:08Z schjetne_ joined #ccl 2016-07-02T13:17:39Z schjetne quit (Ping timeout: 276 seconds) 2016-07-02T13:29:53Z schjetne_ is now known as schjetne 2016-07-02T13:34:01Z schjetne_ joined #ccl 2016-07-02T13:35:21Z schjetne quit (Ping timeout: 246 seconds) 2016-07-02T14:17:14Z pierpa joined #ccl 2016-07-02T14:55:46Z ekinmur joined #ccl 2016-07-02T15:04:28Z ekinmur quit (Max SendQ exceeded) 2016-07-02T15:05:09Z ekinmur joined #ccl 2016-07-02T15:08:32Z schjetne_ quit (Ping timeout: 272 seconds) 2016-07-02T15:09:52Z rumbler31 joined #ccl 2016-07-02T16:14:04Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-02T16:55:36Z hhdave joined #ccl 2016-07-02T16:59:59Z hhdave quit (Ping timeout: 250 seconds) 2016-07-02T18:26:07Z rmetest joined #ccl 2016-07-02T18:30:38Z rmetest quit (Ping timeout: 250 seconds) 2016-07-02T18:58:28Z ro2k joined #ccl 2016-07-02T19:12:18Z rumbler31 quit (Ping timeout: 244 seconds) 2016-07-02T19:15:42Z ro2k quit (Ping timeout: 250 seconds) 2016-07-02T19:16:25Z rumbler31 joined #ccl 2016-07-02T20:25:51Z ekinmur joined #ccl 2016-07-02T20:51:16Z Quadrescence quit (Quit: Leaving) 2016-07-02T21:07:28Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-02T21:24:06Z ekinmur joined #ccl 2016-07-02T21:38:59Z Quadrescence joined #ccl 2016-07-02T22:01:04Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-02T23:18:15Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-02T23:35:57Z kpoeck joined #ccl 2016-07-02T23:58:26Z rumbler31 joined #ccl 2016-07-03T00:00:17Z rumbler31 quit (Remote host closed the connection) 2016-07-03T00:29:10Z kpoeck quit (Quit: Page closed) 2016-07-03T01:10:56Z pierpa quit (Ping timeout: 250 seconds) 2016-07-03T01:43:18Z ekinmur joined #ccl 2016-07-03T02:00:47Z rme quit (Quit: rme) 2016-07-03T02:00:47Z rme quit (Quit: rme) 2016-07-03T02:06:27Z mrottenkolber joined #ccl 2016-07-03T02:06:49Z mrottenkolber left #ccl 2016-07-03T02:16:13Z rme joined #ccl 2016-07-03T02:28:43Z rumbler31 joined #ccl 2016-07-03T02:39:43Z rumbler31 quit (Ping timeout: 244 seconds) 2016-07-03T02:41:10Z rumbler31 joined #ccl 2016-07-03T03:52:03Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-03T04:11:20Z rumbler31 quit (Remote host closed the connection) 2016-07-03T04:22:37Z rme quit (Quit: rme) 2016-07-03T04:22:37Z rme quit (Quit: rme) 2016-07-03T05:13:36Z edgar-rft quit (Quit: edgar-rft) 2016-07-03T05:19:21Z trumbull joined #ccl 2016-07-03T05:23:17Z trumbull left #ccl 2016-07-03T06:05:23Z blender joined #ccl 2016-07-03T06:05:51Z blender quit (Client Quit) 2016-07-03T06:20:10Z edgar-rft joined #ccl 2016-07-03T07:26:07Z MetaHertz joined #ccl 2016-07-03T08:28:55Z MetaHertz quit (Ping timeout: 252 seconds) 2016-07-03T08:36:40Z MetaHertz joined #ccl 2016-07-03T10:14:57Z schjetne joined #ccl 2016-07-03T10:31:17Z rme joined #ccl 2016-07-03T11:27:56Z rme quit (Ping timeout: 185 seconds) 2016-07-03T11:31:46Z mdc_mobile quit (Ping timeout: 186 seconds) 2016-07-03T12:11:57Z eschatologist quit (Read error: Connection reset by peer) 2016-07-03T13:03:37Z rme quit (Quit: rme) 2016-07-03T14:12:53Z ekinmur joined #ccl 2016-07-03T14:30:03Z edgar-rft quit (Quit: edgar-rft) 2016-07-03T15:28:24Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-03T15:46:20Z eschatologist joined #ccl 2016-07-03T16:22:21Z rme joined #ccl 2016-07-03T17:22:03Z eschatol_ joined #ccl 2016-07-03T17:25:55Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-03T18:23:19Z eschatol_ quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-03T18:23:31Z eschatologist joined #ccl 2016-07-03T18:35:59Z eschatologist quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-03T20:09:43Z faheem joined #ccl 2016-07-03T20:52:45Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-03T21:37:50Z pierpa joined #ccl 2016-07-03T21:41:47Z eschatologist joined #ccl 2016-07-03T22:07:51Z mrottenkolber joined #ccl 2016-07-03T22:07:54Z mrottenkolber: Hi 2016-07-03T22:08:23Z mrottenkolber: I am trying to use oprofile to profile my app but I get the following: 2016-07-03T22:08:23Z mrottenkolber: sudo operf /usr/lib/ccl/lx86cl64 -n -I profiled-ccl 2016-07-03T22:08:23Z mrottenkolber: operf: Profiler started 2016-07-03T22:08:23Z mrottenkolber: profiled app killed by signal 11 2016-07-03T22:08:46Z mrottenkolber: any ideas? 2016-07-03T22:09:33Z mrottenkolber: ccl 1.10, Linux 4.4.14, oprofile-1.1.0 2016-07-03T22:11:17Z mrottenkolber: bitrot in ccl::write-elf-symbols-to-file? 2016-07-03T22:11:30Z mrottenkolber: incompatible oprofile version? 2016-07-03T22:21:04Z edgar-rft joined #ccl 2016-07-03T23:33:52Z rumbler31 joined #ccl 2016-07-04T00:16:06Z rumbler31 quit (Ping timeout: 246 seconds) 2016-07-04T00:29:14Z rumbler31 joined #ccl 2016-07-04T01:17:32Z ekinmur joined #ccl 2016-07-04T01:36:36Z mrottenkolber quit (Ping timeout: 246 seconds) 2016-07-04T02:29:56Z rme quit (Quit: rme) 2016-07-04T03:32:33Z rumbler31 quit (Ping timeout: 240 seconds) 2016-07-04T03:40:18Z nocd joined #ccl 2016-07-04T03:51:49Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-04T04:05:15Z rumbler31 joined #ccl 2016-07-04T05:07:21Z pierpa quit (Quit: ERC Version 5.3 (IRC client for Emacs)) 2016-07-04T05:42:08Z DavidGu joined #ccl 2016-07-04T07:22:20Z MetaHertz joined #ccl 2016-07-04T07:25:34Z DavidGu quit (Ping timeout: 258 seconds) 2016-07-04T07:26:57Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-04T07:42:38Z DavidGu joined #ccl 2016-07-04T09:55:46Z okflo joined #ccl 2016-07-04T10:32:10Z hhdave joined #ccl 2016-07-04T11:19:21Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-04T11:57:10Z mrottenkolber joined #ccl 2016-07-04T12:15:31Z nocd quit (Read error: Connection reset by peer) 2016-07-04T12:58:55Z rumbler31 joined #ccl 2016-07-04T13:05:17Z milanj joined #ccl 2016-07-04T13:30:26Z MetaHertz joined #ccl 2016-07-04T13:36:31Z DavidGu quit (Ping timeout: 240 seconds) 2016-07-04T13:52:44Z rme joined #ccl 2016-07-04T13:56:42Z nobody_ joined #ccl 2016-07-04T14:05:05Z nobody_: xxx 2016-07-04T14:06:50Z nobody_ left #ccl 2016-07-04T14:08:56Z DavidGu joined #ccl 2016-07-04T14:24:02Z DavidGu quit (Ping timeout: 260 seconds) 2016-07-04T14:43:33Z DavidGu joined #ccl 2016-07-04T14:43:52Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-04T14:52:11Z pierpa joined #ccl 2016-07-04T14:52:33Z ekinmur joined #ccl 2016-07-04T14:52:34Z pierpa quit (Remote host closed the connection) 2016-07-04T14:56:33Z pierpa joined #ccl 2016-07-04T15:16:09Z mrottenkolber left #ccl 2016-07-04T15:17:33Z milanj quit (Quit: This computer has gone to sleep) 2016-07-04T15:22:26Z rumbler31 joined #ccl 2016-07-04T15:37:35Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-04T15:45:25Z milanj joined #ccl 2016-07-04T16:17:31Z DavidGu quit (Ping timeout: 240 seconds) 2016-07-04T16:19:07Z logrus joined #ccl 2016-07-04T16:40:40Z hhdave quit (Ping timeout: 264 seconds) 2016-07-04T16:51:05Z logrus quit (Quit: Leaving) 2016-07-04T16:54:28Z raphaelss joined #ccl 2016-07-04T16:56:01Z raphaelss left #ccl 2016-07-04T16:56:39Z diphuser joined #ccl 2016-07-04T17:46:17Z DavidGu joined #ccl 2016-07-04T17:50:38Z milanj quit (Quit: This computer has gone to sleep) 2016-07-04T17:58:38Z hhdave joined #ccl 2016-07-04T18:00:27Z okflo quit (Ping timeout: 260 seconds) 2016-07-04T18:01:48Z okflo joined #ccl 2016-07-04T18:02:13Z hhdave_ joined #ccl 2016-07-04T18:03:48Z hhdave quit (Ping timeout: 276 seconds) 2016-07-04T18:03:48Z hhdave_ is now known as hhdave 2016-07-04T18:13:04Z DavidGu quit (Ping timeout: 264 seconds) 2016-07-04T18:30:41Z hhdave quit (Quit: hhdave) 2016-07-04T18:52:35Z eschatologist quit (Excess Flood) 2016-07-04T18:54:38Z eschatologist joined #ccl 2016-07-04T19:08:25Z hhdave joined #ccl 2016-07-04T19:17:58Z hhdave quit (Quit: hhdave) 2016-07-04T19:24:17Z ekinmur joined #ccl 2016-07-04T19:30:10Z kpoeck joined #ccl 2016-07-04T19:43:03Z MetaHertz quit (Ping timeout: 240 seconds) 2016-07-04T19:45:43Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-04T20:00:26Z okflo quit (Remote host closed the connection) 2016-07-04T20:14:38Z ekinmur joined #ccl 2016-07-04T20:17:48Z ekinmur quit (Client Quit) 2016-07-04T20:40:01Z ekinmur joined #ccl 2016-07-04T20:58:46Z kpoeck quit (Ping timeout: 250 seconds) 2016-07-04T21:02:08Z kpoeck joined #ccl 2016-07-04T21:08:10Z hhdave joined #ccl 2016-07-04T21:13:55Z logrus joined #ccl 2016-07-04T21:16:02Z hydan joined #ccl 2016-07-04T21:17:48Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-04T21:23:25Z rumbler31 quit (Read error: Connection reset by peer) 2016-07-04T21:23:42Z rumbler31 joined #ccl 2016-07-04T21:32:34Z kpoeck quit (Ping timeout: 250 seconds) 2016-07-04T21:44:24Z rumbler31: i know office hours are tomorrow, but I got a question. http://paste.lisp.org/display/319836 should this produce a type error? a user on linux says that it does. on osx I can't reproduce. what is the correct behavior? 2016-07-04T22:06:00Z sixbitslacker: i am on linux and it didn't throw an error for me 2016-07-04T22:06:31Z sixbitslacker: 1.12-dev, built from the repo 2016-07-04T22:08:11Z sixbitslacker: sbcl 1.3.4 throws an error, though, on make-instance. 2016-07-04T22:08:13Z sixbitslacker: unknown type specifier: :STRING 2016-07-04T22:08:58Z rumbler31: hmm 2016-07-04T22:10:06Z rumbler31: ok, when I get the code the linux guy was using i'll try it again 2016-07-04T22:10:07Z |3b|: did you mean 'string instead of :string? 2016-07-04T22:10:27Z |3b|: (or maybe just string, not sure if that is evaluated or not) 2016-07-04T22:10:37Z rumbler31: i didn't, but maybe he typed it into irc wrong but into the repl right 2016-07-04T22:11:40Z rumbler31: with 'string, no error either 2016-07-04T22:12:06Z rumbler31: with string, yes error 2016-07-04T22:12:15Z rumbler31: i was doing it wrong 2016-07-04T22:12:47Z sixbitslacker: yes, i get an error with string instead of :string in both ccl and sbcl 2016-07-04T22:12:52Z |3b| isn't sure if it is specified whether an undefined type there should error or not, but it is explicitly undefined behavior if you store anything other than the specified type in the slot 2016-07-04T22:13:15Z |3b|: which means it could error, crash, delete your hard drive, whatever 2016-07-04T22:13:39Z rumbler31: the code comment suggests that such should be done but I couldn't follow it to see if it actually was done 2016-07-04T22:14:15Z rumbler31: it is explicitly an error, which means it should happen no matter the safety level? 2016-07-04T22:14:49Z |3b|: "The consequences of attempting to store in a slot a value that does not satisfy the type of the slot are undefined." 2016-07-04T22:15:19Z rumbler31: how did you find that so quick? 2016-07-04T22:15:19Z |3b| is talking about CL in general though, if you are looking at something CCL specific they may have more specific expectations 2016-07-04T22:15:37Z hhdave quit (Quit: hhdave) 2016-07-04T22:15:46Z |3b| clicked on DEFCLASS in your link, scrolled down to :type 2016-07-04T22:17:42Z rumbler31: i didn't know it did that.... 2016-07-04T22:17:49Z rumbler31: i'm learning all kinds of things tonight 2016-07-04T22:18:47Z milanj joined #ccl 2016-07-04T22:19:15Z rumbler31: paste.lisp.org, more than just a pretty frontend 2016-07-04T22:29:11Z milanj quit (Quit: This computer has gone to sleep) 2016-07-04T22:33:35Z sixbitslacker: previously... http://ircbrowse.net/browse/lisp?events_page=251909 2016-07-04T22:33:53Z sixbitslacker: came up in a search for the consequences sentence 2016-07-04T22:47:47Z hydan quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-04T23:57:21Z diphuser quit (Ping timeout: 250 seconds) 2016-07-05T00:58:55Z eschatologist_ joined #ccl 2016-07-05T01:00:38Z eschatologist_ quit (Client Quit) 2016-07-05T01:01:17Z eschatologist_ joined #ccl 2016-07-05T01:03:42Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-05T01:10:05Z eschatologist_ quit (Ping timeout: 258 seconds) 2016-07-05T01:11:49Z eschatologist_ joined #ccl 2016-07-05T01:11:52Z eschatologist quit (Ping timeout: 272 seconds) 2016-07-05T01:12:40Z eschatologist_ is now known as eschatologist 2016-07-05T01:30:56Z ekinmur joined #ccl 2016-07-05T02:26:16Z pierpa quit (Ping timeout: 264 seconds) 2016-07-05T03:04:51Z DougNYC joined #ccl 2016-07-05T03:25:41Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-05T04:06:17Z DavidGu joined #ccl 2016-07-05T04:22:20Z DavidGu quit (Ping timeout: 250 seconds) 2016-07-05T04:32:57Z DougNYC quit 2016-07-05T06:23:16Z rme quit (Quit: rme) 2016-07-05T06:23:16Z rme quit (Quit: rme) 2016-07-05T06:42:29Z DavidGu joined #ccl 2016-07-05T07:08:15Z milanj joined #ccl 2016-07-05T07:12:10Z logrus quit (Quit: Leaving) 2016-07-05T07:18:22Z MetaHertz joined #ccl 2016-07-05T07:56:50Z hhdave joined #ccl 2016-07-05T08:58:47Z taij33n joined #ccl 2016-07-05T09:07:03Z hydan joined #ccl 2016-07-05T10:05:24Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-05T10:31:51Z hydan quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-05T10:32:17Z hydan joined #ccl 2016-07-05T10:32:54Z hydan quit (Client Quit) 2016-07-05T10:50:45Z edgar-rft quit (Quit: edgar-rft) 2016-07-05T10:53:15Z MetaHertz joined #ccl 2016-07-05T10:55:11Z DavidGu quit (Ping timeout: 240 seconds) 2016-07-05T11:03:02Z okflo joined #ccl 2016-07-05T11:18:55Z okflo is now known as okflo_ 2016-07-05T11:19:30Z okflo_ is now known as okflo 2016-07-05T11:44:35Z knobo1 joined #ccl 2016-07-05T11:51:55Z lisper29 joined #ccl 2016-07-05T12:03:47Z hydan joined #ccl 2016-07-05T12:05:00Z hydan is now known as hydan` 2016-07-05T12:16:29Z rumbler31 joined #ccl 2016-07-05T12:26:00Z mrottenkolber joined #ccl 2016-07-05T12:48:36Z hydan` is now known as hydan 2016-07-05T13:03:38Z wgslayer joined #ccl 2016-07-05T13:06:13Z rme joined #ccl 2016-07-05T13:10:55Z wgslayer quit (Quit: Leaving) 2016-07-05T13:19:36Z ekinmur joined #ccl 2016-07-05T13:28:39Z rumbler31 quit (Ping timeout: 264 seconds) 2016-07-05T13:32:57Z rumbler31 joined #ccl 2016-07-05T13:33:36Z nzambe joined #ccl 2016-07-05T13:37:35Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-05T13:37:41Z lisper29 quit (Quit: This computer has gone to sleep) 2016-07-05T13:44:59Z nzambe joined #ccl 2016-07-05T13:51:39Z WindyFu_ joined #ccl 2016-07-05T13:54:39Z WindyFu_ quit (Client Quit) 2016-07-05T14:00:36Z rme: Hi, everyone. I'm R. Matthew Emerson (that's where the "rme" comes from). I'm CCL committer. I used to be a Clozure Associates employee, but unfortunately, I'm not any more. So, I don't speak for Clozure Associates, obviously. 2016-07-05T14:00:50Z rme: I'm happy to talk about whatever you want as long as it relates to CCL somehow. 2016-07-05T14:01:08Z rme: I don't know everything, so please forgive me if I don't know the answer to your question. 2016-07-05T14:04:43Z mrottenkolber: Hi Matthew :) 2016-07-05T14:04:51Z rme: Hey, Max. 2016-07-05T14:09:39Z lisper29 joined #ccl 2016-07-05T14:10:45Z mrottenkolber: Is profiling using oprofile on Linux still supported? My distribution comes with oprofile-1.1.0 which has a different CLI interface than what's described in the CCL documentation (no opcontrol but operf instead), and when I try to profile a CCL image with elf headers prepended like described it segfaults. 2016-07-05T14:11:36Z mrottenkolber: Also if I do (save-application "/tmp/profiled-ccl" :prepend-kernel "/tmp/elf-symbols") I get a shared library, but really I want a standalone executable to profile, right? 2016-07-05T14:12:13Z rme: The profiling situation is a bit messed up. 2016-07-05T14:13:27Z rme: Recent kernels and perf versions are not fooled by our trick of making a heap image look sort of like an ELF file. 2016-07-05T14:13:58Z rme: Also, they only consider mapped memory from ELF files (and not just some random file, like a heap image) as a place where functions can reside. 2016-07-05T14:14:16Z mrottenkolber: I see 2016-07-05T14:15:24Z rme: gbyers fiddled with it a few months ago, but as far I know, using perf (or presumably oprofile) doesn't work. 2016-07-05T14:18:13Z rumbler31: Greetings Matthew 2016-07-05T14:18:40Z rme: perf would be nice if we could make it work. For instance, it supports the use of a map file to map functions to symbols, so we wouldn't have to write ELF symbols anywhere. 2016-07-05T14:18:45Z rme: Hello rumbler31. 2016-07-05T14:20:57Z rumbler31: I was in the debugger recently, using slime, and I tried to evaluate some code in the context of a selected stack frame, and i got a message similar to the effect of "this is not working right now" 2016-07-05T14:21:07Z rumbler31: forgive me if i'm mangling terms 2016-07-05T14:21:08Z rme: heh 2016-07-05T14:21:59Z rumbler31: I take it you know what I'm talking about :) 2016-07-05T14:22:12Z rme: The function in CCL that doe that is called eval-in-frame, I believe. 2016-07-05T14:22:39Z rme looks 2016-07-05T14:23:04Z edgar-rft joined #ccl 2016-07-05T14:24:56Z rme: apply-in-frame, actually 2016-07-05T14:26:42Z rme: I don't regocnize the error, no, I just thought the message was amusing. 2016-07-05T14:27:00Z rumbler31: ah, well I'll see if I can't reproduce it 2016-07-05T14:27:40Z rme: Sometimes the built-in debugger works when the slime debugger doesn't. 2016-07-05T14:28:12Z rumbler31: I've been getting the impression that I should be trying to track slime trunk as a general rule 2016-07-05T14:29:50Z rme: I'm a slime caveman, so I don't know it well. I don't think gbyers uses slime at all. 2016-07-05T14:30:06Z mrottenkolber: rme: so is contrib/huebner/advice-profiler still a thing? I have played with it a bit and seems to have endured its share of bitrot as well (it prominently displays every item twice) 2016-07-05T14:32:41Z rme: mrottenkolber: It may have bitrotted, I'm afraid. 2016-07-05T14:33:13Z phf: helmut has surrendered his linus role under "community" pressure, and imho the project started to degrade since. the only backend that's reliably up to date is sbcl in trunk 2016-07-05T14:38:37Z rme: mrottenkolber: I fiddled around with a crude userland-based sampling profiler a while ago. I pasted it as http://paste.lisp.org/display/319875 if you happen to feel like toying with it. It is crude: it just samples every millisecond, and records the name of the function that the program counter is in. 2016-07-05T14:38:52Z rme: It's very simple. 2016-07-05T14:41:19Z rumbler31: rme: Regarding development on windows, I played with the provided win32 example and found that only specific windows libraries are brought in to be understood by the ccl ffi framework, do you know of any reason why more of them aren't consumed as a regular part of the windows build step, other than perhaps a lack of demand/use/maintaner time? 2016-07-05T14:41:20Z rme: I've often wondered if I could make CCL work well with dtrace, but the dtrace on OS X (my main platform) doesn't support ustack helpers. http://darksi.de/3.dtrace-ustack-helper/ is something I have bookmarked. 2016-07-05T14:41:23Z WindyFU joined #ccl 2016-07-05T14:46:12Z mrottenkolber: rme: will take a look at it, sounds interesting 2016-07-05T14:46:13Z rme: rumbler31: Are you asking about the interface database that lets you do stuff like (#_CloseHandle h)? 2016-07-05T14:46:26Z rumbler31: rme: yes 2016-07-05T14:47:09Z lisper29 quit (Quit: This computer has gone to sleep) 2016-07-05T14:47:48Z rme: OK. Then yes, it's just a question of demand/interest/time. It's not too hard on Linux and OS X to build the interfaces; I'm not sure about Windows at the moment. 2016-07-05T14:48:04Z rumbler31: and on windows I believe that only a subset of the win32 libraries are brought in for consumption, and it looked like a full list was enumerated in the windows build scripts with most of them commented out, I didn't have a mingw or cygwin environment on that machine otherwise I would have messed with the build script to see what would happen 2016-07-05T14:48:13Z rumbler31: ok, thank you! 2016-07-05T14:49:09Z rme: rumbler31: There's a tool called ffigen4 that is a patched GCC front-end. It processes header files, and writes out s-expression formatted data about the foreign types and functions. 2016-07-05T14:49:33Z rme: See http://trac.clozure.com/ccl/wiki/BuildFFIGEN 2016-07-05T14:50:21Z rme: I'm not sure what the 64-bit Windows situation is now. At one point, we had to cross-compile 64-bit Windows interface databases. 2016-07-05T14:51:04Z rme: This is an area that needs some attention. Relying on an old gcc with patches is probably not a good long-term plan. 2016-07-05T14:51:48Z rumbler31: ahh thank you. 2016-07-05T14:52:03Z rme: Ryan Pavlik wrote a thing using clang's interntal C++ interface to do a similar task. https://github.com/rpav/c2ffi 2016-07-05T14:52:50Z rumbler31: my main workstation died, I keep bouncing from osx to windows and back, so sometimes I care about windows and sometimes I don't. Next time I get windows back I'll dig into this some more 2016-07-05T14:52:57Z rme: Maybe it would be a good idea to make that emit the s-expression based ffigen4 format. We need to do something in this area. 2016-07-05T14:55:34Z phf: rme: a more general question if you don't mind, do you have any personal near term goals with ccl that you're working on? 2016-07-05T14:55:37Z rme: I think it's valuable to support Windows, but it's also true that the Windows port doesn't really get the attention it deserves sometimes, since most (all?) ccl hackers don't run Windows as their main computing system. 2016-07-05T14:57:35Z rme: phf: The thing that I have been thinking about the most is making a good IDE for CCL (mainly targeting the Macintosh, probably). I don't know if that's the best idea, but that's the thing that's been on my mind. 2016-07-05T14:57:46Z rme: We need to do a 1.12 release fairly soon also. 2016-07-05T14:59:18Z fe[nl]ix: rme: does ccl move code objects ? 2016-07-05T14:59:20Z phf: oh right, it's your idea of rethinking the widget system so it's not all #/foo 2016-07-05T14:59:52Z rme: fe[nl]ix: yes, the gc will relocate functions. 2016-07-05T15:00:51Z rme: fe[nl]ix: On x86-64, this is no big deal because we use pc-relative addressing to access literals ("constants"). On x86-32, we have to fix up absolute addresses. 2016-07-05T15:01:13Z fe[nl]ix: precisely because of perf and profiling tools we're working on saving an SBCL core as ELF .o and create an executable with ld 2016-07-05T15:01:33Z fe[nl]ix: currently QPX cannot be effectively profiled and that sucks 2016-07-05T15:01:45Z rme: Yeah 2016-07-05T15:02:08Z fe[nl]ix: also, profiling tools expect C++ kind of code and cannot deal with relocating functions 2016-07-05T15:04:29Z rme: Even when we had oprofile/perf working, you were limited to loading up the code you wanted to profile, saving an image, and then starting up profiling from that image. It didn't work to add code at runtime and have the profile pick that up. 2016-07-05T15:04:52Z fe[nl]ix: yes, that's what we'll be doing 2016-07-05T15:05:46Z fe[nl]ix: fixing the relocating functions case is possible but rather tricky 2016-07-05T15:08:35Z rme: phf: Is there something on your wishlist, or were you just wondering? 2016-07-05T15:09:57Z phf: rme: we briefly talked about better widget system in the past, so i was just wonder, but also hoping that you might still be working on that :> 2016-07-05T15:14:27Z phf: i can even do grunt work of filling out details if there was a framework to fill them out into, i spent a lot of time making ccl/cocoa bearable, but for a recent project just bought capi 2016-07-05T15:15:08Z rme: After working on Opusmodus, believe me, I feel the need. 2016-07-05T15:18:06Z WindyFU quit (Quit: WindyFU) 2016-07-05T15:23:58Z rme: Thanks for coming to my office hours, everyone. 2016-07-05T15:24:37Z phf: oh oh on a personal note is your current work still lisp related? 2016-07-05T15:25:02Z rumbler31: rme: ty 2016-07-05T15:25:45Z rme: phf: I haven't yet started any new full-time work. 2016-07-05T15:27:51Z rme: Of course I very much want to keep working with CCL and CL if that's possible, but I guess we'll see. 2016-07-05T15:28:05Z phf: understood 2016-07-05T15:28:24Z jdz: rme: wish you luck finding that job! 2016-07-05T15:28:33Z jdz: i mean, sincerely 2016-07-05T15:29:09Z rumbler31 quit (Ping timeout: 246 seconds) 2016-07-05T15:29:33Z rme: Thanks, jdz. Please feel free to contact me or Clozure if you have CL or CCL work. 2016-07-05T15:29:46Z rme: (that goes for anyone, of course) 2016-07-05T15:30:16Z jdz: i sure will, after i leave my current job involving Python and Django, and found my own company. 2016-07-05T15:31:09Z mrottenkolber: :-) 2016-07-05T15:31:27Z mrottenkolber left #ccl 2016-07-05T15:31:27Z rme: heh 2016-07-05T15:32:35Z lisper29 joined #ccl 2016-07-05T15:37:59Z pierpa joined #ccl 2016-07-05T15:38:59Z gbyers: Rme: overslept, sorry. Hope it went well. 2016-07-05T15:39:50Z rme: It went well. Part two starts at 4pm Eastern, if you will happen to be around for that. 2016-07-05T15:53:16Z logrus joined #ccl 2016-07-05T15:54:46Z gbyers: will try. first time I've overslept in a very long time. 2016-07-05T16:13:55Z rumbler31 joined #ccl 2016-07-05T16:19:09Z gbyers: fwiw, I started playing around with porting ffigen to gcc 5.x at one point. It would be a lot of work, but did not seem hopeless. 2016-07-05T16:19:34Z pierpa: that's 20:00 UTC if google's not wrong (22:00 CE DST time) 2016-07-05T16:32:47Z edgar-rft quit (Quit: edgar-rft) 2016-07-05T16:33:40Z rumbler31: gbyers: can you talk a little bit more about ffigen's responsibilities? 2016-07-05T16:36:28Z hhdave quit (Ping timeout: 264 seconds) 2016-07-05T16:42:37Z gbyers: It parses .h files. It is sometimes called "the interface translator" in the ccl documentation. It is based on a very old version ofhe GCC frontend. Much of the ffi code in ccl uses the results of that translation. 2016-07-05T16:44:37Z rumbler31: ok 2016-07-05T17:05:47Z hhdave joined #ccl 2016-07-05T17:09:06Z hhdave_ joined #ccl 2016-07-05T17:10:32Z hhdave quit (Ping timeout: 260 seconds) 2016-07-05T17:10:32Z hhdave_ is now known as hhdave 2016-07-05T17:11:21Z hhdave quit (Client Quit) 2016-07-05T17:50:31Z milanj quit (Quit: This computer has gone to sleep) 2016-07-05T18:24:57Z milanj joined #ccl 2016-07-05T18:37:46Z okflo quit (Remote host closed the connection) 2016-07-05T18:58:26Z lisper29 left #ccl 2016-07-05T19:51:41Z lithper joined #ccl 2016-07-05T20:00:30Z rme: Hi everyone. Welcome to the second half of CCL office hours. 2016-07-05T20:00:40Z lithper: Hi RME! 2016-07-05T20:00:44Z pierpa: hullo 2016-07-05T20:01:00Z lithper: How did the first hour go? Did you have a good turnout? 2016-07-05T20:01:25Z rme: I'm here, and maybe gbyers is too. I'll be happy to talk about whatever you want, as long as it relates to CCL somehow. 2016-07-05T20:01:36Z rme: I think it went well. Channel logs are here: http://ccl.clozure.com/irc-logs/ccl/ccl-2016-07.txt 2016-07-05T20:02:34Z gbyers: I'll need to step out later, i think. 2016-07-05T20:03:36Z lithper: Hm, surprised and disappointed to learn that rme is no longe a Clozure employee 2016-07-05T20:04:00Z rme: Well, business is cyclical. It's a bummer to be sure. 2016-07-05T20:04:39Z lithper: What are you working on nowadays? 2016-07-05T20:06:17Z lithper: Hm, “The thing that I have been thinking about the most is making a good IDE for CCL (mainly targeting the Macintosh,” 2016-07-05T20:06:33Z lithper: What’s wrong with the current IDE? 2016-07-05T20:07:57Z rme: I've been a little discouraged, but I'm starting to cheer up. I'm trying to give CCL some love while I'm looking for work. 2016-07-05T20:09:15Z rme: My work on Opusmodus has convinced me that a higher-level interface to Cocoa is highly desireable. Also, as part of that, I think that the editor needs to be more directly programmable in Lisp. It's currently kind of mixed up with the Cocoa text system. 2016-07-05T20:09:34Z rme: That makes it hard to program the editor, say for custom mouse commands and so forth. 2016-07-05T20:10:02Z lithper: Yeah, Hemlock is a bit of a PITA. LMK if you decide to dive into that if you’d like some help. I’ve done some Hemlock hacking in the past. 2016-07-05T20:10:36Z lithper: GB, are you here? 2016-07-05T20:10:39Z rme: Thanks. I need to stop thinking at some point, and start hacking. 2016-07-05T20:11:06Z gbyers: yes 2016-07-05T20:11:22Z lithper: Very sorry about the fire. Glad you’re OK. (Are you OK?) 2016-07-05T20:13:58Z gbyers: I wasn't hurt, but lost everything that I owned. It could have been worse. but I would not wish it on anyone, 2016-07-05T20:14:51Z lithper: Wow, that sucks. What started it if you don’t mind my asking? 2016-07-05T20:15:41Z rumbler31: lithper: I've seen some people ask about a paredit editing mode for the ccl hemlock editor. it seems like core hemlock commands are there to do much of the work, are you aware of anything like this already in existence? 2016-07-05T20:15:41Z rumbler31: oh wow 2016-07-05T20:16:01Z gbyers: it was an electrical fire. 2016-07-05T20:16:48Z lithper: I’ve never used paredit (not even sure what it is). I’ve done some noodling around with paren balancing in the past because I wanted to balance [] and {}. It’s non-trivial. 2016-07-05T20:16:50Z rme: By the way, everyone, if you want to say something, just jump right in. 2016-07-05T20:17:56Z lithper: @gb: serous bummer about the CCL sources. I hope another copy turns up some day. 2016-07-05T20:18:37Z rme: The last time I held office hours, I think I was teaching one of the last sections of VAX assembly language. I remember being pretty bored when nobody showed up, so don't be shy. 2016-07-05T20:18:38Z gbyers: you mean MCL sources>? 2016-07-05T20:18:56Z lithper: Yeah 2016-07-05T20:19:56Z rumbler31: lithper: its a mode in emacs that helps with managing parens and sexps, like you can expand-contract sexps, or make a new sexp that surrounds a number of sexps forward/backwards of the cursor, automatically closes parens and puts the cursor inside, etc 2016-07-05T20:21:16Z lithper: @rumbler31: Yeah, I see. I looked it up. There’s no reason you couldn’t do something like that in Hemlock, but the problem is that Hemlock has this really weird internal representation of text that is not at all easy to deal with. But if you want this, why not just use emacs and slime? 2016-07-05T20:22:30Z lithper: You might be able to do an 80/20 solution with the right keybindings for Hemlock built-ins for moving around and cutting/pasting sexprs 2016-07-05T20:22:32Z rumbler31: lithper: I use emacs/slime, just curious about the native ccl editor on osx 2016-07-05T20:22:57Z lithper: Nothing like that built in AFAIK. RME? 2016-07-05T20:23:04Z PuercoPop: rme: Do you know if Opus Modus can work as a library in non OSX plataforms? 2016-07-05T20:23:08Z PuercoPop: *platforms 2016-07-05T20:24:28Z rme: PuercoPop: As far as I know, Opusmodus is only distributed as a complete application. The non-UI parts of it could fairly easily be separated out. 2016-07-05T20:24:37Z hydan quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-05T20:24:44Z rme: But I don't know if the client is interested in doing that. You could send mail to him and ask. 2016-07-05T20:25:41Z lithper: Wow, never heard of Opusmodus before. That looks really cool! 2016-07-05T20:25:47Z PuercoPop: rme: will do, thanks. 2016-07-05T20:25:47Z rme: You'd might want some non-OS X way to send MIDI, unless you just wanted to produce MIDI files and play them wherever. 2016-07-05T20:26:44Z rme: PuercoPop: He's quite enthusiastic and cordial. 2016-07-05T20:27:14Z lithper: OK, I have a question: has anyone here looked at urbit? 2016-07-05T20:28:08Z gbyers: rme: good luck. ttyl. 2016-07-05T20:28:15Z rme: The coolest sexp-based Hemlock command I know of that's not bound to a key is Extract Form. 2016-07-05T20:28:17Z rme: See you gbyers. 2016-07-05T20:28:24Z lithper: TTFN GB! 2016-07-05T20:28:44Z rme: I think emacs has something like it called raise-sexp or something. 2016-07-05T20:29:01Z lithper: Re urbit: I know that doesn’t sound like a CCL question but bear with me. 2016-07-05T20:29:08Z mrottenkolber joined #ccl 2016-07-05T20:31:50Z rme: I'm ignorant about urbit. 2016-07-05T20:32:47Z logrus quit (Quit: Leaving) 2016-07-05T20:33:52Z lithper: OK, well, maybe we should talk about this off-line or on the newsgroup then. But urbit is a new social-networky sort of thing that is also a self-described attempt to reinvent computing from the ground up. It’s based on these two new languages called Nock and Hoon. 2016-07-05T20:34:24Z lithper: I would have written the whole thing off as completely bogus except that they did the equivalent of a kickstarter a few days ago and raised $200k in two days. 2016-07-05T20:36:06Z lithper: That kind of traction ought to attract a competitor (ahem) 2016-07-05T20:37:33Z pierpa: completely bogus :) 2016-07-05T20:41:39Z lithper: OK, well, that went over like a lead balloon. Next topic… 2016-07-05T20:41:56Z lithper: What would it take to get Hemlock running on Windows? 2016-07-05T20:42:58Z rme: I think the Cocotron approach is a fool's errand. 2016-07-05T20:43:57Z lithper: That’s too bad, because that would seem to be the path of least resistance if it had worked 2016-07-05T20:44:12Z lithper: Not that I doubt you, but what’s wrong with Cocotron? 2016-07-05T20:45:05Z mrottenkolber: rme: I played around with your tiny profiler and I like it :-) I don't think I understand how it works though, I would have expected the total number of samples to increase when I decrease the sleep duration? 2016-07-05T20:45:29Z rme: It's just too hard to duplicate all of Cocoa. It's hard enough to get stuff working with Cocoa, but it's harder still to debug problerms when the problem might be in Cocotron itself. 2016-07-05T20:46:00Z rme: And Coctron is 32-bit only, which is a big flaw nowadays. 2016-07-05T20:47:30Z rme: mrottenkolber: I assume that there's a limit to how fast ccl::map-call-frames can do its thing. 2016-07-05T20:47:54Z lithper: OK, so what will work? Isn’t Hemlock supposed to be display-agnostic? Can it just be hooked into a native Windows text window? 2016-07-05T20:48:33Z rme: mrottenkolber: It suspends the thead, examines it, and then resumes it. 2016-07-05T20:49:31Z mrottenkolber: I see 2016-07-05T20:49:51Z lithper: Wow, this is kind of cool! When I click on ccl::map-call-frames it invokes CCL! From Colloquy! 2016-07-05T20:50:12Z rme: The Hemlock in the IDE has gotten a bit intertwined with the rest of the IDE. I would probably start anew from portable Hemlock. 2016-07-05T20:50:34Z rme: I think that was sellout who implemented that ccl: trick. It's kind of cute, I agree. 2016-07-05T20:51:56Z rme: mrottenkolber: Like I say, that profiler is just me playing around. I think it tells the truth, but maybe I am doing it wrong. 2016-07-05T20:52:02Z lithper: Where do ccl: events get delivered? I would love to be able to write a proper handler for ccl: links. Lots of interesting possibilities there. 2016-07-05T20:52:50Z lithper: (And lots of potential security holes too!) 2016-07-05T20:53:53Z rme: lithper: See ccl:cocoa-ide;cocoa-editor.lisp and look for "defclass url-handler-command". 2016-07-05T20:57:03Z lithper: Cool! Thanks! 2016-07-05T21:03:23Z rme: The other thing I would say about making progress on Windows is that someone would need to invest in learning it well and using it. 2016-07-05T21:04:32Z phf: there's a sampling profiler in cmu ai archives, i've not looked at it, but it claims to be portable (with just these 3 system specific forms!!!) 2016-07-05T21:04:43Z rme: I feel less antipathy towards MS and Windows now that they've been taken down a peg, but still, I'm really not at home on Windows. 2016-07-05T21:04:54Z phf: http://www.cs.cmu.edu/Groups/AI/util/lang/lisp/code/tools/metering/metering.cl 2016-07-05T21:06:53Z lithper: It’s been an hour, and I don’t know if you’re sticking around or not, but in case not, thanks for doing this RME! 2016-07-05T21:07:06Z rme: I'm not 100% happy with OS X, but it's still better than anything else around, at least at the moment. 2016-07-05T21:07:23Z rme: My pleasure. I hope it's fun/useful. Maybe I'll do it again. 2016-07-05T21:07:39Z lithper: +1 to that sentiment. Say, what about hemlock on Linux? 2016-07-05T21:08:11Z lithper: Apple does seem to be in the process of jumping the shark so I need a backup plan 2016-07-05T21:08:17Z rme: I'm uncertain what approach would be best to take on Linux. 2016-07-05T21:09:14Z rme: Basic X11? GTK something? Qt? Wayland? It's pretty confusing. 2016-07-05T21:09:27Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-05T21:09:53Z lithper: Yeah, that’s pretty much as far as I got too. You’d think that building a basic text editor on the world 2016-07-05T21:10:03Z lithper: ’s third most popular OS would be a solved problem 2016-07-05T21:10:31Z rme: Motif/Xt was the last thing I used in that world, and it was horrible. Truly hideous. 2016-07-05T21:11:14Z mrottenkolber: Probably best to use GTK3 if you really want to target Linux, Qt is super cross platform but a beast 2016-07-05T21:11:24Z mrottenkolber: did you see cl-ui? 2016-07-05T21:11:54Z mrottenkolber: its alpha/broken but th underlying C library seems *neat* 2016-07-05T21:12:32Z rme: My impression is that basic X11 would be the most straightforward, but it's not even clear how you're supposed to draw text now. Apparently it's no longer cool to use XDrawString; you're supposed to use the RENDER extension or something. What a mess. 2016-07-05T21:12:41Z rme: mrottenkolber: link? 2016-07-05T21:12:43Z mrottenkolber: its a binding for libui: https://github.com/andlabs/libui#screenshots 2016-07-05T21:13:06Z mrottenkolber: https://github.com/jinwoo/cl-ui <- the (broken) CL binding 2016-07-05T21:13:17Z lithper: Hm, that looks very promising 2016-07-05T21:13:32Z lithper: Thx for the pointer! 2016-07-05T21:13:47Z mrottenkolber: I tested it on the weekend and it segfaults my CCL more often than not but I got a window that was responsive and all 2016-07-05T21:14:30Z rme: I'm kind of anti-cross-platform, but I'm bookmarking that. 2016-07-05T21:14:31Z mrottenkolber: if cl-ui was stable and had a team it would be beautiful 2016-07-05T21:15:02Z mrottenkolber: thing is it apparently looks native on all platforms (on linux it did for me) and the C lib seems tiny 2016-07-05T21:15:36Z lithper: If the C is solid then fixing the bindings should be relatively easy 2016-07-05T21:15:58Z mrottenkolber: yeah its a fresh project, but cl-ui has huge potential 2016-07-05T21:16:00Z phf: let's do more ffi from lisp, that's going to improve things around here 2016-07-05T21:16:27Z mrottenkolber: not sure, don't like segfaults 2016-07-05T21:16:37Z phf: woosh 2016-07-05T21:16:47Z mrottenkolber: FFI use should be limited and very well tested ;-) 2016-07-05T21:17:36Z rme: I did the Opusmodus UI via the FFI. It hurt a lot. 2016-07-05T21:18:16Z lithper: Which FFI? The actual C FFI? Or the ObjC bridge? And why did it hurt? 2016-07-05T21:18:40Z rme: Some of the pain was probably self-inflicted due to incompetence, but still. 2016-07-05T21:19:36Z rme: It uses the Objective-C bridge. Memory management (Objective-C uses reference couting) is probably the biggest challenge. 2016-07-05T21:20:50Z rme: You can say (make-instance 'ns:ns-mutable-array), for instance, but even though that looks "normal", the returned object is a foreign object, and you have to remember to release it. 2016-07-05T21:21:43Z rme: And the "do UI stuff on the main thread" restriction that Cocoa has is also sometimes a bother. 2016-07-05T21:21:57Z gbyers: ARC is left as an exercise. 2016-07-05T21:22:12Z lithper: ARC? 2016-07-05T21:22:43Z rme: Apple's compiler uses static analysis and naming conventions to determine when to automatically insert retains and releases. 2016-07-05T21:22:47Z phf: capi has same restriction (code needs to run on gui thread), and they have a handy set of utility functions that make it almost invisible 2016-07-05T21:22:54Z rme: s/compiler/C compiler/ 2016-07-05T21:23:07Z lithper: Naming conventions??? Really? 2016-07-05T21:23:12Z rme: No lie. 2016-07-05T21:23:58Z rme: If a method name starts with copy, e.g., copyFoo, then the compiler assumes that it will need to relase the returned reference. 2016-07-05T21:24:04Z rme: As an example. 2016-07-05T21:24:34Z lithper: Wow. Pardon me while I pick my jaw up off the floor. 2016-07-05T21:25:03Z MetaHertz quit (Ping timeout: 276 seconds) 2016-07-05T21:25:03Z lithper: Re memory management: whenever I’ve run into those kinds of issues I just wrap my objc objects in Lisp wrappers that have finalize methods that release the ObjC object. 2016-07-05T21:25:21Z rme: I don't know as much (well, hardly anything) about how the Swift compiler does it, but it's the same idea: the compiler manages the reference counting as needed. 2016-07-05T21:25:25Z phf: and then you're going to run out of heap, because gc doesn't know about your allocation sizes 2016-07-05T21:25:30Z gbyers: Apple tried at one point to implenent a GC for objc, it is best forgotten, 2016-07-05T21:26:04Z phf: using finalizers for foreign heap free's is a Bad Idea 2016-07-05T21:26:15Z lithper: Why? 2016-07-05T21:26:27Z phf: lithper: read my immediate previous message 2016-07-05T21:27:12Z lithper: “gc doesn't know about your allocation sizes” — sorry, I still don’t get it. 2016-07-05T21:27:33Z rme: It's less bad with reference counting. With something like malloc, you have to be double-dog sure that nobody (like in foreign code) is hanging on to a pointer to the malloc'd memory. 2016-07-05T21:28:38Z phf: lithper: well, because gc doesn't know anything about the nature of the data on the other side. what size it is, what it references, what references it. all it sees is a tiny wrapper with a number in it 2016-07-05T21:28:40Z mrottenkolber: Sorry to interrupt: what are the ccl internal functions to find out source locations of functions? 2016-07-05T21:28:47Z lithper: Yeah, I should not have mentioned malloc. I was using the term metaphorically. I was assuming ObjC and that all allocations were done using ObjC constructors invoked by (make-instance ’ns:whatever) 2016-07-05T21:29:03Z mrottenkolber is hacking on rme's profiler 2016-07-05T21:29:44Z lithper: @phf: OK, I get that, but I still don’t see what that would be a problem. I consider all ObjC objects to be opaque. 2016-07-05T21:29:46Z phf: so if gc collects an instance that free's a corresponding ~live~ object, you get a very nice heizenbug 2016-07-05T21:29:50Z lithper: At least when coding in Lisp. 2016-07-05T21:30:15Z lithper: OK, I understand what you’re saying. So let me clarify: 2016-07-05T21:30:18Z milanj quit (Quit: This computer has gone to sleep) 2016-07-05T21:30:26Z rme: I think it's called function-source-location. You might want to look at the swank backend for ccl; I know it calls exported functions that do the job. 2016-07-05T21:30:31Z phf: lithper: it's not up for a ~debate~, you asked i answered 2016-07-05T21:31:48Z lithper: Yes, I understand. But I wanted to clarify for the sake of others who may be lurking. 2016-07-05T21:32:53Z rme: mrottenkolber: Sorry, it's function-source-note 2016-07-05T21:33:17Z lithper: I don’t wrap memory blocks that are returned by calling malloc. I actually never call malloc from Lisp. I wrap ObjC objects that are returned by ObjC constructors invoked through calls to MAKE-INSTANCE 2016-07-05T21:33:36Z lithper: Those are reference-counted, and using finalizers to release (not free) them works just fine. 2016-07-05T21:33:37Z emaczen quit (Ping timeout: 260 seconds) 2016-07-05T21:33:57Z lithper: At least in my experience 2016-07-05T21:34:10Z rme: OK, last call for office hours! 2016-07-05T21:36:22Z rme: Thank you everyone. See you next time. (I'm often in the channel at other times, so feel free to drop in any time.) 2016-07-05T21:36:56Z mrottenkolber: was great, thanks! 2016-07-05T21:37:14Z lithper: +1! 2016-07-05T21:37:21Z gbyers: thanks, rme! 2016-07-05T21:37:47Z lithper left #ccl 2016-07-05T21:45:47Z fila joined #ccl 2016-07-05T21:47:41Z mrottenkolber left #ccl 2016-07-05T21:59:18Z fila left #ccl 2016-07-05T22:06:01Z rme quit (Ping timeout: 252 seconds) 2016-07-05T22:07:40Z rumbler31 joined #ccl 2016-07-05T22:11:01Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-05T22:29:10Z edgar-rft joined #ccl 2016-07-05T22:36:56Z edgar-rft quit (Quit: edgar-rft) 2016-07-05T22:53:58Z milanj joined #ccl 2016-07-06T00:09:41Z milanj quit (Quit: This computer has gone to sleep) 2016-07-06T01:04:26Z ekinmur joined #ccl 2016-07-06T01:04:35Z ekinmur quit (Client Quit) 2016-07-06T01:16:34Z ekinmur joined #ccl 2016-07-06T01:20:56Z arthurc_ joined #ccl 2016-07-06T01:23:01Z arthurc_ left #ccl 2016-07-06T02:11:13Z rumbler31 quit (Ping timeout: 258 seconds) 2016-07-06T02:14:24Z rumbler31 joined #ccl 2016-07-06T02:15:38Z logrus joined #ccl 2016-07-06T02:42:17Z DavidGu joined #ccl 2016-07-06T02:49:58Z pierpa quit (Ping timeout: 250 seconds) 2016-07-06T02:59:44Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-06T03:03:16Z DavidGu joined #ccl 2016-07-06T04:05:11Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T04:08:32Z rumbler31 quit (Remote host closed the connection) 2016-07-06T04:14:43Z edgar-rft joined #ccl 2016-07-06T04:57:28Z Hank joined #ccl 2016-07-06T05:34:22Z logrus quit (Quit: Leaving) 2016-07-06T05:49:39Z eschatol_ joined #ccl 2016-07-06T05:49:56Z eschatol_ quit (Client Quit) 2016-07-06T06:52:24Z Hank quit (Ping timeout: 250 seconds) 2016-07-06T07:11:12Z MetaHertz joined #ccl 2016-07-06T07:32:51Z Hank joined #ccl 2016-07-06T07:51:49Z milanj joined #ccl 2016-07-06T08:49:53Z MetaHertz quit (Ping timeout: 258 seconds) 2016-07-06T08:50:21Z DavidGu quit (Quit: DavidGu) 2016-07-06T08:50:49Z DavidGu joined #ccl 2016-07-06T09:52:10Z hydan joined #ccl 2016-07-06T11:18:54Z Hank quit (Ping timeout: 250 seconds) 2016-07-06T11:53:12Z DavidGu quit (Quit: DavidGu) 2016-07-06T13:00:53Z ekinmur joined #ccl 2016-07-06T13:19:21Z rme joined #ccl 2016-07-06T13:32:08Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T13:38:24Z ekinmur joined #ccl 2016-07-06T13:52:48Z edgar-rft quit (Quit: edgar-rft) 2016-07-06T14:15:43Z schjetne quit (Ping timeout: 258 seconds) 2016-07-06T14:26:29Z MetaHertz joined #ccl 2016-07-06T14:44:15Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T14:49:15Z schjetne joined #ccl 2016-07-06T14:54:50Z ekinmur joined #ccl 2016-07-06T15:14:20Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T15:18:26Z ekinmur joined #ccl 2016-07-06T17:06:55Z logrus joined #ccl 2016-07-06T17:33:15Z rme quit (Quit: rme) 2016-07-06T17:33:15Z rme quit (Quit: rme) 2016-07-06T17:34:24Z rme joined #ccl 2016-07-06T17:37:00Z edgar-rft joined #ccl 2016-07-06T17:44:52Z pierpa joined #ccl 2016-07-06T18:06:08Z rumbler31 joined #ccl 2016-07-06T18:15:23Z edgar-rft quit (Quit: edgar-rft) 2016-07-06T19:30:52Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-06T19:31:26Z nzambe joined #ccl 2016-07-06T19:48:16Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T19:50:12Z ekinmur joined #ccl 2016-07-06T20:03:42Z hydan quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T20:05:15Z hydan joined #ccl 2016-07-06T20:10:04Z hydan quit (Ping timeout: 264 seconds) 2016-07-06T20:47:32Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-06T21:58:04Z rme: It's kind of sad that I've let tickets stay in the Trac for years. 2016-07-06T22:00:16Z rme: There are some that are 9 years old. 2016-07-06T22:03:38Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-06T22:06:42Z rumbler3_ joined #ccl 2016-07-06T22:09:27Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-06T22:17:19Z DougNYC joined #ccl 2016-07-06T22:55:02Z lisper29 joined #ccl 2016-07-06T23:20:42Z lisper29 quit (Quit: This computer has gone to sleep) 2016-07-06T23:27:21Z lisper29 joined #ccl 2016-07-06T23:57:08Z lisper29 quit (Quit: Leaving) 2016-07-07T00:31:03Z ekinmur joined #ccl 2016-07-07T00:31:21Z ekinmur quit (Client Quit) 2016-07-07T01:40:24Z ekinmur joined #ccl 2016-07-07T01:45:08Z logrus quit (Quit: Leaving) 2016-07-07T02:08:09Z rumbler31 joined #ccl 2016-07-07T02:10:40Z rumbler3_ quit (Ping timeout: 264 seconds) 2016-07-07T02:22:44Z DavidGu joined #ccl 2016-07-07T02:50:23Z edgar-rft joined #ccl 2016-07-07T03:28:58Z milanj quit (Quit: This computer has gone to sleep) 2016-07-07T03:42:12Z DougNYC quit 2016-07-07T03:42:16Z rumbler31 quit (Remote host closed the connection) 2016-07-07T04:01:41Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T04:26:09Z edgar-rft quit (Quit: edgar-rft) 2016-07-07T05:16:36Z Honza_ joined #ccl 2016-07-07T05:18:01Z Honza_ quit (Client Quit) 2016-07-07T05:36:18Z edgar-rft joined #ccl 2016-07-07T06:11:08Z edgar-rft quit (Quit: edgar-rft) 2016-07-07T06:22:18Z rme quit (Quit: rme) 2016-07-07T06:36:38Z DavidGu1 joined #ccl 2016-07-07T06:38:12Z DavidGu quit (Ping timeout: 258 seconds) 2016-07-07T06:38:12Z DavidGu1 is now known as DavidGu 2016-07-07T06:51:42Z milanj joined #ccl 2016-07-07T07:03:03Z MetaHertz joined #ccl 2016-07-07T07:14:07Z Quadrescence quit (Read error: Connection reset by peer) 2016-07-07T07:14:53Z Quadrescence joined #ccl 2016-07-07T08:13:09Z hydan joined #ccl 2016-07-07T08:41:15Z pierpa quit (Ping timeout: 264 seconds) 2016-07-07T09:26:48Z hhdave joined #ccl 2016-07-07T10:11:08Z milanj quit (Quit: This computer has gone to sleep) 2016-07-07T10:19:05Z milanj joined #ccl 2016-07-07T10:25:58Z rme joined #ccl 2016-07-07T10:55:30Z edgar-rft joined #ccl 2016-07-07T11:08:14Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-07T11:55:19Z DavidGu quit (Remote host closed the connection) 2016-07-07T12:20:54Z rme: Good morning. 2016-07-07T12:29:37Z edgar-rft quit (Quit: edgar-rft) 2016-07-07T12:31:19Z lisper29` joined #ccl 2016-07-07T12:35:36Z jackdaniel: hey o/ 2016-07-07T12:44:35Z MetaHertz joined #ccl 2016-07-07T13:18:44Z ekinmur joined #ccl 2016-07-07T14:08:30Z mrottenkolber joined #ccl 2016-07-07T14:10:51Z mrottenkolber: So I am playing around with a combinatory parsing library I am working on, testing various “optimizations”: my theory was less consing == faster execution, but my tests do not support that 2016-07-07T14:12:00Z phf: mrottenkolber: and for good reason, consing is ~cheap~ if you think about it. the expensive part of consing is GC 2016-07-07T14:12:42Z rme: lukego gave a talk once where he advocated consing like crazy for performance. I think it was at some ELS or other. 2016-07-07T14:13:36Z rme: It's certainly true that the egc is quite fast at cleaning up after short-lived objects. 2016-07-07T14:13:37Z mrottenkolber: yeah so my current takeaway is that the GC is my friend 2016-07-07T14:14:56Z mrottenkolber: still kinda hard to accept that I can't improve the performance of my naive code from years ago ;-) 2016-07-07T14:16:05Z phf: there's some time and effect where you improve your algorithm and accidentally introduce some data structure reuse, and both your consing goes down, and your cache locality up, and you get mad gains 2016-07-07T14:16:14Z phf: *some time cause and effect 2016-07-07T14:18:04Z phf: but the only reason you want to eliminate consing is to reduce the frequency of gc on long running processes (in which case the strategy is often to keep the number and size of long lived consed objects low enough long enough that you can restart a process and do a switch over at a regular intervals) 2016-07-07T14:19:19Z mrottenkolber: I basically only have short lived objects 2016-07-07T14:19:29Z mrottenkolber: in this particular case 2016-07-07T14:24:54Z edgar-rft joined #ccl 2016-07-07T14:39:46Z lisper29` left #ccl 2016-07-07T14:45:33Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T14:47:12Z ekinmur joined #ccl 2016-07-07T15:51:54Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T15:58:18Z ekinmur joined #ccl 2016-07-07T16:11:05Z harlanc joined #ccl 2016-07-07T16:15:38Z harlanc left #ccl 2016-07-07T16:29:31Z pierpa joined #ccl 2016-07-07T16:41:11Z hhdave quit (Ping timeout: 244 seconds) 2016-07-07T17:31:27Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T17:33:06Z ekinmur joined #ccl 2016-07-07T18:11:37Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T18:12:26Z ekinmur joined #ccl 2016-07-07T18:42:36Z mrottenkolber quit (Ping timeout: 244 seconds) 2016-07-07T19:21:27Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T19:36:18Z hydan quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-07T19:36:42Z hydan joined #ccl 2016-07-07T19:40:23Z hydan quit (Client Quit) 2016-07-07T20:59:18Z edgar-rft quit (Quit: edgar-rft) 2016-07-07T21:20:05Z MetaHertz quit (Ping timeout: 250 seconds) 2016-07-07T22:12:37Z ekinmur joined #ccl 2016-07-07T23:27:52Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-07T23:30:58Z DougNYC joined #ccl 2016-07-07T23:38:37Z milanj quit (Quit: This computer has gone to sleep) 2016-07-08T00:51:04Z DougNYC quit 2016-07-08T01:51:26Z ekinmur joined #ccl 2016-07-08T03:07:44Z pierpa quit (Ping timeout: 272 seconds) 2016-07-08T03:37:06Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T04:11:36Z edgar-rft joined #ccl 2016-07-08T05:06:15Z rme quit (Quit: rme) 2016-07-08T06:34:56Z hydan joined #ccl 2016-07-08T06:35:57Z milanj joined #ccl 2016-07-08T06:48:50Z edgar-rft quit (Quit: edgar-rft) 2016-07-08T07:20:01Z milanj quit (Quit: This computer has gone to sleep) 2016-07-08T08:13:40Z MetaHertz joined #ccl 2016-07-08T08:17:54Z hhdave joined #ccl 2016-07-08T08:24:37Z fUD quit (Quit: Connection closed for inactivity) 2016-07-08T08:25:44Z milanj joined #ccl 2016-07-08T09:25:24Z edgar-rft joined #ccl 2016-07-08T10:19:13Z hydan quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T10:33:56Z billstclair quit (Ping timeout: 250 seconds) 2016-07-08T10:35:16Z billstclair joined #ccl 2016-07-08T11:04:08Z hydan joined #ccl 2016-07-08T12:19:40Z rme joined #ccl 2016-07-08T13:17:30Z ekinmur joined #ccl 2016-07-08T14:47:37Z Blkt joined #ccl 2016-07-08T15:29:13Z hydan quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-08T16:03:22Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T16:04:48Z ekinmur joined #ccl 2016-07-08T16:09:08Z ekinmur quit (Client Quit) 2016-07-08T16:11:50Z ekinmur joined #ccl 2016-07-08T16:38:32Z hhdave quit (Ping timeout: 244 seconds) 2016-07-08T16:54:06Z logrus joined #ccl 2016-07-08T17:49:27Z pierpa joined #ccl 2016-07-08T17:55:59Z eschatologist quit (Quit: ZNC 1.6.3+deb1 - http://znc.in) 2016-07-08T17:58:02Z eschatologist joined #ccl 2016-07-08T17:58:20Z logrus quit (Quit: Leaving) 2016-07-08T18:59:43Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T19:00:36Z ekinmur joined #ccl 2016-07-08T19:02:09Z hhdave joined #ccl 2016-07-08T19:06:47Z hhdave quit (Ping timeout: 260 seconds) 2016-07-08T19:08:12Z hhdave joined #ccl 2016-07-08T19:19:28Z edgar-rft quit (Quit: edgar-rft) 2016-07-08T19:31:38Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-08T19:45:02Z ekinmur quit (Read error: Connection reset by peer) 2016-07-08T19:45:21Z ekinmur joined #ccl 2016-07-08T19:54:07Z lisper29 joined #ccl 2016-07-08T19:58:12Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T20:03:41Z ekinmur joined #ccl 2016-07-08T20:08:24Z hhdave quit (Quit: hhdave) 2016-07-08T20:29:32Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T20:30:20Z ekinmur joined #ccl 2016-07-08T20:48:51Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T20:49:54Z ekinmur joined #ccl 2016-07-08T21:09:24Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-08T21:33:59Z milanj quit (Quit: This computer has gone to sleep) 2016-07-08T22:03:50Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-08T22:37:21Z lisper29 quit (Quit: This computer has gone to sleep) 2016-07-08T22:44:22Z lisper29 joined #ccl 2016-07-08T22:44:45Z lisper29 is now known as Guest89654 2016-07-08T23:26:19Z Guest89654 left #ccl 2016-07-08T23:26:56Z lisper29` joined #ccl 2016-07-08T23:52:40Z rumbler31 joined #ccl 2016-07-09T01:23:29Z ekinmur joined #ccl 2016-07-09T01:23:35Z rumbler31 quit (Remote host closed the connection) 2016-07-09T01:28:17Z ekinmur quit (Ping timeout: 260 seconds) 2016-07-09T01:54:18Z ekinmur joined #ccl 2016-07-09T02:00:47Z lisper29` quit (Quit: This computer has gone to sleep) 2016-07-09T02:07:41Z lisper29` joined #ccl 2016-07-09T02:09:38Z lisper29` left #ccl 2016-07-09T03:15:24Z jdz quit (Ping timeout: 258 seconds) 2016-07-09T03:19:59Z jdz joined #ccl 2016-07-09T03:54:45Z edgar-rft joined #ccl 2016-07-09T04:00:39Z rme quit (Quit: rme) 2016-07-09T04:09:22Z pierpa quit (Ping timeout: 272 seconds) 2016-07-09T04:18:38Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-09T04:18:45Z rme joined #ccl 2016-07-09T06:23:04Z milanj joined #ccl 2016-07-09T06:41:49Z milanj quit (Quit: This computer has gone to sleep) 2016-07-09T07:00:50Z MetaHertz joined #ccl 2016-07-09T07:11:48Z MetaHertz quit (Quit: Leaving) 2016-07-09T08:27:31Z rme quit (Quit: rme) 2016-07-09T08:27:31Z rme quit (Quit: rme) 2016-07-09T09:45:23Z hhdave_ joined #ccl 2016-07-09T09:50:12Z DavidGu joined #ccl 2016-07-09T09:57:31Z schjetne quit (Ping timeout: 240 seconds) 2016-07-09T10:00:48Z milanj joined #ccl 2016-07-09T10:49:55Z DavidGu quit (Quit: DavidGu) 2016-07-09T12:11:22Z milanj quit (Quit: This computer has gone to sleep) 2016-07-09T12:41:16Z milanj joined #ccl 2016-07-09T14:31:24Z milanj quit (Quit: This computer has gone to sleep) 2016-07-09T14:32:01Z rme joined #ccl 2016-07-09T16:10:32Z pierpa joined #ccl 2016-07-09T16:35:08Z MetaHertz joined #ccl 2016-07-09T16:35:15Z dr-Hertz joined #ccl 2016-07-09T16:35:30Z dr-Hertz quit (Remote host closed the connection) 2016-07-09T17:13:07Z edgar-rft quit (Quit: edgar-rft) 2016-07-09T17:34:19Z mrottenkolber joined #ccl 2016-07-09T17:34:25Z mrottenkolber: Hi! 2016-07-09T17:34:53Z mrottenkolber: I just noticed a huge slowdown when using WITH-SLOTS instead of LET plus structure accessors, any idea why that is? 2016-07-09T17:37:54Z rme: with-slots expands to symbol-macrolet, so maybe the compiler doesn't deal with that as well as a plain let? I don't really know right off hand. 2016-07-09T17:53:58Z mrottenkolber quit (Ping timeout: 272 seconds) 2016-07-09T17:57:40Z milanj joined #ccl 2016-07-09T18:00:21Z jackdaniel: I think that using with-slots accesses the slot with the accessor on each reference, while let binds just a value, hence the latter indeed should be faster 2016-07-09T18:01:27Z jackdaniel: perhaps one could create with-slots* which just binds the values to the variables 2016-07-09T18:02:36Z jackdaniel: (but then setf wouldn't work) 2016-07-09T18:17:22Z logrus joined #ccl 2016-07-09T18:24:22Z edgar-rft joined #ccl 2016-07-09T18:35:37Z eschatologist quit (Remote host closed the connection) 2016-07-09T18:37:45Z eschatologist joined #ccl 2016-07-09T19:03:20Z mrottenkolber joined #ccl 2016-07-09T19:43:47Z MetaHertz quit (Ping timeout: 260 seconds) 2016-07-09T20:01:17Z mrottenkolber quit (Ping timeout: 260 seconds) 2016-07-09T20:34:11Z schjetne joined #ccl 2016-07-09T20:47:03Z schjetne quit (Ping timeout: 240 seconds) 2016-07-09T22:28:35Z logrus quit (Quit: Leaving) 2016-07-09T22:44:36Z logrus joined #ccl 2016-07-09T23:14:48Z copec joined #ccl 2016-07-09T23:48:20Z logrus quit (Quit: Leaving) 2016-07-09T23:48:51Z hhdave_ quit (Quit: hhdave_) 2016-07-10T00:07:56Z PuercoPop left #ccl 2016-07-10T00:51:46Z milanj quit (Quit: This computer has gone to sleep) 2016-07-10T01:18:51Z pierpa quit (Ping timeout: 240 seconds) 2016-07-10T03:29:34Z MetaHertz joined #ccl 2016-07-10T04:59:52Z edgar-rft quit (*.net *.split) 2016-07-10T04:59:53Z copec quit (*.net *.split) 2016-07-10T04:59:54Z eschatologist quit (*.net *.split) 2016-07-10T04:59:54Z billstclair quit (*.net *.split) 2016-07-10T04:59:55Z faheem quit (*.net *.split) 2016-07-10T04:59:56Z brucem quit (*.net *.split) 2016-07-10T04:59:56Z |3b| quit (*.net *.split) 2016-07-10T05:35:06Z rumbler31 joined #ccl 2016-07-10T05:35:26Z MetaHertz quit (Quit: Leaving) 2016-07-10T06:09:01Z rumbler3_ joined #ccl 2016-07-10T06:11:12Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-10T06:35:26Z edgar-rft joined #ccl 2016-07-10T06:45:31Z schjetne joined #ccl 2016-07-10T07:17:22Z milanj joined #ccl 2016-07-10T08:54:02Z copec joined #ccl 2016-07-10T08:54:15Z eschatologist joined #ccl 2016-07-10T08:54:15Z billstclair joined #ccl 2016-07-10T08:54:15Z faheem joined #ccl 2016-07-10T08:54:15Z |3b| joined #ccl 2016-07-10T08:54:15Z brucem joined #ccl 2016-07-10T09:19:36Z rumbler3_ quit (Remote host closed the connection) 2016-07-10T09:31:17Z rme quit (Quit: rme) 2016-07-10T09:31:17Z rme quit (Quit: rme) 2016-07-10T09:54:55Z schjetne_ joined #ccl 2016-07-10T09:56:09Z schjetne quit (Ping timeout: 246 seconds) 2016-07-10T10:02:13Z schjetne_ is now known as schjetne 2016-07-10T10:27:57Z schjetne quit (Ping timeout: 276 seconds) 2016-07-10T12:32:49Z schjetne joined #ccl 2016-07-10T13:33:41Z milanj quit (Quit: This computer has gone to sleep) 2016-07-10T14:49:48Z milanj joined #ccl 2016-07-10T15:04:12Z pierpa joined #ccl 2016-07-10T15:13:01Z milanj quit (Quit: This computer has gone to sleep) 2016-07-10T15:35:31Z milanj joined #ccl 2016-07-10T16:24:00Z rme joined #ccl 2016-07-10T17:19:08Z milanj quit (Quit: This computer has gone to sleep) 2016-07-10T17:25:19Z logrus joined #ccl 2016-07-10T18:03:31Z MetaHertz joined #ccl 2016-07-10T18:04:13Z rumbler31 joined #ccl 2016-07-10T18:39:35Z MetaHertz quit (Quit: Leaving) 2016-07-10T18:40:55Z MetaHertz joined #ccl 2016-07-10T19:29:18Z MetaHertz quit (Remote host closed the connection) 2016-07-10T20:37:00Z schjetne quit (Ping timeout: 276 seconds) 2016-07-10T20:49:19Z schjetne joined #ccl 2016-07-10T21:07:47Z milanj joined #ccl 2016-07-10T21:09:02Z edgar-rft quit (Quit: edgar-rft) 2016-07-10T21:10:41Z ekinmur joined #ccl 2016-07-10T21:16:11Z DougNYC joined #ccl 2016-07-10T21:19:35Z schjetne quit (Remote host closed the connection) 2016-07-10T21:20:55Z schjetne joined #ccl 2016-07-10T21:21:09Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-10T22:00:12Z schjetne quit (Ping timeout: 276 seconds) 2016-07-10T22:07:31Z rumbler31 quit (Ping timeout: 240 seconds) 2016-07-10T22:09:04Z rumbler31 joined #ccl 2016-07-10T23:08:19Z rme quit (Ping timeout: 184 seconds) 2016-07-10T23:09:22Z rme quit (Ping timeout: 260 seconds) 2016-07-10T23:44:33Z milanj quit (Quit: This computer has gone to sleep) 2016-07-11T01:58:06Z pierpa quit (Ping timeout: 276 seconds) 2016-07-11T02:01:31Z ekinmur joined #ccl 2016-07-11T02:08:40Z rumbler3_ joined #ccl 2016-07-11T02:11:45Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-11T02:33:20Z logrus quit (Quit: Leaving) 2016-07-11T02:41:48Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-11T04:03:47Z rumbler3_ quit (Remote host closed the connection) 2016-07-11T04:14:54Z ChibaPet joined #ccl 2016-07-11T04:33:25Z DougNYC quit 2016-07-11T05:30:38Z edgar-rft joined #ccl 2016-07-11T06:04:54Z milanj joined #ccl 2016-07-11T06:30:19Z schjetne joined #ccl 2016-07-11T07:21:05Z pmicossi joined #ccl 2016-07-11T08:06:30Z knobo1 quit (Ping timeout: 246 seconds) 2016-07-11T08:20:23Z knobo1 joined #ccl 2016-07-11T08:21:36Z carvite quit (Ping timeout: 276 seconds) 2016-07-11T08:22:03Z carvite joined #ccl 2016-07-11T08:48:26Z schjetne quit (Ping timeout: 272 seconds) 2016-07-11T08:56:16Z milanj quit (Quit: This computer has gone to sleep) 2016-07-11T09:14:06Z Tennis joined #ccl 2016-07-11T10:03:24Z schjetne joined #ccl 2016-07-11T11:00:38Z pmicossi quit (Read error: Connection reset by peer) 2016-07-11T11:27:30Z milanj joined #ccl 2016-07-11T12:49:24Z rme joined #ccl 2016-07-11T13:03:35Z ekinmur joined #ccl 2016-07-11T13:27:06Z schjetne quit (Ping timeout: 276 seconds) 2016-07-11T13:32:37Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-11T13:35:40Z ekinmur joined #ccl 2016-07-11T13:43:42Z pierpa joined #ccl 2016-07-11T15:33:35Z nzambe joined #ccl 2016-07-11T15:55:14Z stassats joined #ccl 2016-07-11T15:57:35Z stassats: ccl seems to have the same problem as sbcl on fifos and poll() on osx 2016-07-11T15:58:10Z rme: stassats: What problem is that? 2016-07-11T15:58:32Z stassats: transforming sbcl symbols into ccl symbols and pasting 2016-07-11T16:00:51Z stassats: rme: http://paste.lisp.org/display/320171#3 2016-07-11T16:01:06Z stassats: the second read-char blocks, it signals EOF on linux 2016-07-11T16:01:15Z stassats: and it signals EOF if no with-timeout is present 2016-07-11T16:01:40Z stassats: basically it looks like poll/select on closed fifos just block 2016-07-11T16:03:02Z rme: I seem to recall that poll of fifos just doesn't work on OS X. 2016-07-11T16:03:37Z stassats: i guess it's not that big of a problem for ccl, since it only uses poll when timeouts are present 2016-07-11T16:03:51Z stassats: on sbcl it always calls select, sigh 2016-07-11T16:05:05Z stassats: echo -n a > /private/tmp/fifo seems to wake the poll up 2016-07-11T16:08:08Z stassats: i can probably just forgo polling on fifo fds on sbcl, that'll lose timeouts, but will at least work 2016-07-11T17:20:43Z MetaHertz joined #ccl 2016-07-11T17:26:00Z MetaHertz quit (Ping timeout: 258 seconds) 2016-07-11T17:27:40Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-11T17:28:34Z ekinmur joined #ccl 2016-07-11T17:46:00Z MetaHertz joined #ccl 2016-07-11T18:21:12Z MetaHertz quit (Ping timeout: 244 seconds) 2016-07-11T18:22:21Z MetaHertz joined #ccl 2016-07-11T18:29:21Z MetaHertz quit (Ping timeout: 276 seconds) 2016-07-11T18:30:09Z MetaHertz joined #ccl 2016-07-11T18:40:06Z milanj quit (Quit: This computer has gone to sleep) 2016-07-11T19:05:56Z stassats quit (Ping timeout: 272 seconds) 2016-07-11T19:07:13Z stassats joined #ccl 2016-07-11T19:11:37Z stassats quit (Ping timeout: 260 seconds) 2016-07-11T19:13:45Z stassats joined #ccl 2016-07-11T19:17:51Z stassats quit (Ping timeout: 240 seconds) 2016-07-11T19:19:48Z stassats joined #ccl 2016-07-11T20:02:30Z milanj joined #ccl 2016-07-11T20:22:56Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-11T20:24:39Z ekinmur joined #ccl 2016-07-11T20:27:12Z stassats: kqueue doesn't work on closed fifos either 2016-07-11T20:30:12Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-11T21:24:03Z rumbler31 joined #ccl 2016-07-11T21:44:10Z edgar-rft quit (Quit: edgar-rft) 2016-07-11T23:10:33Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-11T23:32:22Z ekinmur joined #ccl 2016-07-11T23:36:44Z ekinmur quit (Ping timeout: 250 seconds) 2016-07-11T23:37:25Z milanj quit (Quit: This computer has gone to sleep) 2016-07-12T00:07:20Z MetaHertz joined #ccl 2016-07-12T00:40:20Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-12T00:41:40Z MetaHertz joined #ccl 2016-07-12T00:47:30Z ekinmur joined #ccl 2016-07-12T01:05:51Z MetaHertz quit (Ping timeout: 276 seconds) 2016-07-12T01:16:07Z MetaHertz joined #ccl 2016-07-12T01:21:27Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-12T01:22:38Z MetaHertz joined #ccl 2016-07-12T01:24:01Z rumbler3_ joined #ccl 2016-07-12T01:26:25Z rumbler31 quit (Ping timeout: 244 seconds) 2016-07-12T01:35:17Z MetaHertz quit (Remote host closed the connection) 2016-07-12T01:35:36Z MetaHertz joined #ccl 2016-07-12T01:40:31Z MetaHertz quit (Ping timeout: 240 seconds) 2016-07-12T01:49:22Z stassats quit (Ping timeout: 272 seconds) 2016-07-12T01:50:50Z MetaHertz joined #ccl 2016-07-12T02:09:52Z MetaHertz quit (Ping timeout: 260 seconds) 2016-07-12T02:11:18Z MetaHertz joined #ccl 2016-07-12T02:22:38Z DavidGu joined #ccl 2016-07-12T02:30:09Z rumbler3_ quit (Remote host closed the connection) 2016-07-12T02:49:37Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-12T02:49:56Z DougNYC joined #ccl 2016-07-12T03:06:16Z pierpa quit (Ping timeout: 252 seconds) 2016-07-12T03:34:38Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-12T03:39:51Z MetaHertz quit (Ping timeout: 240 seconds) 2016-07-12T04:01:34Z DavidGu quit (Ping timeout: 258 seconds) 2016-07-12T04:42:42Z milanj joined #ccl 2016-07-12T04:47:47Z rme quit (Quit: rme) 2016-07-12T04:54:44Z DavidGu joined #ccl 2016-07-12T05:04:37Z DougNYC quit 2016-07-12T06:01:00Z schjetne joined #ccl 2016-07-12T06:02:02Z milanj quit (Quit: This computer has gone to sleep) 2016-07-12T06:07:41Z edgar-rft joined #ccl 2016-07-12T06:56:26Z LahlStephan joined #ccl 2016-07-12T07:03:30Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-12T07:05:10Z eschatologist joined #ccl 2016-07-12T07:07:44Z milanj joined #ccl 2016-07-12T07:07:52Z DavidGu joined #ccl 2016-07-12T07:24:28Z DavidGu quit (Ping timeout: 244 seconds) 2016-07-12T07:24:46Z DavidGu joined #ccl 2016-07-12T07:29:33Z milanj_ joined #ccl 2016-07-12T07:31:45Z milanj quit (Ping timeout: 246 seconds) 2016-07-12T07:58:58Z MetaHertz joined #ccl 2016-07-12T07:59:57Z DavidGu quit (Remote host closed the connection) 2016-07-12T08:00:58Z DavidGu joined #ccl 2016-07-12T08:49:44Z DavidGu quit (Quit: DavidGu) 2016-07-12T10:40:49Z stassats joined #ccl 2016-07-12T11:24:18Z tfb joined #ccl 2016-07-12T11:30:31Z stassats quit (Ping timeout: 240 seconds) 2016-07-12T11:31:18Z stassats joined #ccl 2016-07-12T11:35:42Z stassats quit (Ping timeout: 246 seconds) 2016-07-12T11:36:49Z stassats joined #ccl 2016-07-12T11:40:51Z stassats quit (Ping timeout: 240 seconds) 2016-07-12T11:44:17Z stassats joined #ccl 2016-07-12T12:56:17Z ekinmur joined #ccl 2016-07-12T12:56:46Z ekinmur quit (Client Quit) 2016-07-12T12:59:16Z ekinmur joined #ccl 2016-07-12T13:02:51Z eschatologist quit (Ping timeout: 240 seconds) 2016-07-12T13:25:57Z eschatologist joined #ccl 2016-07-12T13:30:01Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-12T13:32:07Z ekinmur joined #ccl 2016-07-12T13:34:00Z rme joined #ccl 2016-07-12T13:59:47Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-12T14:00:43Z ekinmur joined #ccl 2016-07-12T14:27:15Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-12T14:38:18Z milanj_ joined #ccl 2016-07-12T14:48:45Z stassats quit (Ping timeout: 276 seconds) 2016-07-12T15:38:08Z stassats joined #ccl 2016-07-12T15:42:37Z stassats quit (Ping timeout: 250 seconds) 2016-07-12T15:43:07Z stassats joined #ccl 2016-07-12T15:44:19Z hjudt joined #ccl 2016-07-12T15:50:26Z stassats quit (Ping timeout: 272 seconds) 2016-07-12T15:53:38Z stassats joined #ccl 2016-07-12T15:58:03Z stassats quit (Ping timeout: 244 seconds) 2016-07-12T16:00:06Z stassats joined #ccl 2016-07-12T16:04:33Z stassats quit (Ping timeout: 240 seconds) 2016-07-12T16:08:41Z stassats joined #ccl 2016-07-12T16:21:03Z MetaHertz quit (Remote host closed the connection) 2016-07-12T16:23:52Z MetaHertz joined #ccl 2016-07-12T16:31:38Z pierpa joined #ccl 2016-07-12T17:22:36Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-12T17:47:17Z MetaHertz quit (Ping timeout: 260 seconds) 2016-07-12T17:59:45Z tfb quit (Quit: Leaving...) 2016-07-12T18:20:35Z milanj_ joined #ccl 2016-07-12T18:48:47Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-12T20:41:58Z Blkt quit (Read error: Connection reset by peer) 2016-07-12T20:41:58Z fe[nl]ix quit (Read error: Connection reset by peer) 2016-07-12T20:43:13Z Blkt joined #ccl 2016-07-12T20:43:36Z fe[nl]ix joined #ccl 2016-07-12T21:32:13Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-12T23:50:30Z edgar-rft quit (Quit: edgar-rft) 2016-07-13T01:24:23Z rumbler31 joined #ccl 2016-07-13T01:54:12Z milanj_ joined #ccl 2016-07-13T02:56:04Z stassats quit (Ping timeout: 272 seconds) 2016-07-13T03:15:03Z stassats joined #ccl 2016-07-13T03:16:25Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-13T03:19:40Z stassats quit (Ping timeout: 264 seconds) 2016-07-13T03:40:12Z pierpa quit (Remote host closed the connection) 2016-07-13T03:40:29Z pierpa joined #ccl 2016-07-13T03:44:58Z pierpa` joined #ccl 2016-07-13T03:49:16Z pierpa quit (Ping timeout: 272 seconds) 2016-07-13T03:49:31Z pierpa` quit (Remote host closed the connection) 2016-07-13T03:49:45Z pierpa` joined #ccl 2016-07-13T03:55:16Z pierpa` is now known as pierpa 2016-07-13T04:33:33Z pierpa quit (Ping timeout: 240 seconds) 2016-07-13T04:44:21Z DavidGu joined #ccl 2016-07-13T04:54:40Z milanj_ joined #ccl 2016-07-13T05:24:58Z rumbler3_ joined #ccl 2016-07-13T05:28:12Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-13T05:32:24Z rme quit (Quit: rme) 2016-07-13T05:32:24Z rme quit (Quit: rme) 2016-07-13T06:00:30Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-13T06:28:53Z MetaHertz joined #ccl 2016-07-13T07:35:07Z pmicossi joined #ccl 2016-07-13T07:36:35Z milanj_ joined #ccl 2016-07-13T08:53:10Z rumbler3_ quit (Remote host closed the connection) 2016-07-13T08:54:58Z MetaHertz quit (Remote host closed the connection) 2016-07-13T08:55:32Z edgar-rft joined #ccl 2016-07-13T08:56:10Z MetaHertz joined #ccl 2016-07-13T08:56:48Z MetaHertz quit (Remote host closed the connection) 2016-07-13T08:57:04Z MetaHertz joined #ccl 2016-07-13T10:24:52Z hhdave joined #ccl 2016-07-13T10:37:44Z Quadrescence quit (Ping timeout: 258 seconds) 2016-07-13T10:54:46Z Quadrescence joined #ccl 2016-07-13T11:02:38Z pmicossi quit (Quit: going home) 2016-07-13T11:47:16Z stassats joined #ccl 2016-07-13T11:52:58Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-13T11:54:19Z nzambe joined #ccl 2016-07-13T11:59:30Z rme joined #ccl 2016-07-13T12:00:34Z tfb joined #ccl 2016-07-13T12:42:11Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-13T13:01:32Z milanj_ joined #ccl 2016-07-13T14:32:05Z sam___ joined #ccl 2016-07-13T14:34:52Z sam___ left #ccl 2016-07-13T15:20:59Z edgar-rft quit (Quit: edgar-rft) 2016-07-13T15:40:07Z tfb quit (Remote host closed the connection) 2016-07-13T16:22:29Z tfb joined #ccl 2016-07-13T16:35:06Z hhdave quit (Ping timeout: 276 seconds) 2016-07-13T17:43:01Z pierpa joined #ccl 2016-07-13T17:55:33Z edgar-rft joined #ccl 2016-07-13T17:59:48Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-13T19:26:12Z milanj_ joined #ccl 2016-07-13T19:26:51Z stassats quit (Ping timeout: 240 seconds) 2016-07-13T19:29:32Z stassats joined #ccl 2016-07-13T19:31:53Z rumbler31 joined #ccl 2016-07-13T19:45:57Z rumbler31 quit (Ping timeout: 246 seconds) 2016-07-13T20:23:03Z MetaHertz quit (Ping timeout: 246 seconds) 2016-07-13T20:50:50Z milanj_ quit (Quit: This computer has gone to sleep) 2016-07-13T21:37:48Z rumbler31 joined #ccl 2016-07-13T21:47:25Z stassats quit (Ping timeout: 250 seconds) 2016-07-13T21:48:05Z stassats joined #ccl 2016-07-13T21:52:32Z stassats quit (Ping timeout: 260 seconds) 2016-07-13T21:55:09Z stassats joined #ccl 2016-07-13T21:59:33Z stassats quit (Ping timeout: 250 seconds) 2016-07-13T22:00:06Z stassats joined #ccl 2016-07-13T22:04:11Z stassats quit (Ping timeout: 240 seconds) 2016-07-13T22:05:36Z stassats joined #ccl 2016-07-13T23:16:39Z DougNYC joined #ccl 2016-07-13T23:28:00Z rumbler31 quit (Ping timeout: 244 seconds) 2016-07-13T23:28:47Z rumbler31 joined #ccl 2016-07-13T23:40:19Z pierpa quit (Read error: Connection reset by peer) 2016-07-13T23:44:10Z pierpa joined #ccl 2016-07-13T23:47:44Z pierpa quit (Read error: Connection reset by peer) 2016-07-13T23:52:18Z pierpa joined #ccl 2016-07-14T00:01:07Z milanj joined #ccl 2016-07-14T01:14:23Z milanj quit (Quit: This computer has gone to sleep) 2016-07-14T01:38:22Z rumbler3_ joined #ccl 2016-07-14T01:40:58Z rumbler31 quit (Ping timeout: 250 seconds) 2016-07-14T02:05:33Z edgar-rft quit (Quit: edgar-rft) 2016-07-14T02:21:48Z stassats quit (Ping timeout: 246 seconds) 2016-07-14T03:09:09Z rumbler3_ quit (Remote host closed the connection) 2016-07-14T03:17:37Z pierpa quit (Ping timeout: 252 seconds) 2016-07-14T04:16:57Z DougNYC quit 2016-07-14T06:23:48Z rme quit (Quit: rme) 2016-07-14T06:40:11Z tfeb joined #ccl 2016-07-14T06:40:40Z tfeb quit (Remote host closed the connection) 2016-07-14T06:40:55Z tfeb joined #ccl 2016-07-14T06:43:38Z tfb quit (Ping timeout: 272 seconds) 2016-07-14T07:34:46Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-14T07:35:13Z milanj joined #ccl 2016-07-14T07:44:05Z milanj quit (Quit: This computer has gone to sleep) 2016-07-14T07:52:20Z MetaHertz joined #ccl 2016-07-14T07:56:51Z tfeb quit (Ping timeout: 240 seconds) 2016-07-14T07:59:15Z MetaHertz quit (Quit: Leaving) 2016-07-14T07:59:53Z MetaHertz joined #ccl 2016-07-14T08:23:55Z pillton quit (Quit: ERC (IRC client for Emacs 24.5.1)) 2016-07-14T08:27:49Z MetaHertz quit (Ping timeout: 252 seconds) 2016-07-14T08:37:00Z DavidGu quit (Ping timeout: 272 seconds) 2016-07-14T09:02:35Z MetaHertz joined #ccl 2016-07-14T09:29:04Z pierpa joined #ccl 2016-07-14T10:00:18Z MetaHertz quit (Quit: Leaving) 2016-07-14T10:01:51Z MetaHertz joined #ccl 2016-07-14T10:18:10Z MetaHertz quit (Remote host closed the connection) 2016-07-14T10:18:47Z MetaHertz joined #ccl 2016-07-14T10:27:13Z milanj joined #ccl 2016-07-14T10:41:26Z hhdave joined #ccl 2016-07-14T11:02:24Z pierpa quit (Ping timeout: 244 seconds) 2016-07-14T11:38:03Z LahlStephan quit (Remote host closed the connection) 2016-07-14T11:41:07Z StephanLahl joined #ccl 2016-07-14T11:51:27Z DavidGu joined #ccl 2016-07-14T11:51:28Z DavidGu quit (Client Quit) 2016-07-14T12:37:47Z edgar-rft joined #ccl 2016-07-14T12:54:21Z stassats joined #ccl 2016-07-14T13:14:48Z rme joined #ccl 2016-07-14T13:45:04Z nzambe joined #ccl 2016-07-14T14:31:02Z MetaHertz quit (Quit: Leaving) 2016-07-14T14:31:39Z stassats quit (Ping timeout: 244 seconds) 2016-07-14T14:32:55Z stassats joined #ccl 2016-07-14T14:37:41Z stassats quit (Client Quit) 2016-07-14T14:53:35Z stassats joined #ccl 2016-07-14T14:54:41Z MetaHertz joined #ccl 2016-07-14T15:00:30Z stassats quit (Ping timeout: 258 seconds) 2016-07-14T15:01:03Z stassats joined #ccl 2016-07-14T15:05:29Z stassats quit (Ping timeout: 258 seconds) 2016-07-14T16:42:28Z hhdave quit (Ping timeout: 258 seconds) 2016-07-14T16:47:09Z MetaHertz quit (Remote host closed the connection) 2016-07-14T16:47:31Z MetaHertz joined #ccl 2016-07-14T17:01:33Z stassats joined #ccl 2016-07-14T17:05:33Z milanj quit (Quit: This computer has gone to sleep) 2016-07-14T17:08:36Z MetaHertz quit (Remote host closed the connection) 2016-07-14T17:09:07Z MetaHertz joined #ccl 2016-07-14T18:12:00Z stassats quit (Quit: ZNC 1.6.2+deb1 - http://znc.in) 2016-07-14T18:16:09Z stassats joined #ccl 2016-07-14T18:16:18Z stassats left #ccl 2016-07-14T18:42:21Z pierpa joined #ccl 2016-07-14T19:22:17Z MetaHertz quit (Ping timeout: 260 seconds) 2016-07-14T21:01:19Z hhdave joined #ccl 2016-07-14T21:02:35Z hhdave_ joined #ccl 2016-07-14T21:05:39Z hhdave quit (Ping timeout: 246 seconds) 2016-07-14T21:05:39Z hhdave_ is now known as hhdave 2016-07-14T22:20:19Z eschatologist quit (Quit: ZNC 1.6.3+deb1 - http://znc.in) 2016-07-14T22:21:03Z eschatologist joined #ccl 2016-07-14T22:21:35Z hhdave quit (Quit: hhdave) 2016-07-14T22:30:10Z edgar-rft quit (Quit: edgar-rft) 2016-07-14T23:53:50Z DougNYC joined #ccl 2016-07-15T00:34:57Z pierpa quit (Ping timeout: 260 seconds) 2016-07-15T00:44:43Z emaczen joined #ccl 2016-07-15T00:54:26Z emaczen quit (Changing host) 2016-07-15T00:54:26Z emaczen joined #ccl 2016-07-15T00:54:31Z nzambe quit (Ping timeout: 252 seconds) 2016-07-15T04:39:54Z DougNYC quit 2016-07-15T06:06:36Z edgar-rft joined #ccl 2016-07-15T06:30:03Z rme quit (Quit: rme) 2016-07-15T06:30:04Z rme quit (Quit: rme) 2016-07-15T07:01:40Z pmicossi joined #ccl 2016-07-15T07:30:59Z eschatologist: Interesting. CCL built from trunk on my Raspberry Pi 2 with up-to-date Raspbian loads QuickLisp just fine, but CCL built from trunk on my Raspberry Pi 3 with up-to-date Rapsbian fails to load QuickLisp: Error: The value 4294967188 is not of the expected type INTEGER. 2016-07-15T08:00:57Z hhdave joined #ccl 2016-07-15T08:11:18Z MetaHertz joined #ccl 2016-07-15T08:28:04Z edgar-rft: eschatologist: Obviously something has broken CCL's integrity. 2016-07-15T08:53:30Z DavidGu joined #ccl 2016-07-15T09:16:00Z tfb joined #ccl 2016-07-15T09:22:26Z tfeb joined #ccl 2016-07-15T09:23:08Z tfeb quit (Remote host closed the connection) 2016-07-15T09:23:21Z tfb quit (Ping timeout: 258 seconds) 2016-07-15T09:28:42Z edgar-rft quit (Quit: edgar-rft) 2016-07-15T09:39:54Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-15T10:59:11Z vydd_ joined #ccl 2016-07-15T11:04:19Z vydd_ is now known as vydd 2016-07-15T11:04:19Z vydd quit (Changing host) 2016-07-15T11:04:19Z vydd joined #ccl 2016-07-15T11:41:50Z edgar-rft joined #ccl 2016-07-15T12:24:23Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-15T12:25:09Z eschatologist joined #ccl 2016-07-15T12:33:07Z DavidGu quit (Quit: DavidGu) 2016-07-15T12:44:52Z rme joined #ccl 2016-07-15T12:45:41Z MetaHertz joined #ccl 2016-07-15T13:08:16Z schjetne quit (Ping timeout: 272 seconds) 2016-07-15T13:23:18Z milanj joined #ccl 2016-07-15T14:02:27Z vydd quit 2016-07-15T14:03:02Z nzambe joined #ccl 2016-07-15T14:12:21Z schjetne joined #ccl 2016-07-15T14:40:51Z MetaHertz quit (Ping timeout: 276 seconds) 2016-07-15T15:00:46Z MajorTennis joined #ccl 2016-07-15T15:04:40Z Tennis quit (Ping timeout: 264 seconds) 2016-07-15T15:05:01Z MajorTennis is now known as Tennis 2016-07-15T15:40:20Z pmicossi quit (Quit: going home) 2016-07-15T15:48:56Z phoe_krk joined #ccl 2016-07-15T15:49:08Z phoe_krk: Hey hi!Greetings from #mcclim! 2016-07-15T15:49:13Z phoe_krk inserts a space up there. 2016-07-15T15:49:44Z phoe_krk: I'm curious about https://github.com/robert-strandh/McCLIM/issues/50 - I suspect a possible bug in CCL's threading that causes a big slowdown in McCLIM when rendering stuff in general. 2016-07-15T15:49:53Z phoe_krk: Could anyone knowledgeable in CCL take a look at it? 2016-07-15T15:52:51Z pierpa joined #ccl 2016-07-15T16:17:19Z rme: phoe_krk: That's weird. Does McCLIM use process-wait, by any chance? 2016-07-15T16:18:06Z rme: Because process-wait busy-waits. 2016-07-15T16:18:14Z phoe_krk: rme: Gimme a second. 2016-07-15T16:18:22Z phoe_krk: It seems that CLX does. 2016-07-15T16:18:30Z phoe_krk: But give us a while. We'll give you a stacktrace. 2016-07-15T16:19:56Z phoe_krk: http://paste.lisp.org/display/320538 2016-07-15T16:20:08Z phoe_krk: XLIB does, which is CLX. 2016-07-15T16:20:23Z phoe_krk: Is there an alternative, rme? Like a "non-busy" wait? 2016-07-15T16:20:25Z rme: After looking at the code for process-wait, I see that it doesn't exactly busy-wait, but it checks the predicate and then sleeps a bit (namely for *ns-per-tick* nanoseconds). 2016-07-15T16:20:53Z phoe_krk: rme: where does the value for ccl::*ns-per-tick* come from? I mean, I can see the ticks-per-second thing. 2016-07-15T16:21:01Z phoe_krk: But it seems sort of huge, too huge. 2016-07-15T16:22:09Z phoe_krk: As you can see, I could lower it by a factor of 10e3 without any adverse weird effects. 2016-07-15T16:22:35Z phoe_krk: (I actually lowered it to 10 and had the McCLIM demo running, but that's sort of radical.) 2016-07-15T16:22:43Z rme: phoe_krk: I don't know the internals of CLX. Usually, I'd think you'd want to wait on some event (with #_select or #_poll maybe), or wait on some semaphore. 2016-07-15T16:23:38Z phoe_krk: rme: neither do I, we'd need to ask beach about it. But my question is - why does lowering that variable help? 2016-07-15T16:23:51Z rme: It appears that CCL initializes *ticks-per-second* by calling (#_sysconf #$_SC_CLK_TCK). On my Mac, this is 100, which is pretty small. 2016-07-15T16:24:06Z rme: Well, you poll more frequently, right? 2016-07-15T16:24:35Z phoe_krk: On my Linux, it's the same. 2016-07-15T16:24:38Z phoe_krk: 100. 2016-07-15T16:24:44Z phoe_krk: And I can poll much more frequently. 2016-07-15T16:24:54Z rme: *ns-per-tick* is computed from *ticks-per-second* 2016-07-15T16:24:57Z phoe_krk: I mean, uh. 100 ticks per second? That's pretty slow. 2016-07-15T16:25:00Z phoe_krk: I got that one, rme,. 2016-07-15T16:25:10Z phoe_krk: It's computed from that value and it's computed as a pretty huge sleep value. 2016-07-15T16:25:16Z phoe_krk: It needn't sleep that much. 2016-07-15T16:25:17Z rme: Sorry. 2016-07-15T16:25:42Z phoe_krk: Um, sorry? 2016-07-15T16:25:56Z phoe_krk: You didn't do anything bad. 2016-07-15T16:26:18Z rme: Sorry I didn't mean to insult your intelligence --- "I got that one". No worries, anyway. 2016-07-15T16:26:27Z phoe_krk: Oh! No problem. 2016-07-15T16:27:14Z phoe_krk: My question is: is that value computed correctly? 2016-07-15T16:27:34Z phoe_krk: If I can lower it by a factor of 1000 and get a performance *boost*? 2016-07-15T16:28:44Z phoe_krk: I mean - what's the reasoning behind that value? 2016-07-15T16:29:08Z rme: I doubt if there's any subtle reasoning behind it. 2016-07-15T16:29:33Z rme: I'd be willing to change *ticks-per-second* to be at least 1000 without a second thought. 2016-07-15T16:30:58Z phoe_krk: Where can I suggest a patch then? 2016-07-15T16:31:27Z rme: I'll just do it right now. 2016-07-15T16:31:58Z phoe_krk: Modify the value at which it's computed? 2016-07-15T16:32:13Z phoe_krk: Or make a (max 1000 ...) call? 2016-07-15T16:32:21Z rme: (max 1000 ...) 2016-07-15T16:32:31Z phoe_krk: Okay. 2016-07-15T16:33:16Z phoe_krk: I'd actually give it an even higher value. 1000 is once a milisecond, which is not a lot these days, especially given graphics rendering. 2016-07-15T16:33:19Z rme: That's just a factor of 10 difference, though. 2016-07-15T16:33:24Z phoe_krk: But - I'm not a maintainer or anything. 2016-07-15T16:33:42Z phoe_krk: I gave it a 1000x boost and it was acting much better. 2016-07-15T16:33:57Z hhdave quit (Ping timeout: 276 seconds) 2016-07-15T16:34:55Z MetaHertz joined #ccl 2016-07-15T16:35:10Z rme: The real problem is probably the fact that CLX is process-wait at all. 2016-07-15T16:36:15Z rme: Dang, I typed min instead of max and committed that. What an idiot I am. 2016-07-15T16:36:51Z phoe_krk: rme: oops 2016-07-15T16:37:54Z jdz: quick, force push while nobody has noticed the blunder! 2016-07-15T16:38:20Z phoe_krk: git star-wars 2016-07-15T16:38:30Z phoe_krk: it's an alias for git --force push 2016-07-15T16:38:32Z jdz: yes, use the force! 2016-07-15T16:38:35Z rme: Subversion doesn't have that. :-) 2016-07-15T16:38:46Z phoe_krk: git git! 2016-07-15T16:38:53Z jdz: i only have git fap which is an alias for git fetch --all --prune 2016-07-15T16:39:01Z phoe_krk: ...thta's bad 2016-07-15T16:39:10Z rme: Hey, this is a family channel. 2016-07-15T16:39:42Z phoe_krk: Lisper families go in here? 2016-07-15T16:39:53Z jdz: we're all one big family 2016-07-15T16:40:20Z phoe_krk: yay 2016-07-15T16:40:23Z phoe_krk channelhug 2016-07-15T16:41:31Z phoe_krk: rme: https://github.com/sharplispers/clx/blob/2aa969ee2abd777a76765e6713986cf3f5919be8/dep-openmcl.lisp#L496 2016-07-15T16:42:19Z MetaHertz quit (Ping timeout: 244 seconds) 2016-07-15T16:42:38Z jdz: that smells like the old days of lisp threading implementations where processes would have run reasons, which could be predicates 2016-07-15T16:43:31Z phoe_krk: jdz: this *are* the old days of lisp threading implementations. 2016-07-15T16:43:37Z phoe_krk: s/are/is/ 2016-07-15T16:43:43Z phoe_krk: This is CLX, the code that's older than me. 2016-07-15T16:44:26Z MetaHertz joined #ccl 2016-07-15T16:45:01Z jdz: yes, the comments mention "Zetalisp primitives", "TI Explorer" and "Symbolics" 2016-07-15T16:45:38Z jdz: the code probably could be updated to get rid of the process-bloc completely 2016-07-15T16:45:43Z phoe_krk: good lords 2016-07-15T16:45:52Z rme: It looks like the sbcl version of xlib::process-wait busy-waits, calling sched_yield in a loop with no sleep in there at all. 2016-07-15T16:45:54Z phoe_krk: Like - turning that into a NOP? 2016-07-15T16:46:58Z jdz: it's been many years since i've done anything with CLX 2016-07-15T16:47:02Z rme: xlib::process-block, I should say 2016-07-15T16:47:42Z phoe_krk: I see. 2016-07-15T16:47:59Z phoe_krk: So we should actually use this as a goal. 2016-07-15T16:48:09Z phoe_krk: Because SBCL does this really fast. 2016-07-15T16:48:20Z tfb joined #ccl 2016-07-15T16:48:23Z phoe_krk: When I lower that sleep value, CCL goes within the ~85% speed range which is very acceptable. 2016-07-15T16:48:50Z phoe_krk: When I go with the original sleep value, CCL is 0.085% of SBCL's speed. 2016-07-15T16:48:59Z rme: So, yeah, I'd probably change process-block to do what sbcl does. (loop (when (apply predicate args) (return) (#_sched_yield))) and call it a day. 2016-07-15T16:49:40Z rme: Or just leave out the yield and just busy-wait, for that matter. 2016-07-15T16:51:47Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-15T16:52:01Z nzambe joined #ccl 2016-07-15T16:53:29Z rme: Does that make sense, phoe_krk? I would avoid using ccl:process-wait in CLX, since ccl:process-wait tries to avoid busy-waiting by sleeping periodically, and it looks like that is interfering with how CLX works. 2016-07-15T16:54:14Z phoe_krk: rme: yes! 2016-07-15T16:54:28Z phoe_krk: We're talking on #clim now about the possibility of leveraging bordeaux-threads for actual threads. 2016-07-15T16:54:48Z phoe_krk: Because CLX doesn't use them in the slightest. 2016-07-15T16:55:42Z phoe_krk: But wait. 2016-07-15T16:55:55Z phoe_krk: This actually... uses CCL locks. 2016-07-15T16:56:04Z phoe_krk: It just seems that not to the full extent. 2016-07-15T16:56:34Z phoe_krk: The question is, do we need PROCESS-BLOCK which waits until a predicate returns a non-nil value. 2016-07-15T16:57:04Z phoe_krk: Shouldn't we rather rewrite the code so it uses BT's process waking mechanism? 2016-07-15T16:58:04Z phoe_krk: But that's a structural/design challenge, actually. 2016-07-15T16:58:23Z phoe_krk: We'd need to find usages of process-block, understand them and refactor the code. 2016-07-15T16:58:36Z rme: I couldn't say; I'm totally unfamiliar with CLX. It sounds like CLX may be designed around the lispm process model. 2016-07-15T16:59:05Z phoe_krk: Yes, it is. 2016-07-15T16:59:20Z phoe_krk: I popped up some of its code and the first thing I noticed was #+lispm. 2016-07-15T16:59:35Z phoe_krk: Which was like, wat, I wasn't even born when this stuff was written. 2016-07-15T17:00:09Z rme: You're standing on the shoulders of giants! 2016-07-15T17:00:55Z phoe_krk: I am! 2016-07-15T17:01:06Z phoe_krk: And these giants are wrapped in parens! 2016-07-15T17:04:26Z nzambe quit (K-Lined) 2016-07-15T17:06:15Z jdz: and they have hairy armpits! 2016-07-15T17:06:52Z phoe_krk: ... 2016-07-15T17:07:06Z phoe_krk: family channel dayum 2016-07-15T17:09:41Z phoe_krk: < beach> Thanks mmos! Thanks #ccl. :) 2016-07-15T17:09:44Z phoe_krk: Just so you people know. 2016-07-15T17:10:09Z phoe_krk: So - the workaround is to increase the ccl::*ticks-per-second*. 2016-07-15T17:10:20Z phoe_krk: Which is actually less of a workaround and more of a software upgrade. 2016-07-15T17:10:47Z phoe_krk: The solution is to fix CLX in that matter. Maybe that function can be optimized away completely by understanding what calls it and what it wanted to achieve by calling that. 2016-07-15T17:11:59Z rme: I guess so. I'd prefer it if the definition of process-block in dep-openmcl.lisp were changed not to call ccl:process-wait, and to be like sbcl's version. That way you're not messing around with unexported symbols like ccl::*ticks-per-second*. 2016-07-15T17:12:09Z phoe_krk: Ayup. 2016-07-15T17:12:31Z phoe_krk: But, um, I think these are two separate things, and they both deserve to be fixed. 2016-07-15T17:12:36Z phoe_krk: s/things/issues/ 2016-07-15T17:12:56Z phoe_krk: I'll file a CLX issue later today. 2016-07-15T17:13:12Z Tennis quit (Quit: Leaving) 2016-07-15T17:15:50Z tfb quit (Remote host closed the connection) 2016-07-15T17:17:07Z rme: Oh, *ticks-per-second* is exported (!). But setting it or binding it won't fix process-wait, because process-wait uses ccl::*ns-per-tick*, which is initialized from the value of *ticks-per-second*, but obviously won't be changed by setting *ticks-per-second* to something new. 2016-07-15T17:17:51Z emaczen quit (Read error: Connection reset by peer) 2016-07-15T17:18:15Z tfb joined #ccl 2016-07-15T17:19:01Z rme: Anyway: process-wait = bad news 2016-07-15T17:20:42Z phoe_krk: bad news! 2016-07-15T17:20:51Z phoe_krk: So this needs to be terribly fixed in CLX. 2016-07-15T17:21:06Z phoe_krk: rme: also no doubt it's exported. 2016-07-15T17:21:22Z phoe_krk: It's basically available everywhere on the system as it's basically a global kernel flag. 2016-07-15T17:21:22Z rme: process-wait? Certainly. 2016-07-15T17:22:23Z rme: All this made sense with lispm-style processes. With OS-scheduled threads, not so much now. 2016-07-15T17:23:27Z rme: Let's just rewrite everything in JavaScript and make everything run in a web browser. That'll make everything simpler and better, right? 2016-07-15T17:24:07Z phoe_krk: Yes! Things like left-padding a string take too much effort right now and they could be simplified and put in a library where everyone can load it from. 2016-07-15T17:24:08Z rme: And fashionable! 2016-07-15T17:24:13Z tfb quit (Remote host closed the connection) 2016-07-15T17:24:37Z phoe_krk: (rme: imagine what would happen if Xach removed SPLIT-SEQUENCE from the Quicklisp archives.) 2016-07-15T17:24:45Z tfb joined #ccl 2016-07-15T17:26:35Z jdz: we don't need javascript for that: https://gitlab.com/mbabich/trivial-left-pad 2016-07-15T17:26:45Z phoe_krk: ^ 2016-07-15T17:27:20Z jdz: ofc this would not exist without the javascript version being there first 2016-07-15T17:27:31Z jdz: all the things we can learn from the modern languages! 2016-07-15T17:28:02Z phoe_krk: yes 2016-07-15T17:28:21Z phoe_krk: lesson I: don't make repos the npm waY 2016-07-15T17:28:25Z phoe_krk: bbl 2016-07-15T17:40:33Z phoe_krk quit (Ping timeout: 240 seconds) 2016-07-15T17:41:23Z tfb quit (Remote host closed the connection) 2016-07-15T17:44:03Z milanj quit (Quit: Leaving) 2016-07-15T18:18:25Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-15T18:20:26Z eschatologist joined #ccl 2016-07-15T18:38:35Z nzambe joined #ccl 2016-07-15T18:39:14Z phoe_krk joined #ccl 2016-07-15T18:39:16Z phoe_krk: 2016-07-15T16:48:59Z rme: So, yeah, I'd probably change process-block to do what sbcl does. (loop (when (apply predicate args) (return) (#_sched_yield))) and call it a day. 2016-07-15T18:39:21Z phoe_krk: rme: is that loop well formed? 2016-07-15T18:39:37Z phoe_krk: That WHEN seems to return well before #_sched_yield gets called every single time! 2016-07-15T18:39:53Z rme: I typed in in IRC, so it might not be. 2016-07-15T18:40:10Z rme: But I hope you get the idea. 2016-07-15T18:41:36Z phoe_krk: I'll try to. 2016-07-15T18:42:10Z phoe_krk: But wait. 2016-07-15T18:42:17Z phoe_krk: (defun process-block (whostate predicate &rest predicate-args) 2016-07-15T18:42:17Z phoe_krk: (declare (ignore whostate)) 2016-07-15T18:42:17Z phoe_krk: (or (apply predicate predicate-args) 2016-07-15T18:42:17Z phoe_krk: (error "Program tried to wait with no scheduler."))) 2016-07-15T18:42:19Z phoe_krk: This is SBCL. 2016-07-15T18:42:26Z phoe_krk: There's... no loop whatsoever here. 2016-07-15T18:42:33Z phoe_krk: Just a single APPLY. 2016-07-15T18:42:52Z phoe_krk: And no yield. 2016-07-15T18:43:09Z rme: https://github.com/sharplispers/clx/blob/2aa969ee2abd777a76765e6713986cf3f5919be8/dependent.lisp#L1171 2016-07-15T18:43:26Z phoe_krk: Oh wait. 2016-07-15T18:43:28Z phoe_krk: That's the wrong line. 2016-07-15T18:43:52Z phoe_krk: Got it. I'll modify the CCL code. 2016-07-15T18:49:33Z rme: phoe_krk: Instead of #_sched_yield, I'd recommend ccl:process-allow-schedule. That'll cover the MS Windows case. 2016-07-15T18:52:34Z ekinmur joined #ccl 2016-07-15T18:52:39Z eschatologist quit (Ping timeout: 250 seconds) 2016-07-15T18:56:00Z phoe_krk: rme: Thanks. 2016-07-15T18:56:04Z ekinmur quit (Client Quit) 2016-07-15T18:56:09Z phoe_krk: Is that an exported function? 2016-07-15T18:57:51Z eschatologist joined #ccl 2016-07-15T19:01:18Z rme: yes, it is 2016-07-15T19:02:52Z phoe_krk: Got it. I'll use that. 2016-07-15T19:05:50Z phoe_krk: This whole issue had a positive side effect. 2016-07-15T19:06:01Z phoe_krk: I actually had to install CCL to fix it. 2016-07-15T19:06:17Z rme: Ha! 2016-07-15T19:06:21Z phoe_krk: And since it's not a matter of `apt install ccl`... 2016-07-15T19:06:47Z hhdave joined #ccl 2016-07-15T19:07:32Z rme: It's still a single command to install CCL, though. Did you follow http://ccl.clozure.com/download.html? 2016-07-15T19:10:08Z phoe_krk: Nah, I grabbed the tarball release. 2016-07-15T19:10:17Z hhdave_ joined #ccl 2016-07-15T19:10:31Z phoe_krk: Unzipped it into /opt/ccl/ and made a symlink. 2016-07-15T19:10:57Z phoe_krk: Like, /usr/bin/ccl -> /opt/ccl/lx86cl64woo386 2016-07-15T19:11:04Z phoe_krk: Or whatever other funny name that proper binary has. 2016-07-15T19:11:23Z hhdave quit (Ping timeout: 258 seconds) 2016-07-15T19:11:23Z hhdave_ is now known as hhdave 2016-07-15T19:13:15Z rme: Oh, that's OK, too. I keep making those tar archives for releases, but I never really know when someone uses them. My web analytics are poor to non-existent. 2016-07-15T19:13:44Z phoe_krk: Actually, umm. 2016-07-15T19:13:46Z phoe_krk: https://github.com/sharplispers/clx/blob/master/dependent.lisp#L1164 <- 2016-07-15T19:13:59Z phoe_krk: Should I defun ccl::yield like this for CCL? 2016-07-15T19:14:19Z phoe_krk: And call the yield function instead of calling the #_ internal one? 2016-07-15T19:14:36Z phoe_krk: Because I think it would be good style. 2016-07-15T19:14:46Z rme: I wouldn't. That function yield is the equivalent of ccl:process-allow-schedule. 2016-07-15T19:15:12Z phoe_krk: I see. 2016-07-15T19:15:46Z phoe_krk: Would the declarations in that case matter? 2016-07-15T19:15:49Z phoe_krk: Like, speed (safety 0)? 2016-07-15T19:16:01Z rme: As I mentioned, I would say don't call #_sched_yield. Call ccl:process-allow-scheule, which will work on Windows as well as Unix-oids. 2016-07-15T19:16:16Z phoe_krk: Lldkfjlskdf. 2016-07-15T19:16:17Z phoe_krk: I meant that. 2016-07-15T19:16:22Z phoe_krk: To call that function. 2016-07-15T19:18:18Z rme: No; declarations won't affect anything relating to calling ccl:process-allow-schedule, so they're not necessary. 2016-07-15T19:18:54Z phoe_krk: I see. 2016-07-15T19:19:02Z phoe_krk - +2 to Lisp smarts. 2016-07-15T19:29:27Z MetaHertz quit (Ping timeout: 276 seconds) 2016-07-15T20:03:47Z hhdave quit (Quit: hhdave) 2016-07-15T20:39:31Z phoe_krk: https://github.com/sharplispers/clx/pull/31 2016-07-15T21:19:56Z tfb joined #ccl 2016-07-16T01:57:58Z emaczen joined #ccl 2016-07-16T01:58:38Z emaczen quit (Changing host) 2016-07-16T01:58:38Z emaczen joined #ccl 2016-07-16T02:03:51Z jdz quit (Ping timeout: 258 seconds) 2016-07-16T02:09:34Z jdz joined #ccl 2016-07-16T02:17:12Z warcat joined #ccl 2016-07-16T02:17:37Z warcat: Is there a way to "unload" a loaded compiled file? 2016-07-16T02:19:20Z edgar-rft quit (Quit: edgar-rft) 2016-07-16T02:21:35Z rme: warcat: No. You can try to undo certain things. For instance, if your file defined a function "foo", you could say (unintern 'foo) or perhaps (fmakunbound 'foo) to get rid of that function. 2016-07-16T02:37:10Z warcat quit (Ping timeout: 250 seconds) 2016-07-16T02:37:40Z pierpa quit (Ping timeout: 264 seconds) 2016-07-16T02:38:50Z warcat joined #ccl 2016-07-16T02:39:19Z rme: warcat: Did you see my reply? 2016-07-16T02:39:27Z warcat: yes, I'm trying it now 2016-07-16T02:39:47Z warcat: hope it works, it was in a macro... clm's definstrument 2016-07-16T02:39:52Z rme: OK. I didn't know if you had disconnected and didn't see. 2016-07-16T02:41:59Z warcat: > Error: The value NIL is not of the expected type MACPTR. 2016-07-16T02:42:02Z warcat: > Error: The value NIL is not of the expected type MACPTR. 2016-07-16T02:42:07Z warcat: whoops sorry 2016-07-16T02:42:18Z warcat: I guess I'll have to restart ccl 2016-07-16T04:04:04Z Ando joined #ccl 2016-07-16T04:06:05Z Ando: I'm having trouble compiling CCl 1.11 from the tarball for hardfloat arm6 lon linux, ldd reports armcl not a dynamic executable. Anyone heard of this? 2016-07-16T04:10:27Z rme: Ando: It might be related to http://trac.clozure.com/ccl/ticket/1351. 2016-07-16T04:11:29Z rme: Try editing the Makefile, and removing the "-T ./armlinux.x" bit. 2016-07-16T04:11:33Z rme: See also http://trac.clozure.com/ccl/ticket/1349 2016-07-16T04:11:48Z Ando: I'll try that now thanks 2016-07-16T04:13:18Z Ando: No browser atm :) 2016-07-16T04:14:11Z Ando: On an unrelated note, I couldn't get svn to d/l the repo, it just hung. I had to d/l the tarball 2016-07-16T04:14:22Z Ando: 2016-07-16T04:14:27Z rme: The development version of ccl (and therefore the next release, when it comes out) will require armv7, by the way. 2016-07-16T04:14:34Z Ando: Meh 2016-07-16T04:14:45Z Ando: :D I'll stick with 1.11 then 2016-07-16T04:14:56Z Ando: If I can get it to work that is 2016-07-16T04:15:28Z eschatologist: That's a little bit of a bummer since the original Raspberry Pi and Pi 0 are armv6, and CCL works great on them. 2016-07-16T04:15:48Z Ando: That compiled, thank you ver much :) 2016-07-16T04:15:52Z eschatologist: But I know how every platform adds maintenance overhead, believe me. ;) 2016-07-16T04:16:21Z Ando: I do believe you :D 2016-07-16T04:16:23Z rme: Yeah, Clozure had a customer who wanted to use a armv6 device, so we added it. 2016-07-16T04:16:33Z Ando: Ah, OK 2016-07-16T04:17:40Z rme: But, now I think the customer uses a different board. And armv7 has a couple of nice instructions that are handy. like movt, dmb, and I forget what else at the moment. 2016-07-16T04:18:16Z rme: Ando: So it works? Good. 2016-07-16T04:18:29Z Ando: Thing is, Lisp is timeless, and the last version is forever (disregarding bugfixes) so at least we have it 2016-07-16T04:19:14Z rme: Right. 1.11 doesn't magically quit working. 2016-07-16T04:19:21Z Ando: Yep, just did (*3 4) to test it, as you do. Now to add it as emacs inferior lisp mode... :D Many thanks 2016-07-16T04:19:43Z Ando: I'm still using Clisp on an Atari Falcon from <2000, no problems 2016-07-16T04:20:23Z Ando: bit slow but thats what you get with ~50Mhz lol 2016-07-16T04:21:10Z rme: MCL 2.01 still runs on my Macintosh SE with 4 MB of memory and an I-don't-remember-how-slow MC68000. 2016-07-16T04:21:30Z rme: 8 MHz, maybe? 2016-07-16T04:21:30Z Ando: M68K vanilla was 8MHz 2016-07-16T04:21:47Z Ando: Nice machine :) 2016-07-16T04:22:16Z rme: It has a nice keyboard. :-) 2016-07-16T04:22:48Z Ando: I'd kill for a PS/2 Atari ST keyboard right now 2016-07-16T04:23:17Z rme: Totally off-topic, but http://www.thoughtstuff.com/rme/weblog/?p=21 is something I made on that Macintosh SE back in the day. 2016-07-16T04:23:22Z rme: Fun times. 2016-07-16T04:23:49Z rme: But Ando doesn't have a browser. Pity. :-) 2016-07-16T04:24:07Z rme: At least lisp works. 2016-07-16T04:24:10Z Ando: I don't have a browser atm, but I've saved the chatlog to look later 2016-07-16T04:24:26Z Ando: Got lisp, got emacs, don't need much else lol 2016-07-16T04:24:49Z Ando: hehe 2016-07-16T04:26:42Z Ando: Just gonna restart emacs to see if I can get ccl working with slime, thanks again for your help 2016-07-16T04:26:53Z rme: My pleasure. Good luck. 2016-07-16T04:27:01Z Ando: Tnx :) 2016-07-16T04:27:11Z Ando quit (Remote host closed the connection) 2016-07-16T04:28:00Z phf: i still have a dedicated ibook for mcl work.. 2016-07-16T04:29:02Z rme: I have a Mac OS X 10,6 Server VM for that, now that my PowerBook G4 croaked. 2016-07-16T04:29:58Z rme: I could dust off the Blue & White G3 in the closet, maybe. Or the PowerMac 8600 in the closet. 2016-07-16T04:30:29Z rme: I threw out all my old Sun SPARC hardware so I could have room for newer obsolete computers. 2016-07-16T04:30:41Z phf: ha 2016-07-16T04:32:44Z rme: Shoulda kept the NeXT cube. I really liked that machine. 2016-07-16T04:33:08Z phf: i have mcl running on a handmedown ibook g3, that someone from 2600 meeting gave me when i moved to this country. that was also the combination i used to make my first consulting money, so it has sentimental value for me 2016-07-16T04:34:51Z phf: i periodically consider ebaying a beefier machine, but never got to it 2016-07-16T04:36:49Z rme: Yeah, I think about grabbing a PowerMac G5 to keep the PowerPC port of CCL alive, but then I get over it. It's a pity that the port is more-or-less dead. If IBM or someone donated a Power8 machine, maybe I'd think about bringing it back to life. 2016-07-16T04:37:53Z rme: I think CCL assumes that the page size is 4K, so that assumption would need to be revised. (I'm pretty sure the Linux on those Power8 machines uses a 64K page size.) 2016-07-16T04:43:33Z phf: hmm, i thought that linux has a fixed page size at 4k, but it would make sense that it would support something bigger on power8. as far as i recall power8 can do variable sizes 2016-07-16T07:29:03Z rme quit (Quit: rme) 2016-07-16T07:48:58Z MetaHertz joined #ccl 2016-07-16T09:28:11Z eschatologist quit (Max SendQ exceeded) 2016-07-16T09:30:12Z eschatologist joined #ccl 2016-07-16T10:06:30Z warcat quit (Ping timeout: 244 seconds) 2016-07-16T10:27:31Z MetaHertz quit (Ping timeout: 252 seconds) 2016-07-16T10:47:50Z DavidGu joined #ccl 2016-07-16T11:14:57Z edgar-rft joined #ccl 2016-07-16T11:43:31Z DavidGu quit (Ping timeout: 240 seconds) 2016-07-16T12:00:06Z StephanLahl quit (Remote host closed the connection) 2016-07-16T12:14:26Z rme joined #ccl 2016-07-16T14:12:28Z tfb quit (Remote host closed the connection) 2016-07-16T14:17:10Z MetaHertz joined #ccl 2016-07-16T14:40:54Z DougNYC joined #ccl 2016-07-16T14:48:26Z pierpa joined #ccl 2016-07-16T15:50:39Z rumbler31 joined #ccl 2016-07-16T15:53:31Z MetaHertz quit (Ping timeout: 240 seconds) 2016-07-16T16:12:28Z MetaHertz joined #ccl 2016-07-16T16:14:12Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-16T16:17:12Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-16T16:57:31Z MetaHertz joined #ccl 2016-07-16T17:16:06Z MetaHertz quit (Ping timeout: 272 seconds) 2016-07-16T19:03:40Z MetaHertz joined #ccl 2016-07-16T19:07:07Z warcat joined #ccl 2016-07-16T19:11:49Z StephanLahl joined #ccl 2016-07-16T20:06:31Z ekinmur joined #ccl 2016-07-16T20:28:34Z MetaHertz quit (Ping timeout: 244 seconds) 2016-07-16T21:11:29Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-16T21:14:14Z gendl quit (Ping timeout: 258 seconds) 2016-07-16T21:16:04Z gendl joined #ccl 2016-07-16T22:02:09Z warcat quit (Ping timeout: 246 seconds) 2016-07-16T22:50:42Z emaczen quit (Ping timeout: 276 seconds) 2016-07-16T22:55:27Z emaczen joined #ccl 2016-07-16T23:01:35Z edgar-rft quit (Quit: edgar-rft) 2016-07-16T23:10:46Z warcat joined #ccl 2016-07-16T23:15:24Z warcat quit (Ping timeout: 276 seconds) 2016-07-16T23:18:46Z Blkt quit (Remote host closed the connection) 2016-07-16T23:18:47Z fe[nl]ix quit (Read error: Connection reset by peer) 2016-07-16T23:20:19Z Blkt joined #ccl 2016-07-16T23:20:38Z fe[nl]ix joined #ccl 2016-07-17T01:01:40Z leb joined #ccl 2016-07-17T01:05:50Z ekinmur joined #ccl 2016-07-17T01:45:42Z leb quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-17T01:51:33Z rumbler31 joined #ccl 2016-07-17T02:39:07Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-17T03:10:30Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-17T03:22:24Z pierpa quit (Ping timeout: 246 seconds) 2016-07-17T03:29:04Z rumbler31 joined #ccl 2016-07-17T03:36:24Z rumbler31 quit (Ping timeout: 246 seconds) 2016-07-17T03:42:32Z rumbler31 joined #ccl 2016-07-17T04:00:44Z edgar-rft joined #ccl 2016-07-17T04:32:02Z rme quit (Quit: rme) 2016-07-17T04:32:02Z rme quit (Quit: rme) 2016-07-17T06:01:02Z DougNYC quit 2016-07-17T06:16:03Z rumbler31 quit (Ping timeout: 264 seconds) 2016-07-17T06:30:20Z rumbler31 joined #ccl 2016-07-17T08:54:05Z ccl-logbot joined #ccl 2016-07-17T08:54:05Z 2016-07-17T08:54:05Z names: ccl-logbot DavidGu rumbler31 edgar-rft fe[nl]ix Blkt emaczen gendl StephanLahl eschatologist jdz phoe_krk nzambe schjetne Quadrescence hjudt carvite knobo1 ChibaPet brucem |3b| faheem billstclair taij33n akkad kini sigjuice mjl mdc_mobile Intensity john-mcaleely gbyers gz__ clop jackdaniel dustinm` sixbitslacker phf @ChanServ copec 2016-07-17T09:25:45Z DavidGu quit (Ping timeout: 276 seconds) 2016-07-17T10:28:48Z rme joined #ccl 2016-07-17T11:30:32Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-17T11:39:20Z rumbler31 joined #ccl 2016-07-17T12:47:49Z MetaHertz joined #ccl 2016-07-17T13:01:58Z rme quit (Quit: rme) 2016-07-17T13:54:49Z MetaHertz quit (Ping timeout: 250 seconds) 2016-07-17T14:28:31Z rumbler31 quit (Ping timeout: 240 seconds) 2016-07-17T14:35:11Z ekinmur joined #ccl 2016-07-17T15:41:03Z rumbler31 joined #ccl 2016-07-17T15:48:07Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-17T16:09:52Z MetaHertz joined #ccl 2016-07-17T16:19:46Z DougNYC joined #ccl 2016-07-17T16:22:27Z rme joined #ccl 2016-07-17T16:25:09Z phoe_krk quit (Quit: leaving) 2016-07-17T16:37:35Z rumbler31 quit (Ping timeout: 244 seconds) 2016-07-17T16:59:36Z pierpa joined #ccl 2016-07-17T17:08:12Z MetaHertz quit (Quit: Leaving) 2016-07-17T17:18:13Z rumbler31 joined #ccl 2016-07-17T17:26:04Z MetaHert` joined #ccl 2016-07-17T17:59:31Z rumbler31 quit (Ping timeout: 252 seconds) 2016-07-17T18:04:34Z rumbler31 joined #ccl 2016-07-17T18:48:57Z MetaHert` quit (Quit: ERC Version 5.3 (IRC client for Emacs)) 2016-07-17T18:53:39Z rme quit (Quit: rme) 2016-07-17T18:53:40Z rme quit (Quit: rme) 2016-07-17T18:54:26Z schjetne quit (*.net *.split) 2016-07-17T18:54:26Z hjudt quit (*.net *.split) 2016-07-17T18:54:27Z clop quit (*.net *.split) 2016-07-17T18:54:27Z jackdaniel quit (*.net *.split) 2016-07-17T18:54:27Z dustinm` quit (*.net *.split) 2016-07-17T18:59:56Z ekinmur joined #ccl 2016-07-17T19:11:17Z MetaHertz joined #ccl 2016-07-17T19:13:24Z jackdaniel joined #ccl 2016-07-17T19:13:25Z hjudt joined #ccl 2016-07-17T19:13:28Z dustinm` joined #ccl 2016-07-17T19:13:44Z clop joined #ccl 2016-07-17T19:16:57Z schjetne joined #ccl 2016-07-17T19:22:50Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-17T19:59:34Z MetaHertz quit (Ping timeout: 240 seconds) 2016-07-17T20:19:44Z ekinmur joined #ccl 2016-07-17T21:16:56Z rme joined #ccl 2016-07-17T21:21:52Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-17T22:04:38Z DougNYC quit (Remote host closed the connection) 2016-07-17T22:08:07Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-17T22:48:03Z rumbler31 joined #ccl 2016-07-17T23:33:13Z DougNYC joined #ccl 2016-07-17T23:38:19Z DougNYC quit (Ping timeout: 252 seconds) 2016-07-18T00:26:17Z DougNYC joined #ccl 2016-07-18T00:57:38Z ekinmur joined #ccl 2016-07-18T03:05:32Z eschatologist: Is there a guide to building CCL if you check out ^/trunk rather than a specific arch version? 2016-07-18T03:05:54Z eschatologist: I'd like to switch to building it from the git mirror, because that makes it *much* easier to maintain local patches. 2016-07-18T03:06:24Z rme: Hang on a second, and let me see if I have notes for that already written up. 2016-07-18T03:06:50Z eschatologist: cool 2016-07-18T03:07:39Z rme: Here's this: https://lists.clozure.com/pipermail/openmcl-devel/2015-December/011164.html 2016-07-18T03:07:50Z rme: Let me know if that makes sense. It's what I do myself when I check out ccl. 2016-07-18T03:08:55Z rme: svn co http://svn.clozure.com/publicsvn/openmcl/trunk/darwinx86/ccl is handly if you jjust want to grab a working ccl with one command, but it's not so convenient for actually hacking on ccl itself, because it uses svn externals is a sort of weird way. 2016-07-18T03:09:15Z rme: Also, I take patches, by the way. 2016-07-18T03:09:58Z eschatologist: Heh. Alas, I probably wouldn't be able to share. 2016-07-18T03:12:20Z rme: One reason I suggested changing CCL's license to Apache 2 is that I hoped that make make it easier for people to contribute. But I understand that even that may not even matter for certain organizations. 2016-07-18T03:18:00Z eschatologist: Yep. It's not a matter of CCL's own license. 2016-07-18T03:18:30Z eschatologist: The details on that web site make sense. 2016-07-18T03:18:36Z eschatologist: That'd let me pull the ARMv6 stuff back in. :) 2016-07-18T03:23:33Z rme: It's possible that a paying client might be able to get armv6 back in trunk, or maybe keep 1.11 supported until the armv6 hardware goes away. (I only want to pointi out the paid option, so please don't be offended. I'm glad to see people using CCL in any case.) 2016-07-18T03:24:45Z eschatologist: No worries. 2016-07-18T03:31:48Z DougNYC quit 2016-07-18T03:32:55Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-18T03:35:10Z rme quit (Quit: rme) 2016-07-18T03:35:10Z rme quit (Quit: rme) 2016-07-18T05:18:56Z pierpa quit (Ping timeout: 272 seconds) 2016-07-18T06:07:13Z warcat joined #ccl 2016-07-18T08:50:57Z MetaHertz joined #ccl 2016-07-18T09:14:35Z Tennis joined #ccl 2016-07-18T10:04:04Z MetaHertz quit (Ping timeout: 264 seconds) 2016-07-18T10:52:22Z MetaHertz joined #ccl 2016-07-18T11:09:43Z warcat quit (Quit: leaving) 2016-07-18T11:50:54Z hhdave joined #ccl 2016-07-18T12:13:48Z wgslayer joined #ccl 2016-07-18T12:37:13Z wgslayer quit (Quit: Leaving) 2016-07-18T12:55:27Z MetaHertz quit (Remote host closed the connection) 2016-07-18T13:15:48Z knobo1: How do I reload foreign library pointers in a saved image? 2016-07-18T13:18:14Z ekinmur joined #ccl 2016-07-18T13:46:51Z rme joined #ccl 2016-07-18T14:46:05Z pmicossi joined #ccl 2016-07-18T15:25:18Z pmicossi quit (Quit: ERC (IRC client for Emacs 24.5.1)) 2016-07-18T16:11:57Z hhdave quit (Ping timeout: 260 seconds) 2016-07-18T17:04:28Z mjl quit (Ping timeout: 272 seconds) 2016-07-18T17:04:28Z gbyers quit (Ping timeout: 272 seconds) 2016-07-18T17:06:23Z sigjuice quit (Ping timeout: 272 seconds) 2016-07-18T17:06:23Z sixbitslacker quit (Ping timeout: 272 seconds) 2016-07-18T17:06:26Z gbyers joined #ccl 2016-07-18T17:07:01Z hhdave joined #ccl 2016-07-18T17:11:52Z eschatologist quit (Ping timeout: 264 seconds) 2016-07-18T17:12:56Z mjl joined #ccl 2016-07-18T17:13:09Z sigjuice joined #ccl 2016-07-18T17:14:40Z dfigrish joined #ccl 2016-07-18T17:14:45Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-18T17:14:49Z dfigrish: hello everyone! 2016-07-18T17:16:15Z nzambe joined #ccl 2016-07-18T17:18:10Z rme: Hello 2016-07-18T17:18:14Z eschatologist joined #ccl 2016-07-18T17:32:40Z hhdave quit (Quit: hhdave) 2016-07-18T17:57:54Z pmicossi joined #ccl 2016-07-18T18:09:18Z dfigrish: rme: hi Matthew 2016-07-18T18:16:51Z pierpa joined #ccl 2016-07-18T18:42:36Z hhdave joined #ccl 2016-07-18T19:27:17Z ralfD joined #ccl 2016-07-18T19:33:06Z DavidGu1 joined #ccl 2016-07-18T19:51:21Z pierpa quit (Read error: Connection reset by peer) 2016-07-18T19:51:36Z pierpa joined #ccl 2016-07-18T19:57:40Z DavidGu1 quit (Ping timeout: 252 seconds) 2016-07-18T20:01:51Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-18T20:07:04Z ralfD quit (Ping timeout: 264 seconds) 2016-07-18T20:11:33Z rumbler31 joined #ccl 2016-07-18T21:35:50Z emaczen quit (Changing host) 2016-07-18T21:35:50Z emaczen joined #ccl 2016-07-18T21:40:23Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-18T21:49:53Z DougNYC joined #ccl 2016-07-18T22:30:32Z dfigrish quit (Ping timeout: 260 seconds) 2016-07-18T23:13:37Z hhdave quit (Quit: hhdave) 2016-07-19T00:28:44Z edgar-rft quit (Quit: edgar-rft) 2016-07-19T00:43:40Z ekinmur joined #ccl 2016-07-19T00:48:11Z ekinmur quit (Ping timeout: 240 seconds) 2016-07-19T00:50:01Z rumbler31: ccl::%file-kind is reporting that directories on my machine are of type :special and not :directory, causing problems. i'm on osx 10.11 with ccl 1.12-dev-r16752 2016-07-19T00:52:51Z rumbler31: when I try to follow the code in slime, I can't figure out how to discover the value of forms like #$S_IFMT, before I went much further I figured i'd ask if this was a known issue 2016-07-19T00:53:27Z rme: rumbler31: Can you please update and do (rebuild-ccl :full t)? I've made changes in the area of code that you mention, and I think that they should be fixed now. 2016-07-19T00:53:38Z rumbler31: sure thing! 2016-07-19T00:54:24Z rme: The particular problem had to do with issues calling the lstat system call. 2016-07-19T00:55:09Z rumbler31: when i get to it and try again i'll let you know, thanks 2016-07-19T00:55:30Z rme: OK. I'm about to step out for a walk, but I'll check back later. 2016-07-19T01:58:28Z ekinmur joined #ccl 2016-07-19T02:07:51Z rumbler31 quit (Ping timeout: 240 seconds) 2016-07-19T02:17:24Z ekinmur quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-19T02:25:02Z ekinmur joined #ccl 2016-07-19T02:58:03Z ekinmur quit (Ping timeout: 244 seconds) 2016-07-19T02:58:43Z ekinmur joined #ccl 2016-07-19T03:19:59Z rumbler31 joined #ccl 2016-07-19T03:36:34Z rumbler31: do you know if it is still necessary to issue a rebuild twice? there was an issue a while back that required this 2016-07-19T03:40:32Z rme: Often you can get away with doing it once, but there are circumstances where you have to do it two times. 2016-07-19T03:42:38Z rumbler31: it looks like ccl::%file-kind now reports directory, so thanks! 2016-07-19T03:42:52Z rme: Glad to hear it. 2016-07-19T03:47:25Z DougNYC quit (Read error: Connection reset by peer) 2016-07-19T03:47:32Z DougNYC joined #ccl 2016-07-19T04:03:24Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T04:24:18Z pierpa quit (Ping timeout: 276 seconds) 2016-07-19T04:34:17Z DougNYC quit 2016-07-19T05:26:05Z rumbler31 quit (Read error: Connection reset by peer) 2016-07-19T05:26:07Z rumbler3_ joined #ccl 2016-07-19T06:02:23Z rme quit (Quit: rme) 2016-07-19T06:02:23Z rme quit (Quit: rme) 2016-07-19T07:22:11Z rumbler3_ quit (Ping timeout: 240 seconds) 2016-07-19T07:25:53Z rumbler31 joined #ccl 2016-07-19T07:26:58Z edgar-rft joined #ccl 2016-07-19T08:07:50Z rumbler31 quit (Remote host closed the connection) 2016-07-19T08:56:38Z john-mcaleely_ joined #ccl 2016-07-19T08:57:11Z john-mcaleely_ quit (Read error: Connection reset by peer) 2016-07-19T09:32:32Z PuercoPop joined #ccl 2016-07-19T10:17:48Z edgar-rft quit (Quit: edgar-rft) 2016-07-19T12:22:01Z rme joined #ccl 2016-07-19T13:18:41Z ekinmur joined #ccl 2016-07-19T13:24:00Z edgar-rft joined #ccl 2016-07-19T13:40:19Z pmicossi` joined #ccl 2016-07-19T13:56:43Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T13:59:50Z moore33 joined #ccl 2016-07-19T14:02:10Z rme: Hello, everyone. Welcome to part 1 of today's ccl office hours. I'll be in the channel for the next hour or so to talk about whatever you want, as long as it applies to ccl somehow. 2016-07-19T14:02:55Z moore33: Hi! Thanks for doing this. 2016-07-19T14:02:59Z rme: Hi, Tim. 2016-07-19T14:03:52Z moore33: Unfortunately, ccl is working too well for my purposes to generate any questions for you :/ 2016-07-19T14:04:16Z rme: Heh. 2016-07-19T14:05:19Z moore33: If anyone is curious, I am running ccl on ARM Linux, doing 3D programming in Lisp using cl-opengl to interface to OpenGL. It all seems to work well. 2016-07-19T14:06:08Z rme: You don't happen to have any suggestions for a sort of "hello world" program using modern OpenGL, do you? I have the recent OpenGL red book, but I never have used the programmable pipeline with shaders and all that. 2016-07-19T14:06:09Z jackdaniel uses ccl on linux-arm box too 2016-07-19T14:07:05Z fpt joined #ccl 2016-07-19T14:07:56Z moore33: rme:opengl-tutorial.org seems OK at first glance. 2016-07-19T14:08:10Z moore33: It follows modern OpenGL practice. 2016-07-19T14:08:35Z rme: Thanks; I'll bookmark it and give it a try. 2016-07-19T14:11:55Z rme: moore33: By the way, I bought that Bordeaux door book on behalf of an architect friend. His wife is going to give it to him next month as a birthday present. 2016-07-19T14:12:52Z rme: http://www.histoires-de-portes.fr is the book in question for other readers. But this isn't really ccl related, is it? 2016-07-19T14:13:38Z moore33: rme: Heh, did you carry it home with you? 2016-07-19T14:15:18Z rme: No, I ordered it from Mollat. Shipping was just 15 EUR, and I think I got a break on TVA, too. 2016-07-19T14:17:08Z moore33: Oh, that's not bad at all. 2016-07-19T14:20:48Z rme: For you ARM users, I suppose I should warn you that the development version of ccl (and therefore the next release) requires an armv7 processor. This means that future ccl versions will run on the Raspberry Pi 3, but not earlier ones. 2016-07-19T14:24:45Z rme: Most everything for several years has used hardware that implements armv7, so this should generally be no big deal (except for RPi users before version 3, like I said). 2016-07-19T14:26:09Z moore33: So BeagleBone Black will continue to be fine. 2016-07-19T14:26:46Z hydan joined #ccl 2016-07-19T14:27:16Z fpt: RE: armv7: Good to know. Thanks. 2016-07-19T14:31:00Z mjl quit (Max SendQ exceeded) 2016-07-19T14:32:01Z ekinmur joined #ccl 2016-07-19T14:32:32Z rme: If you're lurking out there, don't be shy. 2016-07-19T14:34:51Z mjl joined #ccl 2016-07-19T14:38:05Z moore33: rme: Do you have a large startup file for ccl with various customizations, or do you keep it pretty close to the default setup? 2016-07-19T14:38:53Z rme: I have two lines in my ccl-init.lisp file. (ccl::set-development-environment), and (setq ccl::*did-show-marketing-blurb* t). 2016-07-19T14:39:05Z rme: Sometimes I add (setq ccl:*load-preserves-optimization-settings* t) 2016-07-19T14:39:43Z moore33: Does set-development-environment load up the hemlock editor et al? 2016-07-19T14:40:37Z rme: No, that just makes the compiler not complain when functions in the cl or ccl package (I think) get redefined. 2016-07-19T14:40:46Z moore33: Ah, ok. 2016-07-19T14:41:14Z rme: I use that because I work on ccl itself. Most people wouldn't use that. 2016-07-19T14:41:33Z rme: The load-preserves-optimization thing can be useful when loading systems from Quicklisp. 2016-07-19T14:42:37Z rme: Some crazy programmers put things like (declaim (optimize (speed 3) (safety 0))) at the top of a source file, not understanding that the declaim has persistent load-time effects. 2016-07-19T14:42:58Z moore33: Oh right. 2016-07-19T14:43:00Z rme: The setting makes load preseve the optimization settings much in the same way that it preserves, say, *package*. 2016-07-19T14:43:25Z jackdaniel: I usually do (declaim (optimize (debug 3))) ;-) 2016-07-19T14:43:44Z moore33: Yeah, me too, though not in source files. 2016-07-19T14:44:04Z jdz: rme: any chances of having something like SBCL's sb-ext:restrict-compiler-policy? 2016-07-19T14:44:40Z rme: I don't believe (debug 3) really makes much difference on ccl. 2016-07-19T14:44:57Z jdz: for me the important bit is ability to restrict safety to 1 2016-07-19T14:46:07Z jdz: but not important enough to pay for the feature :/ 2016-07-19T14:46:17Z rme: jdz: There's a thing called a compiler-policy object that you can customize. I don't think it's well documented, if at all. 2016-07-19T14:47:26Z jdz: this is one of those things one does not need until they learn about it and then that's the best thing since sliced bread 2016-07-19T14:48:14Z rme: There's http://trac.clozure.com/ccl/wiki/DeclareOptimize 2016-07-19T14:49:27Z rme: Basically, you can call (ccl:new-compiler-policy), fiddle with it as desired, and then use (ccl:set-current-compiler-policy new-policy). 2016-07-19T14:49:43Z jdz: rme: cool, thanks for digging that up! 2016-07-19T14:49:49Z rme: Maybe that's enough to get you started. 2016-07-19T14:50:32Z jdz: yes, when i get back to using CL... 2016-07-19T14:51:02Z jdz: for doing serious stuff, not just doing Jane Street puzzles 2016-07-19T15:03:56Z rme: Thanks everyone for coming to my office hours. I'll be back for part 2 at 20:00 UTC (4:00 pm US/Eastern). 2016-07-19T15:18:07Z pmicossi` quit (Quit: ERC (IRC client for Emacs 24.5.1)) 2016-07-19T15:25:19Z gbyers: The original raspi was an armv6. A year or so ago, an armv7 rasp pi was released. The v3 rasp pi is an armv8. The original armv6 version is not sold anymore, but a lot of them were sold at one point. 2016-07-19T15:49:23Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-19T15:49:35Z nzambe joined #ccl 2016-07-19T15:53:50Z okflo joined #ccl 2016-07-19T16:09:20Z hhdave joined #ccl 2016-07-19T16:41:52Z moore33 quit (Quit: Leaving) 2016-07-19T16:58:29Z pierpa joined #ccl 2016-07-19T17:00:54Z pmicossi quit (Quit: ERC (IRC client for Emacs 24.5.1)) 2016-07-19T17:40:29Z peter___ joined #ccl 2016-07-19T17:41:11Z peter___ quit (Client Quit) 2016-07-19T17:46:25Z lispm joined #ccl 2016-07-19T18:04:05Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T18:04:57Z ekinmur joined #ccl 2016-07-19T18:12:13Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-19T18:12:27Z nzambe joined #ccl 2016-07-19T18:22:23Z hydan quit (Quit: Textual IRC Client: www.textualapp.com) 2016-07-19T18:51:18Z lispm quit (Ping timeout: 250 seconds) 2016-07-19T18:54:29Z dfigrish joined #ccl 2016-07-19T18:55:04Z dfigrish: hi there 2016-07-19T19:20:06Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T19:28:57Z ekinmur joined #ccl 2016-07-19T19:31:01Z ekinmur quit (Client Quit) 2016-07-19T19:32:37Z ekinmur joined #ccl 2016-07-19T20:00:38Z rme: Hello, everyone. Welcome to part 2 of today's CCL office hours. I'll be here in the channel for the next hour or so to talk about whatever you want as long as it relates to CCL, however tangentially. 2016-07-19T20:03:27Z rme: So, please don't be shy. 2016-07-19T20:03:46Z pierpa: hello 2016-07-19T20:07:19Z gbyers is rarely shy 2016-07-19T20:11:28Z edgar-rft quit (Quit: edgar-rft) 2016-07-19T20:13:00Z dfigrish: rme: what is the development plan/strategy for CCL? 2016-07-19T20:13:43Z dfigrish: as I can see, there is no plan at all 2016-07-19T20:14:01Z rme: Other than "make it better"? There's not a 5 year plan, true. 2016-07-19T20:14:51Z rme: Try to take over the world? 2016-07-19T20:15:19Z rme: But seriously, maybe setting down some goals would be a good idea. 2016-07-19T20:15:38Z rme: That is a fair point. 2016-07-19T20:17:00Z dfigrish: But the "better" is a very relatively concept :) I believe that without the clear development plan everything is doomed for stagnation. 2016-07-19T20:21:36Z pierpa: dfigrish: what would you want to see in ccl? 2016-07-19T20:21:41Z fpt: Well what are some interesting or difficult challenges in the Lispish world these days? Is it moving to support smaller architectures, like ARM? Or supporting interfacing with GPUs? 2016-07-19T20:21:59Z gbyers: Rme just asked you what you wanted. I seem to have missed your answer 2016-07-19T20:22:52Z PuercoPop: rme: Hey thanks for doing this! How is boostrap process? What does level0 do? level1? 2016-07-19T20:23:35Z rme: http://ccl.clozure.com/docs/ccl.html#building-clozure-cl-from-its-source-code talks about it a little bit 2016-07-19T20:24:30Z rme: But, generally speaking, the files in level-0 get loaded into lisp datastructures, and then are written out as a bootstrapping heap image. 2016-07-19T20:25:07Z rme: When that image is loaded, it then loads the rest of the fasl files (from level-1 and elsewhere), and the lisp becomes more and more capable as it does so. 2016-07-19T20:26:09Z rme: When that's done, you get a prompt, and then you can run save-application at that point, and you have a basically fully-functional lisp. I would then start up from that image, and run (rebuild-ccl :clean t). 2016-07-19T20:27:44Z dfigrish: pierpa: well, actively development process :-) 2016-07-19T20:27:48Z rme: That's how cross-compiling works. You arrange to load the compiler backend for the target architecture (by ad hoc ways, mostly: this could be improved), and then you compile level-0 and generate the bootstrapping image like I said. 2016-07-19T20:28:24Z rme: Then, you cross compile all the fasls. After that, you go to the target system, load the bootstrapping image, and then it loads the rest of the fasls as usual. 2016-07-19T20:28:56Z eschatologist: rme: The Raspberry Pi was sold from Feb. 2012 and continues to be sold, at least in the form of the Pi Zero. The Pi 2 only came out in Feb. 2015, and the Pi 3 in Feb. 2016. The Pi 2 supports armv7, the Pi 3 is technically armv8/arm64 but also supports armv7. 2016-07-19T20:29:47Z rme: I was mistaken about the RPi situation. Thanks gbyers and eschatologist for clearing it up. It's just the original RPi that's not going to run future ccl. 2016-07-19T20:30:04Z eschatologist: And the RPi Zero which is the $5 compute board that's still sold. 2016-07-19T20:30:09Z eschatologist: :p 2016-07-19T20:30:14Z eschatologist: (meeting time) 2016-07-19T20:31:01Z PuercoPop: rme: thanks. Another question, in retrospect does rolling your own documentation markup language (ccldoc) instead of skribe/texinfo/ReST(although I imagine ReST didn't exist was back then) worth it? 2016-07-19T20:31:38Z gbyers: The original rpi was not available when i looked a few years ago 2016-07-19T20:31:41Z rme: http://trac.clozure.com/ccl/wiki/CrossCompiling might be interesting for someone. 2016-07-19T20:32:07Z rme: Ah, the documentation... 2016-07-19T20:32:20Z gbyers: A.few weeks ago. 2016-07-19T20:32:57Z dfigrish: how about to minimize and clean everything up, remove old stuff and focus on the x86-64 and ARM platforms? 2016-07-19T20:33:08Z gbyers: Trac also has a wiki markup language. 2016-07-19T20:33:20Z rme: We found that we didn't like texinfo. Then, someone converted the documentation to docbook, which I really, really, didn't like. I toyed with using TeX, and gbyers toyed with LyX, but nobody was really keen on anything. 2016-07-19T20:33:38Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T20:35:04Z rme: So, I think Gail (gz) must have gotten tired of all the whining, and she wrote ccldoc. It's handy because the document is represented as a lisp data structure, and it's really quite easy to write an output formatter for it. 2016-07-19T20:35:32Z ekinmur joined #ccl 2016-07-19T20:36:27Z rme: I personally have had a long-term goal that ccl's documentation could be like the hardcopy of CLTL. 2016-07-19T20:36:36Z alms_clozure joined #ccl 2016-07-19T20:37:21Z gbyers: dfigrish: the ppc is not supported anymore. The x8632 still is, as are x8664 and armv7. 2016-07-19T20:37:24Z pierpa: I wish there was a CLTLIII updated the final standard. And then an updated standard :) 2016-07-19T20:38:01Z pierpa: *updated to 2016-07-19T20:38:35Z rme: So, we want HTML and some sort of hardcopy-style output (probably TeX/LaTeX). I don't know if systems like Sphinx support that, but they come with a lot of other foreign ecosystem stuff. Markdown and ReST (well, Markdown; I'm not 100% sure about ReST) are a little *too* simple. 2016-07-19T20:39:13Z rme: PuercoPop: That's probably way, way, more than you wanted to hear. 2016-07-19T20:40:07Z phf: (fwiw there's a lot more practical i386 devices, than there is ARM still, so dropping 32bit support would be shortsighted) 2016-07-19T20:40:32Z PuercoPop: thanks! I've seen hu.dwim doing a sexp syntax docbook and recently CommonDoc as a way to represent a text document with CLOS. I've been wanting to write some extensions in the spirit of Racket's scribble, to easily link the online source repository and stuff 2016-07-19T20:40:50Z rme: Racket's scribble is very nice indeed. 2016-07-19T20:41:25Z PuercoPop: Ideally even a CL documentation browser for CommonDoc would be nice 2016-07-19T20:41:41Z dfigrish: gbyers: uh, ok. What about the library/? I mean, does anybody using the stuff like openmcl-gtk-support.lisp or jni.lisp etc? 2016-07-19T20:42:13Z PuercoPop: it makes it every easy to contribute documentation and keep it up to date. I've contributed a few typos due to the low energy barrier to do so 2016-07-19T20:42:20Z rme: Where did CommonDoc come from? I'm not sure I've ever seen it before today. 2016-07-19T20:44:10Z rme: It looks to be using an approach very much like ccldoc's. 2016-07-19T20:45:17Z rme: http://commondoc.github.io for those watching 2016-07-19T20:45:41Z gbyers: Dfigrish; probably not, at least in the case of the old gtk example. 2016-07-19T20:46:10Z PuercoPop: rme: Eudoxia wrote it after talking or reading a design of Robert Standh iirc 2016-07-19T20:47:10Z rme: I remember telling Robert about ccldoc. In any case, I think the approach is a good idea. 2016-07-19T20:47:41Z rme: But Robert had had the idea long before I mentioned ccldoc. 2016-07-19T20:49:34Z dfigrish: gbyers: then I would like to remove the stuff like this (or move it to the special old-stuff-place) 2016-07-19T20:50:31Z gbyers: Why? 2016-07-19T20:50:46Z dfigrish: gbyers: the goal is to facilitate maintenance. 2016-07-19T20:51:28Z dfigrish: gbyers: every kind of stuff needs care 2016-07-19T20:52:04Z gbyers: Is maintenance hard because an pld gtk example is distributed? I didnt realize that. 2016-07-19T20:54:05Z dfigrish: gbyers: the unuseful stuff brings some kind of chaos which simply prevents 2016-07-19T20:55:32Z phf: dfigrish: that's not true in universal sense, i use stale ccl code all the time because it often has very interesting if aborted ideas that can be reused 2016-07-19T20:57:04Z dfigrish: phf: I'm talking about the CCL repository :-) Why the aborted ideas should be inside it? 2016-07-19T21:00:08Z phf: dfigrish: because often times that's the only place quality lisp code is available. you're not going to go and create "repository of excellent lisp code" for all the "useless" things that you've killed, so discovering those pieces is going to become significantly harder than just doing cd ccl; grep -r 2016-07-19T21:01:22Z dfigrish: phf: how about the "obsolete but quality list code" repository? 2016-07-19T21:01:26Z gbyers: Many years ago, people used gtk 1 more than they do now. I think of the example as having some (often very small) value today, and i don't see the existence of that file as being a source of chaos for anyone. 2016-07-19T21:01:46Z rme: Let's talk about something else. 2016-07-19T21:02:05Z dfigrish: gbyers: except for me ;-) 2016-07-19T21:02:09Z gbyers: Yes, sorry 2016-07-19T21:02:13Z dfigrish: rme: okay ;-) 2016-07-19T21:05:05Z dfigrish: well, see you later. There is already midnight. Bye. 2016-07-19T21:05:15Z rme: See you. 2016-07-19T21:09:19Z rme: Please feel free to speak up, if you're lurking out there! 2016-07-19T21:10:10Z dfigrish quit (Ping timeout: 252 seconds) 2016-07-19T21:17:19Z rme: Well, thank you everyone for coming to office hours. It's fun to talk about ccl. 2016-07-19T21:17:27Z fpt: I mostly had gossip monitoring in mind. But WRT IDE stuff goals on the To Do page, "Bring back Fred," has there been any movement on that, or is there likely to be in the foreseeable future? 2016-07-19T21:19:27Z rme: fpt: I think about that a lot. I'm not sure if "bring back Fred" is the way to go, but I do feel a need to improve the editor in the IDE so that mouse and other event handling can be done in Lisp. 2016-07-19T21:20:30Z rme: Right now, the editor is a hybrid of NSTextView and Hemlock. 2016-07-19T21:23:13Z rme: I know I sound like a broken record, but if I could get some funding, I'd love to work on that. But I know that nobody is going to say, "here's $20k, show me what you've got after two months". Maybe the best thing to do is to get something promising working, and then see if interest comes. 2016-07-19T21:24:14Z fpt: Kickstart Son-of-Fred? 2016-07-19T21:24:22Z rme: Have you seen Swift Playgrounds in Xcode lately? That is pretty cool stuff, and CL would be a great fit for that kind of usage. 2016-07-19T21:26:43Z fpt: Oh yeah! SP looks way cool. Something like that could be a fabulous way to play with macros. 2016-07-19T21:27:11Z gbyers: Several months ago, someone report ed an issue: they needed to use a specific version of libssl and if they ran the ccl editor and clicked on the menubar, the wrong version of libssl was getting loaded. Aside from convincing myself thst no lisp code was likely involved, i did.not get very far on the bug. 2016-07-19T21:29:10Z rme: I also like the Mathematica notebook interface, and have thought some sort of similar "worksheet"-type interface would be neat in a CL context. 2016-07-19T21:29:38Z ekinmur quit (Quit: My Mac has gone to sleep. ZZZzzz…) 2016-07-19T21:30:47Z rme: That "load wrong libssl" bug is extremely strange, and makes approximately no sense. 2016-07-19T21:43:47Z rme: Again, thanks everyone. I'm off for now. 2016-07-19T21:43:54Z pierpa: ty 2016-07-19T21:46:30Z gbyers: I think that ron Garrett was the person who reported it, but im not sure how it was reportef 2016-07-19T21:47:41Z fpt: Thanks for your time today. 2016-07-19T21:51:21Z eschatologist: Loading "the wrong" SSL when clicking the menu bar doesn't surprise me, based on the way Mach-O loading works. 2016-07-19T21:51:35Z edgar-rft joined #ccl 2016-07-19T21:52:23Z eschatologist: Clicking on the menu bar probably brings in more OS libs, which will bring in the OS copy of libssl. 2016-07-19T21:52:36Z eschatologist: Anything that looks up libssl symbols *purely* by name will then find the symbols in that lib. 2016-07-19T22:14:27Z gbyers: That's generally what i thought, which suggests that the real fix likely involves fixing the broken libssl. I think that ron Garrett thought that some lisp code was more directly involved than it likely is. 2016-07-19T22:16:19Z PuercoPop: rme: there is one project for a notebook in CL (and another one to piggy back ipython/juniper). Here is a demo https://vimeo.com/97623064 2016-07-19T22:16:22Z PuercoPop: (and the code https://github.com/Inaimathi/cl-notebook) 2016-07-19T22:31:12Z hhdave quit (Quit: hhdave) 2016-07-19T22:52:20Z eschatologist: The broken (non-OS) libssl could potentially be fixed by using the handle explicitly returned from a dlopen() when using dlsym(), or by ensuring that the non-OS libssl is built with some symbol prefix (which I think OpenSSL supports) so the symbols are entirely distinct. 2016-07-19T23:20:23Z DougNYC joined #ccl 2016-07-19T23:54:53Z edgar-rft quit (Quit: edgar-rft) 2016-07-20T01:22:51Z edgar-rft joined #ccl 2016-07-20T03:11:52Z pierpa quit (Ping timeout: 264 seconds) 2016-07-20T03:16:07Z DougNYC quit 2016-07-20T03:23:37Z edgar-rft quit (Quit: edgar-rft) 2016-07-20T05:25:38Z rme quit (Quit: rme) 2016-07-20T05:25:38Z rme quit (Quit: rme) 2016-07-20T06:54:52Z okflo` joined #ccl 2016-07-20T06:58:52Z okflo quit (Ping timeout: 260 seconds) 2016-07-20T07:08:41Z eschatologist: Was the code for Digitool's CLIM ever made available? 2016-07-20T08:13:17Z vito_ joined #ccl 2016-07-20T08:13:40Z vito_ is now known as Guest96935 2016-07-20T08:28:22Z Guest96935 quit (Quit: Page closed) 2016-07-20T08:34:11Z dfigrish joined #ccl 2016-07-20T08:34:17Z dfigrish: hello 2016-07-20T09:09:00Z consolers` joined #ccl 2016-07-20T09:28:04Z consolers` quit (Ping timeout: 240 seconds) 2016-07-20T09:47:38Z pmicossi joined #ccl 2016-07-20T12:27:03Z rme joined #ccl 2016-07-20T12:42:37Z kai__ joined #ccl 2016-07-20T12:46:54Z kai__: hi all, i'm on linux trying to build ccl from source. i downloaded the binary for linux x64_64 and put it in the ccl-svn directory (with the image). then i run "? (ccl:rebuild-ccl :full t)" but hit a read error: "Read error between positions 17832 and 18582 in .../src/ccl/lib/compile-ccl.lisp". earlier the process complains that constants.cdb and vars.cdb are missing. do you have any ideas what i'm doing wrong? thanks! 2016-07-20T12:51:10Z pierpa joined #ccl 2016-07-20T13:11:24Z rme: kai__: You need a copy of the interface databases too. 2016-07-20T13:11:55Z kai__: are this the cdb files? 2016-07-20T13:12:04Z rme: Yes. 2016-07-20T13:12:38Z kai__: should i just copy all of them from the binary distribution? 2016-07-20T13:13:01Z kai__: probably just elf/libc 2016-07-20T13:13:09Z rme: If you look in the scripts/ directory, there's a file named "get-binaries". You can use this to get them. Or, you can just copy them from somewhere, like you said. 2016-07-20T13:13:42Z kai__: i don't have that script, but i'll just copy it by hand 2016-07-20T13:15:29Z kai__: rme: that worked, thanks! 2016-07-20T13:15:41Z rme: You're welcome. 2016-07-20T13:16:28Z kai__: is there a way to generate those files on my own? from a clean directory? 2016-07-20T13:16:42Z kai__: (i usually like to build stuff from source) 2016-07-20T13:20:13Z edgar-rft joined #ccl 2016-07-20T13:21:57Z rme: kai__: It's possible, but it's a hassle. The manual talks about it http://ccl.clozure.com/docs/ccl.html#the-interface-translator 2016-07-20T13:22:52Z rme: The .cdb files just contain information from .h files in a more convenient form. 2016-07-20T13:26:07Z rme: http://trac.clozure.com/ccl/wiki/CustomFramework has a few more details 2016-07-20T13:27:03Z kai__: cool, thanks for the info, i'll check it out :) 2016-07-20T13:28:25Z rme: I would advise you not to bother with it unless you really are interested for some reason. 2016-07-20T13:30:18Z kai__: Makes sense, I just like to know where the binary stuff I'm using is coming from. When building entierely from source I mostly feel much better about software, so I usually like to checkout what's inside binary blobs :) 2016-07-20T13:31:00Z kai__: But yes, I'm not overly sceptical about ccl, so let's see :) Thanks for you help! 2016-07-20T13:31:13Z rme: Good luck! 2016-07-20T13:39:47Z rme quit (Ping timeout: 185 seconds) 2016-07-20T13:42:32Z pierpa: I found out that (SETQ X X) is not optimized away by the compiler. Is this worth a ticket? (I had to complicate slightly a macro to make it remove this vacuous assignment, would be nice if this wasn't necessary) 2016-07-20T13:52:51Z rme: It's probably worth a ticket. 2016-07-20T13:53:05Z pierpa: ok 2016-07-20T13:53:07Z pierpa: will do 2016-07-20T13:54:21Z rme: I don't know if it's always safe to do that; for instance, in a with-accessors form. But when the thing being setq'd isn't a symbol-macro, it seems like it should be ok. 2016-07-20T13:55:34Z pierpa: k 2016-07-20T13:55:47Z pierpa: in my case it is a plain variable 2016-07-20T14:25:51Z dfigrish quit (Ping timeout: 240 seconds) 2016-07-20T14:45:59Z dfigrish joined #ccl 2016-07-20T15:02:49Z tfb joined #ccl 2016-07-20T15:08:07Z kai__ quit (Quit: leaving) 2016-07-20T15:26:14Z rme quit (Quit: rme) 2016-07-20T15:28:09Z pmicossi quit (Quit: going home) 2016-07-20T15:29:03Z edgar-rft quit (Quit: edgar-rft) 2016-07-20T15:43:16Z rme joined #ccl 2016-07-20T15:51:35Z tfb quit (Remote host closed the connection) 2016-07-20T16:04:34Z dfigrish quit (Ping timeout: 240 seconds) 2016-07-20T17:10:56Z dfigrish joined #ccl 2016-07-20T17:44:22Z edgar-rft joined #ccl 2016-07-20T19:24:28Z DavidGu joined #ccl 2016-07-20T19:28:31Z DavidGu quit (Ping timeout: 240 seconds) 2016-07-20T19:55:13Z okflo` quit (Remote host closed the connection) 2016-07-20T20:14:35Z rme quit (Ping timeout: 183 seconds) 2016-07-20T20:15:56Z rme quit (Ping timeout: 250 seconds) 2016-07-20T20:18:45Z rme joined #ccl 2016-07-20T20:22:55Z dfigrish: sea ya 2016-07-20T20:23:57Z dfigrish quit (Quit: rcirc on GNU Emacs 25.1.50.1) 2016-07-20T23:22:04Z DougNYC joined #ccl 2016-07-21T00:22:50Z DougNYC quit 2016-07-21T01:19:51Z pierpa quit (Ping timeout: 240 seconds) 2016-07-21T02:34:10Z DougNYC joined #ccl 2016-07-21T04:42:29Z edgar-rft quit (Quit: edgar-rft) 2016-07-21T05:11:27Z rme quit (Quit: rme) 2016-07-21T05:46:23Z DougNYC quit (Ping timeout: 244 seconds) 2016-07-21T07:02:28Z DougNYC joined #ccl 2016-07-21T07:07:40Z DougNYC quit (Ping timeout: 264 seconds) 2016-07-21T07:53:47Z DavidGu joined #ccl 2016-07-21T07:54:01Z DavidGu quit (Client Quit) 2016-07-21T08:04:21Z hhdave joined #ccl 2016-07-21T08:12:37Z hhdave quit (Ping timeout: 260 seconds) 2016-07-21T08:12:58Z hhdave joined #ccl 2016-07-21T08:20:10Z StephanLahl quit (Remote host closed the connection) 2016-07-21T08:27:59Z DavidGu joined #ccl 2016-07-21T08:51:27Z DougNYC joined #ccl 2016-07-21T08:55:43Z DougNYC quit (Ping timeout: 252 seconds) 2016-07-21T10:58:04Z tfb joined #ccl 2016-07-21T11:10:54Z MajorTennis joined #ccl 2016-07-21T11:14:38Z Tennis quit (Ping timeout: 250 seconds) 2016-07-21T11:40:36Z tfb quit (Remote host closed the connection) 2016-07-21T12:02:42Z MajorTennis quit (Ping timeout: 276 seconds) 2016-07-21T12:27:21Z DougNYC joined #ccl 2016-07-21T12:32:12Z DougNYC quit (Ping timeout: 250 seconds) 2016-07-21T12:59:27Z rme joined #ccl 2016-07-21T13:09:39Z pierpa joined #ccl 2016-07-21T13:29:46Z Tennis joined #ccl 2016-07-21T13:40:18Z StephanLahl joined #ccl 2016-07-21T13:57:40Z edgar-rft joined #ccl 2016-07-21T14:13:30Z rumbler31 joined #ccl 2016-07-21T14:16:10Z DougNYC joined #ccl 2016-07-21T14:55:33Z DougNYC quit (Ping timeout: 240 seconds) 2016-07-21T15:34:56Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-21T15:35:32Z nzambe joined #ccl 2016-07-21T16:08:03Z DougNYC joined #ccl 2016-07-21T16:12:11Z DougNYC quit (Ping timeout: 240 seconds) 2016-07-21T16:23:58Z rumbler31 quit (Ping timeout: 252 seconds) 2016-07-21T16:24:14Z Tennis quit (Quit: Leaving) 2016-07-21T16:37:40Z hhdave quit (Ping timeout: 264 seconds) 2016-07-21T16:57:26Z edgar-rft quit (Quit: edgar-rft) 2016-07-21T17:17:54Z rumbler31 joined #ccl 2016-07-21T17:38:08Z dfigrish joined #ccl 2016-07-21T17:38:13Z dfigrish: hello 2016-07-21T17:43:40Z dfigrish: gbyers: thanks for the ticket 1353 2016-07-21T17:47:00Z eschatologist: Did anyone preserve Macintosh Common Lisp CLIM? 2016-07-21T17:48:21Z rme: I've never had my hands on MCL's CLIM, myself. 2016-07-21T17:50:43Z eschatologist: Bummer. Might be one of those things lost to time. :( 2016-07-21T17:54:05Z rme: Unless it's on an old backup tape in the bottom of a filing cabinet in a disused lavatory behind a door that says "beware of the Leopard" somewhere at Apple, I am afraid that it may be lost to posterity. 2016-07-21T17:54:35Z eschatologist: Or unless one of the people who has it (Rainer, Rolf) can be talked into sharing it. 2016-07-21T17:55:00Z eschatologist: Excuse me, Ralf. 2016-07-21T17:55:03Z eschatologist quit (Quit: ZNC 1.6.3+deb1 - http://znc.in) 2016-07-21T17:55:39Z eschatologist joined #ccl 2016-07-21T17:56:04Z DougNYC joined #ccl 2016-07-21T17:57:50Z edgar-rft joined #ccl 2016-07-21T18:01:12Z DougNYC quit (Ping timeout: 260 seconds) 2016-07-21T18:11:15Z gbyers: For some reason, i think that one of the people who did the mcl clim port may have been bob kerns. I have no idea of how to get in touch with him, and may not remember his involvement correctly. 2016-07-21T18:21:07Z eschatologist: That wouldn't be surprising, since CLIM was an ILA thing and Bob was at ILA from what I've seen. 2016-07-21T18:30:09Z rme: LinkedIn has been urging me to spam Bob Kerns with a "friend request". https://www.linkedin.com/in/bobkerns 2016-07-21T18:34:01Z rme: I should start a side business as a private investigator, since I seem to be good at stalking people. 2016-07-21T18:44:38Z dfigrish: rme: :) 2016-07-21T18:44:39Z lispm joined #ccl 2016-07-21T18:46:26Z lispm: The MCL CLIM belongs to Alice Hartley, AFAIK 2016-07-21T18:53:02Z lispm quit (Quit: lispm) 2016-07-21T18:57:25Z rme: Wow. ESP. Or Google. 2016-07-21T19:08:08Z eschatologist: Has anyone heard from her lately? 2016-07-21T19:10:40Z pierpa quit (Ping timeout: 264 seconds) 2016-07-21T19:13:03Z dfigrish: rme: likelihood irc-logs 2016-07-21T19:44:21Z DougNYC joined #ccl 2016-07-21T19:49:24Z DougNYC quit (Ping timeout: 276 seconds) 2016-07-21T20:11:54Z rumbler31 quit (Ping timeout: 258 seconds) 2016-07-21T20:12:55Z rumbler31 joined #ccl 2016-07-21T20:14:11Z edgar-rft quit (Quit: edgar-rft) 2016-07-21T20:25:09Z dfigrish quit (Quit: rcirc on GNU Emacs 25.1.50.1) 2016-07-21T21:13:00Z edgar-rft joined #ccl 2016-07-21T21:24:57Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-21T21:29:04Z nzambe quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2016-07-21T21:29:15Z nzambe joined #ccl 2016-07-21T21:33:07Z DougNYC joined #ccl 2016-07-21T21:37:03Z DougNYC quit (Ping timeout: 240 seconds) 2016-07-21T21:59:13Z rumbler31 joined #ccl 2016-07-21T22:08:16Z DougNYC joined #ccl 2016-07-21T22:38:03Z DougNYC quit (Ping timeout: 240 seconds) 2016-07-22T00:37:46Z DougNYC joined #ccl 2016-07-22T00:56:49Z DougNYC quit 2016-07-22T01:07:38Z DougNYC joined #ccl 2016-07-22T01:15:47Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-22T01:57:08Z DougNYC quit 2016-07-22T03:03:57Z edgar-rft quit (Quit: edgar-rft) 2016-07-22T07:14:16Z dfigrish joined #ccl 2016-07-22T07:39:12Z rme quit (Quit: rme) 2016-07-22T07:39:12Z rme quit (Quit: rme) 2016-07-22T07:55:40Z hhdave joined #ccl 2016-07-22T08:12:33Z knobo1 quit (Ping timeout: 244 seconds) 2016-07-22T08:13:04Z clop quit (Ping timeout: 244 seconds) 2016-07-22T08:13:27Z clop joined #ccl 2016-07-22T08:13:52Z knobo1 joined #ccl 2016-07-22T09:12:09Z DavidGu quit (Remote host closed the connection) 2016-07-22T09:14:01Z Tennis joined #ccl 2016-07-22T09:39:33Z DavidGu joined #ccl 2016-07-22T12:17:57Z edgar-rft joined #ccl 2016-07-22T12:43:42Z rumbler31 joined #ccl 2016-07-22T13:08:19Z rme joined #ccl 2016-07-22T14:18:38Z |3b| quit (Remote host closed the connection) 2016-07-22T14:23:56Z rumbler31 quit (Ping timeout: 250 seconds) 2016-07-22T14:28:02Z StephanLahl quit (Remote host closed the connection) 2016-07-22T14:34:40Z StephanLahl joined #ccl 2016-07-22T14:46:15Z dfigrish quit (Ping timeout: 276 seconds) 2016-07-22T14:46:49Z mrottenkolber joined #ccl 2016-07-22T14:47:16Z mrottenkolber: Feature request: would it be possible to include a HTML rendered manual in the next release under doc/? 2016-07-22T14:58:51Z rme: That's a reasonable request. 2016-07-22T14:59:51Z rme: I have also been considering putting a version of ccldoc into the ccl repository. That would let users easily build the manual themselves, which might be better. 2016-07-22T15:00:56Z rme: Via. e.g., cd ccl/doc; make html-manual (or something like that). 2016-07-22T15:03:14Z phf: it would be handy to have ccldoc in the repo 2016-07-22T15:04:51Z rme: At one point I think I may have been advocating that it be kept separate, but if so, I've changed my mind. 2016-07-22T15:05:10Z rme gets to work on that 2016-07-22T15:24:05Z |3b| joined #ccl 2016-07-22T15:29:39Z edgar-rft quit (Ping timeout: 272 seconds) 2016-07-22T15:31:22Z dfigrish joined #ccl 2016-07-22T15:39:47Z dustinm` quit (*.net *.split) 2016-07-22T15:42:06Z dustinm` joined #ccl 2016-07-22T15:42:50Z edgar-rft joined #ccl 2016-07-22T16:28:00Z DougNYC joined #ccl 2016-07-22T16:36:11Z hhdave quit (Ping timeout: 240 seconds) 2016-07-22T16:46:12Z Tennis quit (Quit: Leaving) 2016-07-22T16:51:59Z dfigrish quit (Ping timeout: 258 seconds) 2016-07-22T16:54:18Z PuercoPop quit (Ping timeout: 250 seconds) 2016-07-22T16:55:10Z faheem quit (Ping timeout: 250 seconds) 2016-07-22T16:55:30Z faheem joined #ccl 2016-07-22T16:57:42Z PuercoPop joined #ccl 2016-07-22T18:11:45Z vydd joined #ccl 2016-07-22T19:05:03Z dfigrish joined #ccl 2016-07-22T19:39:32Z rumbler31 joined #ccl 2016-07-22T20:44:35Z rumbler31: hello #ccl. I've finally written something complicated enough that is spending an excessive amount of time in garbage collection, and I need help understanding how to reason about the system's behavior in more detail than what (time provides. I gather that the tools for this are likely implementation dependent, so I could use some pointers 2016-07-22T20:45:04Z rumbler31: basically I've implemented some of the pcl practical on reading binary files 2016-07-22T20:46:42Z rumbler31: right now i'm handling the "tagged types with dispatch" by hand, I have a small method structure laid out, and two tagged data chunks specified, both with many fields which come down to setf's on accessor functions of the slot for various elementary data types 2016-07-22T20:48:08Z rumbler31: in a function where I read every header, and then dispatch on type to either construct an object from a known message or simply collect the message data as a sequence on a 700 meg file where the messages are usually under 1000kb each, the function takes around 2000 seconds and around 90% of that time is spent in garbage collection. 2016-07-22T20:49:23Z rumbler31: I haven't had to really know what this means since before this it hasn't been a problem, but since I didn't write anything with an eye towards avoiding this problem I can't say that I'm surprised i've run into it, and I also haven't run into any literature that I've grokked to enough of a degree to recognize the patterns that put me in this situation 2016-07-22T20:52:03Z rumbler31: a binary class is made with a macro that reads in a list of slots to define and their types, and emits a reader that takes a stream and setf's each field based on repeated calls to read-byte for the various types 2016-07-22T20:52:14Z rumbler31: reader funciton* 2016-07-22T20:53:57Z rumbler31: since I figured that something I would do would make this activity slow for a complete read and digest of the file, I'd considered making the classes such that the relevant byte sequence would be first captured, and only individual slots actually parsed on first read, perhaps only written out to the backing byte sequence when a larger notional object write call would be made 2016-07-22T21:23:04Z |3b| 's only suggestion would be to store bulk data in specialized arrays if possible, since those are usually easier to GC than unspecialized arrays 2016-07-22T21:23:24Z |3b|: (and that's in general, not ccl specific) 2016-07-22T21:30:53Z rumbler31: do you mean where I specify things like :element-type 'unsigned-byte? it seems like the implementation is giving me simple vectors, which I guess is ok? although I'm keeping all of them, so they're not garbage, right? 2016-07-22T21:32:53Z |3b|: unsigned-byte isn't a fixed size, so probably upgrades to T, i mean things like single-float, (unsigned-byte 32), etc 2016-07-22T21:33:30Z |3b| isn't sure either of those is actually required to be supported, but i'd expect them on any implementation worth optimizing on :) 2016-07-22T21:34:51Z |3b|: UPGRADED-ARRAY-ELEMENT-TYPE will tell you if you will actually get a specialized array for a particular type 2016-07-22T21:36:15Z |3b|: but generally the GC has to look at every element of an unspecialized array, while it can ignore the contents of specialized arrays 2016-07-22T21:37:20Z rumbler31: well that's a simple enough change in a few places 2016-07-22T21:38:06Z rumbler31: this is an instance where I really don't understand what you're talking about, can you expound a little bit more? 2016-07-22T21:39:24Z |3b|: an unspecialized array can hold any value, including references to arbitrary objects. those might be the only reference, so the GC needs to check each element to see if it is a reference to an object on the heap 2016-07-22T21:40:01Z rumbler31: oh I see 2016-07-22T21:40:29Z |3b|: a specialized array generally holds values directly, and the GC knows a single-float or (unsigned-byte 8) can never be a reference to a heap object, and it knows that is the only thing in the specialized array, so it can just skip the contents 2016-07-22T21:42:00Z rumbler31: so one way I could reduce gc work is to (loosely) declare some garbage I expect to make as specific types that wouldn't hold references to objects on the heap (in appropriate places) 2016-07-22T21:42:44Z |3b| is more thinking about the live data than the garbage 2016-07-22T21:43:40Z |3b|: if the whole array is garbage, it might not need to look at the elements either way (depending on the implementation of the GC) 2016-07-22T21:45:06Z |3b|: and this is specifically limited to contents of specialized arrays, type declarations in general are less likely to affect GC 2016-07-22T21:45:24Z |3b|: (though if they let the compiler avoid creating heap objects for some values, it could still help) 2016-07-22T21:45:34Z rumbler31: ok 2016-07-22T21:45:36Z rumbler31: in hindsight that didn't make sense 2016-07-22T21:45:51Z rumbler31: if I consider something to be garbage, then I'm assuming the gc already knows its garbage, reducing its work here would be reworking things so that no garbage is produced 2016-07-22T21:46:55Z rumbler31: that makes sense in general, but discovering that this is the thing that needs to be done (or measuring how much such change impacts performance) and doing it are both alien to me 2016-07-22T21:47:53Z |3b|: reducing garbage might help as well, though it isn't always worth the effort 2016-07-22T21:48:22Z |3b|: finding the garbage is usually the expensive part, not collecting it 2016-07-22T21:50:54Z |3b|: so if you have a lot of complicated data structures, that could make looking for the garbage more expensive, then if you go back and modify old ones to avoid creating garbage, you might lose some optimizations where the GC would say "this page hasn't been modified since last time so i don't need to look at it for now" but it would have been able to collect fresh garbage quickly 2016-07-22T21:52:25Z |3b| isn't sure if i'm remembering that last part right, or if CCL does that particular thing, but generally point is that there are multiple factors to consider rather than just "amount of consing" 2016-07-22T21:52:57Z rumbler31: I should probably make a smaller test file, but the one i'm testing with right now takes 2000 seconds to run and every run is somewhere north of 90% of time in GC 2016-07-22T21:54:36Z |3b|: might also be some parameters for tuning the GC that would help, have to wait for someone that knows CCL specifics for that though :) 2016-07-22T22:32:11Z rumbler31: of course this is odd, I changed two declarations of vectors i was collecting from 'unsigned-byte to '(unsigned-byte 8) and re-ran. first re-run takes teh same amount of time, 2000 seconds and 90% gc, second run, no other changes just for fun, 200 seconds.... 2016-07-22T22:33:44Z rumbler31: rather, I recompiled the changed code, and also made a copy of the file, its currently in a veracypt volume, into an encrypted volume made with disk utility. I ran the recompiled code on the copy, 1800 seconds, I rerun the code again on the original, 200 seconds. 2016-07-22T22:35:05Z rumbler31: 3rd run, 200 seconds 2016-07-22T22:35:32Z rumbler31: 3rd run original file *200 seconds 2016-07-22T22:42:01Z rumbler31: well, 4th run on the copy, 200 seconds 2016-07-22T22:42:47Z rumbler31: when it would take forever, memory usage by the image would raise to my system limit. with the change it doesn't rise aboe 2 gig or so 2016-07-22T22:42:50Z rumbler31: all with a change from 'unsigned-byte to '(unsigned-byte 8) 2016-07-22T22:46:42Z rumbler31: I'm wondering how memory usage and tons of unspecified arrays are related. so far what you've told me is that i'd expect the time spent in GC to be very high due to having to check all the unspecified arrays for heap references 2016-07-22T22:47:24Z rumbler31: but why should it cause high memory usage? because i'm creating them faster than they can be "cleared" by the gc process and must be stored until such time? 2016-07-22T22:48:25Z rumbler31: but even then, i'm storing them myself, so my usage of the data is the same in both cases, I store the array 2016-07-22T22:48:33Z dfigrish quit (Ping timeout: 240 seconds) 2016-07-22T23:11:38Z fpt quit (Ping timeout: 250 seconds) 2016-07-22T23:22:07Z mrottenkolber quit (Ping timeout: 250 seconds) 2016-07-22T23:27:37Z gbyers: rumbler31: : what kind of machine are you using? 2016-07-22T23:30:07Z rumbler31: mbp, 2.7 ghz core i5, 8 gigs ram 2016-07-22T23:30:44Z rumbler31: over on lisp they say it makes sense that if I'm making 'unsigned-byte arrays that the values I store are all boxed, and hence memory usage would be high 2016-07-22T23:30:59Z rumbler31: i was thinking that memory usage and gc were related, 2016-07-22T23:37:56Z rumbler31 quit (Ping timeout: 258 seconds) 2016-07-22T23:38:35Z rumbler31 joined #ccl 2016-07-22T23:47:24Z rme quit (Quit: rme) 2016-07-22T23:47:24Z rme quit (Quit: rme) 2016-07-23T00:03:16Z rme joined #ccl 2016-07-23T00:54:59Z jdz quit (Ping timeout: 258 seconds) 2016-07-23T00:59:29Z jdz joined #ccl 2016-07-23T01:14:29Z DougNYC quit 2016-07-23T02:00:21Z DougNYC joined #ccl 2016-07-23T02:00:38Z rumbler31 quit 2016-07-23T02:07:27Z rumbler31 joined #ccl 2016-07-23T02:12:38Z rumbler31 quit (Remote host closed the connection) 2016-07-23T04:17:50Z vydd_ joined #ccl 2016-07-23T04:19:58Z vydd quit (Ping timeout: 252 seconds) 2016-07-23T04:52:34Z DougNYC quit 2016-07-23T05:10:27Z edgar-rft quit (Quit: edgar-rft) 2016-07-23T07:57:51Z rme quit (Quit: rme) 2016-07-23T08:32:02Z edgar-rft joined #ccl 2016-07-23T08:56:10Z rumbler31 joined #ccl 2016-07-23T09:14:31Z edgar-rft quit (Quit: edgar-rft) 2016-07-23T09:54:29Z dfigrish joined #ccl 2016-07-23T10:00:14Z brucem quit (Quit: ZNC - http://znc.sourceforge.net) 2016-07-23T10:58:18Z dfigrish quit (Ping timeout: 258 seconds) 2016-07-23T11:01:15Z brucem joined #ccl 2016-07-23T11:02:02Z brucem quit (Changing host) 2016-07-23T11:02:02Z brucem joined #ccl 2016-07-23T11:28:27Z mrottenkolber joined #ccl 2016-07-23T12:10:41Z vydd_ quit (Remote host closed the connection) 2016-07-23T12:26:50Z tfb joined #ccl 2016-07-23T12:31:30Z tfb quit (Client Quit) 2016-07-23T12:54:07Z rumbler3_ joined #ccl 2016-07-23T12:56:59Z rumbler31 quit (Ping timeout: 260 seconds) 2016-07-23T13:00:19Z rumbler3_ quit (Read error: Connection reset by peer) 2016-07-23T13:00:26Z rumbler31 joined #ccl 2016-07-23T13:07:57Z vydd joined #ccl 2016-07-23T13:43:59Z rme joined #ccl 2016-07-23T14:06:46Z rumbler31 quit (Remote host closed the connection) 2016-07-23T15:01:31Z rme quit (Ping timeout: 182 seconds) 2016-07-23T15:03:44Z rme quit (Ping timeout: 250 seconds) 2016-07-23T15:05:12Z rme joined #ccl 2016-07-23T15:12:32Z vydd quit (Ping timeout: 244 seconds) 2016-07-23T15:13:20Z vydd joined #ccl 2016-07-23T15:13:20Z vydd quit (Changing host) 2016-07-23T15:13:20Z vydd joined #ccl 2016-07-23T15:53:38Z edgar-rft joined #ccl 2016-07-23T15:54:55Z vydd quit (Remote host closed the connection) 2016-07-23T16:04:29Z rme quit (Ping timeout: 187 seconds) 2016-07-23T16:04:59Z rme quit (Ping timeout: 250 seconds) 2016-07-23T16:05:14Z vydd joined #ccl 2016-07-23T16:56:09Z DougNYC joined #ccl 2016-07-23T17:11:43Z vydd quit (Remote host closed the connection) 2016-07-23T17:30:19Z vydd joined #ccl 2016-07-23T18:09:33Z dfigrish joined #ccl 2016-07-23T20:19:26Z dfigrish: folks, is there known issues/bugs/gotchas with close-shared-library ? 2016-07-23T21:16:07Z mrottenkolber quit (Ping timeout: 250 seconds) 2016-07-23T21:16:58Z mrottenkolber joined #ccl 2016-07-23T22:23:08Z rumbler31 joined #ccl 2016-07-23T22:29:32Z hhdave joined #ccl 2016-07-23T22:31:39Z DougNYC quit 2016-07-23T22:33:09Z hhdave_ joined #ccl 2016-07-23T22:34:29Z hhdave quit (Ping timeout: 260 seconds) 2016-07-23T22:34:29Z hhdave_ is now known as hhdave 2016-07-23T22:38:11Z dfigrish quit (Quit: rcirc on GNU Emacs 25.1.50.1) 2016-07-23T22:58:34Z nzambe quit (Ping timeout: 240 seconds) 2016-07-23T23:09:27Z mrottenkolber left #ccl 2016-07-23T23:20:45Z hhdave quit (Quit: hhdave) 2016-07-24T01:36:15Z john-mcaleely quit (Remote host closed the connection) 2016-07-24T01:36:28Z john-mcaleely joined #ccl 2016-07-24T02:26:42Z rumbler31 quit (Ping timeout: 276 seconds) 2016-07-24T02:29:51Z rumbler31 joined #ccl 2016-07-24T04:10:57Z kini quit (Remote host closed the connection) 2016-07-24T04:34:20Z rumbler31 quit (Remote host closed the connection)