2016-08-30T00:15:54Z DeadTrickster quit (Ping timeout: 276 seconds)
2016-08-30T00:18:26Z oleo quit (Read error: Connection reset by peer)
2016-08-30T00:19:02Z oleo joined #sbcl
2016-08-30T00:19:02Z oleo quit (Changing host)
2016-08-30T00:19:02Z oleo joined #sbcl
2016-08-30T00:30:00Z karswell quit (Remote host closed the connection)
2016-08-30T00:30:19Z angavrilov_ joined #sbcl
2016-08-30T00:30:19Z angavrilov quit (Ping timeout: 265 seconds)
2016-08-30T00:30:48Z mason quit (Ping timeout: 265 seconds)
2016-08-30T00:31:19Z karswell joined #sbcl
2016-08-30T00:32:07Z mason joined #sbcl
2016-08-30T00:38:12Z trinque joined #sbcl
2016-08-30T00:39:14Z attila_lendvai quit (Ping timeout: 260 seconds)
2016-08-30T00:51:59Z DeadTrickster joined #sbcl
2016-08-30T00:57:30Z karswell quit (Ping timeout: 276 seconds)
2016-08-30T00:59:38Z karswell joined #sbcl
2016-08-30T01:05:53Z sjl quit (Read error: Connection reset by peer)
2016-08-30T01:16:48Z karswell quit (Ping timeout: 265 seconds)
2016-08-30T01:18:54Z karswell joined #sbcl
2016-08-30T01:41:37Z stassats quit (Ping timeout: 252 seconds)
2016-08-30T01:55:29Z em1l joined #sbcl
2016-08-30T01:59:09Z em1l_ quit (Ping timeout: 260 seconds)
2016-08-30T02:11:05Z DougNYC joined #sbcl
2016-08-30T03:03:16Z DougNYC quit
2016-08-30T03:20:30Z DavidGu1 joined #sbcl
2016-08-30T03:21:59Z DavidGu quit (Ping timeout: 260 seconds)
2016-08-30T03:21:59Z DavidGu1 is now known as DavidGu
2016-08-30T04:05:45Z cromachina quit (Read error: Connection reset by peer)
2016-08-30T04:06:09Z cromachina joined #sbcl
2016-08-30T04:32:06Z shka_ joined #sbcl
2016-08-30T04:33:22Z foom quit (Ping timeout: 250 seconds)
2016-08-30T04:41:38Z oleo quit (Quit: Leaving)
2016-08-30T04:45:23Z foom joined #sbcl
2016-08-30T05:26:17Z shka_ quit (Ping timeout: 244 seconds)
2016-08-30T05:35:18Z shka_ joined #sbcl
2016-08-30T05:41:12Z myrkraverk: I'm looking at
2016-08-30T05:41:14Z myrkraverk: http://sbcl.org/manual/sbcl.pdf
2016-08-30T05:41:34Z myrkraverk: Is \smallertt\phantomconcurrency:\llapsb-ext:hash-table-weakness really the name of the function?
2016-08-30T05:41:51Z myrkraverk: Or, if not, can the pdf be recompiled?
2016-08-30T05:42:07Z shka_ quit (Ping timeout: 258 seconds)
2016-08-30T05:42:15Z Bike: lol, tex is hard, eh?
2016-08-30T06:02:52Z scymtym quit (Ping timeout: 264 seconds)
2016-08-30T06:20:33Z ASau joined #sbcl
2016-08-30T06:37:55Z __SiCC__ left #sbcl
2016-08-30T06:51:03Z gingerale joined #sbcl
2016-08-30T07:21:40Z scymtym joined #sbcl
2016-08-30T07:45:21Z Bike quit (Quit: like a turntable made out of flesh)
2016-08-30T07:51:22Z scymtym: myrkraverk: see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=815205#95 . despite the final message in that discussion, rebuilding with texinfo 6.1.0.dfsg.18 still produces the borked pdf on my machine
2016-08-30T07:51:43Z gingerale quit (Remote host closed the connection)
2016-08-30T07:52:34Z scymtym: and the sbcl.pdf contained in the sbcl-doc 1.3.8-1 Debian package has the problem as well
2016-08-30T09:01:05Z scymtym: sorry for not noticing earlier, but we may have two failing tests since the most recent commit: https://ci.cor-lab.org/job/sbcl-master/1707/featureset=4,label=ubuntu_trusty_64bit/consoleFull (search for "unexpected-failure")
2016-08-30T09:01:25Z scymtym: i'm going to check whether this is just our ci setup or an actual regression
2016-08-30T09:17:06Z scymtym: happens locally as well
2016-08-30T09:26:07Z scymtym: dougk_: ^
2016-08-30T09:26:45Z scymtym: the PRINT-UNREADABLY-FUNCTION probably has to be changed to expect a BASE-STRING
2016-08-30T09:27:44Z scymtym: in the (print array *print-readably* :element-type) test, (make-array 0 :element-type nil) seems change element type to CHARACTER after the roundtrip
2016-08-30T09:40:56Z ASau quit (Ping timeout: 258 seconds)
2016-08-30T09:56:13Z jackdaniel:  /win 17
2016-08-30T10:25:29Z Xof joined #sbcl
2016-08-30T10:25:49Z stassats joined #sbcl
2016-08-30T10:53:29Z loke__ is now known as loke
2016-08-30T11:30:40Z eschatologist quit (Excess Flood)
2016-08-30T11:30:52Z eschatologist joined #sbcl
2016-08-30T12:05:26Z sjl joined #sbcl
2016-08-30T13:35:36Z cromachina quit (Read error: Connection reset by peer)
2016-08-30T13:37:22Z Xof: grmbl grmbl people don't run regression tests grmbl
2016-08-30T13:37:38Z stassats: referring to what?
2016-08-30T13:38:44Z Xof: the print-readably strings
2016-08-30T13:39:12Z stassats: but we have finally all strings readable
2016-08-30T13:39:22Z Xof: no we don't
2016-08-30T13:39:34Z Xof: (make-array 0 :element-type nil)
2016-08-30T13:39:50Z stassats: well, that's an outcast
2016-08-30T13:40:09Z Xof: right, but running the existing regression tests would have caught it :-/
2016-08-30T13:41:02Z stassats: but they weren't printable before, were they?
2016-08-30T13:41:30Z Xof: but they correctly signalled a print-not-readable error rather than silently doing the wrong thing
2016-08-30T13:41:43Z stassats: oh, i see
2016-08-30T13:42:06Z stassats: are working on a fix? otherwise i can try
2016-08-30T13:42:34Z Xof: I'm on it
2016-08-30T13:42:45Z stassats: ok, testing builds on windows then
2016-08-30T13:44:23Z Xof: thanks
2016-08-30T13:46:07Z nzambe joined #sbcl
2016-08-30T13:53:48Z stassats: success building on windows and ppc
2016-08-30T13:55:58Z Xof: good news
2016-08-30T13:57:58Z Xof: running tests on my fix, then (hopefully) commit, then release dance
2016-08-30T13:59:44Z stassats: arm64 is there too, which means i can build all the binaries i usually build
2016-08-30T13:59:54Z stassats: (when will i automate this step?)
2016-08-30T14:00:59Z Xof: remind me: do you normally build arm32?
2016-08-30T14:01:06Z Xof: if not, I can do one this month
2016-08-30T14:02:26Z stassats: i don't, too afraid my build host is too new
2016-08-30T14:02:46Z stassats: libc versions and whatnot
2016-08-30T14:02:55Z Xof: ok, I'll use my ancient RPi
2016-08-30T14:03:00Z Xof: "ancient"
2016-08-30T14:31:39Z oleo joined #sbcl
2016-08-30T15:26:33Z logrus joined #sbcl
2016-08-30T16:12:47Z Bike joined #sbcl
2016-08-30T16:21:07Z rudolfochrist joined #sbcl
2016-08-30T16:34:31Z karswell quit (Read error: Connection reset by peer)
2016-08-30T16:44:54Z gingerale joined #sbcl
2016-08-30T16:46:39Z rudolfochrist quit (Ping timeout: 265 seconds)
2016-08-30T16:48:33Z fiddlerwoaroof quit (Read error: Connection reset by peer)
2016-08-30T16:53:02Z fiddlerwoaroof joined #sbcl
2016-08-30T16:56:54Z karswell joined #sbcl
2016-08-30T17:00:53Z attila_lendvai joined #sbcl
2016-08-30T17:00:53Z attila_lendvai quit (Changing host)
2016-08-30T17:00:53Z attila_lendvai joined #sbcl
2016-08-30T17:05:39Z attila_lendvai quit (Ping timeout: 260 seconds)
2016-08-30T17:09:56Z shka_ joined #sbcl
2016-08-30T17:11:04Z chris_l joined #sbcl
2016-08-30T17:20:19Z attila_lendvai joined #sbcl
2016-08-30T17:20:19Z attila_lendvai quit (Changing host)
2016-08-30T17:20:19Z attila_lendvai joined #sbcl
2016-08-30T17:26:20Z rpg joined #sbcl
2016-08-30T17:29:57Z slyrus joined #sbcl
2016-08-30T17:32:43Z sjl quit (Ping timeout: 244 seconds)
2016-08-30T17:41:34Z rudolfochrist joined #sbcl
2016-08-30T18:00:12Z rudolfochrist quit (Ping timeout: 265 seconds)
2016-08-30T18:00:16Z rpg quit (Ping timeout: 250 seconds)
2016-08-30T18:04:26Z sjl joined #sbcl
2016-08-30T18:38:36Z scymtym: stassats: do you think paste.lisp.org/display/324382 is a good idea?
2016-08-30T18:39:12Z stassats: that would require some actual thinking
2016-08-30T18:44:55Z scymtym: i just found runtime/arm64-assem.S:376
2016-08-30T18:45:07Z stassats: scymtym: (logior (ldb (byte 7 0) v) (if (zerop v-next) 0 (ash 1 7))) => (dpb (if (zerop v-next) 0 1) (byte 1 7) v)
2016-08-30T18:45:11Z scymtym: things like that make this change a bit risky
2016-08-30T18:45:24Z stassats: yeah, all the assem.S will have to be changed
2016-08-30T18:46:06Z scymtym: the code would need some improvements, of course. i'm just toying with the idea for now
2016-08-30T18:46:20Z stassats: looks reasonable
2016-08-30T18:46:24Z scymtym: it saves about ~ 30 Kb for the core
2016-08-30T18:46:37Z scymtym: so not too much up front
2016-08-30T18:46:58Z scymtym: but if we can make some of the number smaller, there could be further gains
2016-08-30T18:47:03Z scymtym: *numbers
2016-08-30T18:48:00Z scymtym: ok, i will try to prepare an actual patch, then
2016-08-30T18:48:45Z stassats: and arm64-assem is an easy case
2016-08-30T18:49:28Z stassats: i can help rounding up all the assem.S
2016-08-30T18:49:37Z minion quit (Disconnected by services)
2016-08-30T18:49:41Z minion joined #sbcl
2016-08-30T18:49:45Z scymtym: cool, thanks
2016-08-30T18:49:58Z scymtym: i don't have time right now, though
2016-08-30T18:50:11Z DGASAU` joined #sbcl
2016-08-30T18:51:29Z carvite quit (Ping timeout: 264 seconds)
2016-08-30T18:52:04Z akkad quit (Ping timeout: 264 seconds)
2016-08-30T18:52:04Z DGASAU quit (Ping timeout: 264 seconds)
2016-08-30T18:52:52Z DGASAU` is now known as DGASAU
2016-08-30T18:53:27Z carvite joined #sbcl
2016-08-30T19:03:27Z akkad joined #sbcl
2016-08-30T19:04:43Z logrus quit (Quit: Leaving)
2016-08-30T19:30:31Z chris_l quit (Quit: Ex-Chat)
2016-08-30T19:56:34Z scymtym quit (Ping timeout: 258 seconds)
2016-08-30T19:59:58Z slyrus quit (Ping timeout: 252 seconds)
2016-08-30T20:00:45Z igajsin joined #sbcl
2016-08-30T20:00:48Z igajsin quit (Client Quit)
2016-08-30T20:10:44Z rudolfochrist joined #sbcl
2016-08-30T20:12:26Z rpg joined #sbcl
2016-08-30T20:15:27Z rudolfochrist quit (Ping timeout: 244 seconds)
2016-08-30T20:33:59Z scymtym joined #sbcl
2016-08-30T20:41:29Z slyrus joined #sbcl
2016-08-30T20:49:39Z attila_lendvai quit (Ping timeout: 260 seconds)
2016-08-30T20:54:21Z gingerale quit (Remote host closed the connection)
2016-08-30T20:58:03Z slyrus quit (Ping timeout: 276 seconds)
2016-08-30T21:23:16Z scymtym: the most common sc-offsets seem to fit in 7 bits: http://paste.lisp.org/display/324382#3 . so, rearranging sc-offset coding (in MAKE-SC-OFFSET) a bit could make the vast majority of error-break data fit in a single byte
2016-08-30T21:24:08Z stassats: can we automate this?
2016-08-30T21:24:28Z scymtym: determining the frequencies?
2016-08-30T21:24:39Z stassats: yes, and rearranging
2016-08-30T21:24:52Z scymtym: not sure
2016-08-30T21:25:12Z scymtym: there is at least MAKE-SC-OFFSET with corresponding deconding and some code in the runtime
2016-08-30T21:25:32Z scymtym: maybe we can describe the coding and make the coding data driven
2016-08-30T21:25:44Z scymtym: it is not on any critical path, right?
2016-08-30T21:26:46Z stassats: but how do you want to rearrange make-sc-offset?
2016-08-30T21:52:15Z scymtym: something like ...ssssoooooss where s are sc-number bits and o are offset bits. in combination with mapping the current sc numbers 27, 26 and 0 (or similar) to the range [0,3] that would allow most error data to fit in 7 bit, requiring one byte for encoding
2016-08-30T21:53:10Z stassats: ok
2016-08-30T21:53:25Z scymtym: there would be more offset bits, but that's the idea anyway
2016-08-30T22:11:52Z slyrus joined #sbcl
2016-08-30T22:15:08Z rpg quit (Quit: Textual IRC Client: www.textualapp.com)
2016-08-30T22:17:32Z DeadTrickster quit (Ping timeout: 240 seconds)
2016-08-30T22:40:29Z shka_ quit (Ping timeout: 244 seconds)
2016-08-30T22:42:39Z stassats: tried removing all the nfp stuff on x86oids, doesn't make any improvement
2016-08-30T22:42:40Z stassats: oh well
2016-08-30T22:57:49Z DeadTrickster joined #sbcl
2016-08-30T23:08:37Z DeadTrickster quit (Ping timeout: 252 seconds)
2016-08-30T23:38:32Z cromachina joined #sbcl
2016-08-30T23:49:54Z DeadTrickster joined #sbcl
2016-08-30T23:53:56Z fisxoj joined #sbcl