2015-02-23T00:01:27Z ASau joined #sbcl
2015-02-23T00:26:45Z stassats quit (Ping timeout: 250 seconds)
2015-02-23T00:27:08Z gingerale quit (Ping timeout: 252 seconds)
2015-02-23T00:28:12Z ASau quit (Ping timeout: 245 seconds)
2015-02-23T00:48:53Z ASau joined #sbcl
2015-02-23T00:50:52Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-02-23T00:57:18Z pkhuong: ... I find myself using this pattern in a DSL compiler: (defstruct (foo (:constructor make-foo (bar baz))) bar baz) (defun destructure-foo (foo) (values (foo-bar foo) (foo-baz foo)))
2015-02-23T00:58:19Z pkhuong: with many purely functional IRs, there's a nice symmetry, and m-v-call helps.
2015-02-23T02:12:10Z edgar-rft quit (Quit: bye...)
2015-02-23T02:55:08Z LiamH quit (Read error: Connection reset by peer)
2015-02-23T03:30:54Z scymtym quit (Ping timeout: 246 seconds)
2015-02-23T03:39:02Z christoph_debian quit (Ping timeout: 245 seconds)
2015-02-23T03:52:23Z christoph_debian joined #sbcl
2015-02-23T04:59:52Z Quadrescence joined #sbcl
2015-02-23T05:00:47Z nyef quit (Quit: G'night all)
2015-02-23T05:27:57Z csziacobus quit (Quit: csziacobus)
2015-02-23T05:34:40Z pranavrc joined #sbcl
2015-02-23T05:36:58Z psy_ quit (Quit: Leaving)
2015-02-23T07:18:39Z psy_ joined #sbcl
2015-02-23T07:40:50Z pranavrc quit
2015-02-23T07:41:18Z pranavrc joined #sbcl
2015-02-23T07:41:19Z pranavrc quit (Changing host)
2015-02-23T07:41:19Z pranavrc joined #sbcl
2015-02-23T07:51:23Z gingerale joined #sbcl
2015-02-23T07:52:32Z karswell quit (Read error: Connection reset by peer)
2015-02-23T07:53:10Z karswell joined #sbcl
2015-02-23T07:59:46Z echo-area joined #sbcl
2015-02-23T08:04:47Z pranavrc_ joined #sbcl
2015-02-23T08:06:38Z pranavrc quit (Ping timeout: 246 seconds)
2015-02-23T08:15:54Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-02-23T08:29:51Z scymtym joined #sbcl
2015-02-23T08:34:31Z Shinmera joined #sbcl
2015-02-23T08:48:46Z Shinmera quit (Quit: しつれいしなければならないんです。)
2015-02-23T08:51:42Z Shinmera joined #sbcl
2015-02-23T08:57:37Z rjmacready joined #sbcl
2015-02-23T09:00:34Z loke: When you get the FAILED AVER error when two threads call a generic function at the same time, how can I tell which method it failed on?
2015-02-23T09:01:35Z eschatologist joined #sbcl
2015-02-23T09:01:37Z eschatologist quit (Client Quit)
2015-02-23T09:02:40Z Shinmera quit (Remote host closed the connection)
2015-02-23T09:07:31Z Shinmera joined #sbcl
2015-02-23T09:20:49Z ASau quit (Ping timeout: 255 seconds)
2015-02-23T09:25:25Z angavrilov joined #sbcl
2015-02-23T09:31:23Z ASau joined #sbcl
2015-02-23T09:44:59Z chris2 joined #sbcl
2015-02-23T09:45:14Z chris2: hi. i have a small request: simple-local-client in sb-bsd-sockets/tests.lisp
2015-02-23T09:45:26Z chris2: sends a syslog message with level "kern.debug"
2015-02-23T09:45:54Z chris2: could this be changed to have level "user.debug" perhaps?
2015-02-23T09:46:15Z chris2: that would be <15> i think
2015-02-23T09:48:14Z Shinmera- joined #sbcl
2015-02-23T09:49:02Z Shinmera quit (Ping timeout: 265 seconds)
2015-02-23T09:49:35Z pranavrc_ is now known as pranavrc
2015-02-23T09:51:50Z Shinmera- is now known as Shinmera
2015-02-23T10:00:40Z ASau quit (Remote host closed the connection)
2015-02-23T10:02:38Z krl left #sbcl
2015-02-23T10:03:23Z rjmacready quit (Quit: Page closed)
2015-02-23T10:14:18Z ASau joined #sbcl
2015-02-23T10:18:09Z attila_lendvai joined #sbcl
2015-02-23T10:18:30Z attila_lendvai quit (Changing host)
2015-02-23T10:18:30Z attila_lendvai joined #sbcl
2015-02-23T10:18:54Z selat joined #sbcl
2015-02-23T10:55:04Z Shinmera quit (Read error: Connection reset by peer)
2015-02-23T11:16:55Z psy_ quit (Ping timeout: 255 seconds)
2015-02-23T11:22:10Z psy_ joined #sbcl
2015-02-23T11:22:34Z psy_ quit (Max SendQ exceeded)
2015-02-23T11:23:33Z psy_ joined #sbcl
2015-02-23T11:33:18Z Shinmera joined #sbcl
2015-02-23T11:46:23Z stassats joined #sbcl
2015-02-23T11:49:57Z pranavrc quit (Read error: No route to host)
2015-02-23T12:05:03Z edgar-rft joined #sbcl
2015-02-23T12:32:18Z Cymew joined #sbcl
2015-02-23T12:35:35Z oleo joined #sbcl
2015-02-23T12:37:47Z oleo__ quit (Ping timeout: 245 seconds)
2015-02-23T13:03:20Z Shinmera quit (Ping timeout: 265 seconds)
2015-02-23T13:09:16Z Shinmera joined #sbcl
2015-02-23T13:15:48Z psy_ quit (Quit: Leaving)
2015-02-23T13:38:16Z attila_lendvai quit (Ping timeout: 264 seconds)
2015-02-23T14:04:32Z minion quit (Remote host closed the connection)
2015-02-23T14:04:32Z specbot quit (Remote host closed the connection)
2015-02-23T14:07:12Z minion joined #sbcl
2015-02-23T14:07:51Z specbot joined #sbcl
2015-02-23T14:42:12Z eudoxia joined #sbcl
2015-02-23T14:53:00Z nyef joined #sbcl
2015-02-23T14:53:12Z nyef: G'morning all.
2015-02-23T14:56:22Z psy_ joined #sbcl
2015-02-23T15:01:45Z Shinmera quit (Quit: brb)
2015-02-23T15:04:08Z Shinmera joined #sbcl
2015-02-23T15:09:05Z pkhuong quit (Ping timeout: 246 seconds)
2015-02-23T15:12:23Z rpg joined #sbcl
2015-02-23T15:15:25Z psy_ quit (Ping timeout: 252 seconds)
2015-02-23T15:22:09Z psy_ joined #sbcl
2015-02-23T15:37:43Z eudoxia quit (Quit: Leaving)
2015-02-23T15:55:00Z selat quit (Quit: Lost terminal)
2015-02-23T16:18:26Z stassats: would a return result protocol enforcing for standard GFs be welcome?
2015-02-23T16:20:41Z stassats: probably not now
2015-02-23T16:25:51Z stassats: and i'll expect a lot of code breaking from this, which is good, ought to keep those pesky standard violating people at their toes
2015-02-23T16:34:49Z flip214: stassats: while you're at GFs, how about providing (IIRC) Zachs inlined GFs, too?
2015-02-23T16:37:09Z stassats: there's no point in inlined gfs
2015-02-23T16:40:26Z flip214: stassats: performance is much better - much, much better, if the type can be derived at compile time.
2015-02-23T16:41:09Z stassats: inlining gfs doesn't help you here
2015-02-23T16:41:53Z Cymew quit (Quit: Konversation terminated!)
2015-02-23T16:42:22Z Shinmera quit (Quit: BBL)
2015-02-23T16:42:45Z flip214: why do you say that? I replaced a GF with a `(if (typep ...)) and got a real boost out of that.
2015-02-23T16:43:05Z stassats: because you replaced a gf with `(if (typep ...))
2015-02-23T16:43:12Z stassats: not because you inlined it
2015-02-23T16:43:29Z stassats: generic functions are not `(if (typep ...))
2015-02-23T16:45:25Z flip214: http://discontinuity.info/~pkhuong/gf-sealing.lisp
2015-02-23T16:45:45Z flip214: does something similar.
2015-02-23T16:47:15Z stassats: which requires sealing everything
2015-02-23T16:47:43Z stassats: no new classes can be defined or modified, no new methods can be added
2015-02-23T16:48:01Z stassats: might as well just use structures and typecase
2015-02-23T16:57:13Z Xof: there have been lazily-recomputed inline/sealing solutions before
2015-02-23T16:57:17Z Xof: I think Gerd made one for CMUCL
2015-02-23T17:00:04Z flip214: then a running system just needs to be asked about the number of active objects and method call counts to calculate the "optimal" type tree bisection ;)
2015-02-23T17:04:14Z Xof: no, each call site can have its own dispatch
2015-02-23T17:19:44Z attila_lendvai joined #sbcl
2015-02-23T17:38:37Z eudoxia joined #sbcl
2015-02-23T17:41:24Z pkhuong joined #sbcl
2015-02-23T18:28:46Z Shinmera joined #sbcl
2015-02-23T18:43:16Z krzysz00 quit (Ping timeout: 244 seconds)
2015-02-23T19:35:48Z krzysz00 joined #sbcl
2015-02-23T19:42:37Z krzysz00 quit (Ping timeout: 250 seconds)
2015-02-23T19:44:24Z krzysz00 joined #sbcl
2015-02-23T19:49:42Z Bicyclidine joined #sbcl
2015-02-23T20:15:55Z leuler joined #sbcl
2015-02-23T20:18:23Z leuler: On x86/Linux, with sbcl-1.2.8-81-gd79cc3e, I get a reproducible test failure in step.impure.lisp / STEP-NEXT/2 with an "Unhandled memory fault at #x0." Is this known?
2015-02-23T20:21:47Z stassats: no
2015-02-23T20:22:41Z leuler: Hmm.
2015-02-23T20:29:44Z rpg quit (Quit: rpg)
2015-02-23T20:30:23Z leuler: Also, threads.impure.lisp / FP-MODE-INHERITANCE-THREADS fails, but only when running it from a file containing only this test: "The assertion (= FP-MODE THREAD-FP-MODE) failed with FP-MODE = 1073742413, THREAD-FP-MODE = 589".
2015-02-23T20:31:06Z leuler: The first is: 01000000000000000000001001001101
2015-02-23T20:31:29Z leuler: The second:   00000000000000000000001001001101
2015-02-23T20:31:59Z stassats` joined #sbcl
2015-02-23T20:32:08Z stassats`: leuler: that test never succeeded
2015-02-23T20:32:37Z leuler: It did for me when running the complete file, earlier today.
2015-02-23T20:36:45Z stassats`: and step-next/2 is not reproducible
2015-02-23T20:44:45Z leuler: Well, then thanks for trying.
2015-02-23T20:48:12Z mega1 quit (Ping timeout: 245 seconds)
2015-02-23T20:51:07Z selat joined #sbcl
2015-02-23T20:59:38Z White_Flame quit (Remote host closed the connection)
2015-02-23T21:05:28Z angavrilov quit (Remote host closed the connection)
2015-02-23T21:20:39Z leuler: I guess the fp-mode test should ignore the condition codes in the fp status word.
2015-02-23T21:28:05Z eudoxia quit (Quit: Leaving)
2015-02-23T21:32:25Z selat quit (Quit: Lost terminal)
2015-02-23T21:38:42Z Shinmera quit (Quit: しつれいしなければならないんです。)
2015-02-23T22:51:46Z leuler: stassats`: I'd like to change the fp-mode test to ignore the complete fp status word, thus comparing only the fp control word. That makes it succeed here. You originally wrote that test -- would this change be OK with you?
2015-02-23T22:52:34Z Bicyclidine quit (Ping timeout: 272 seconds)
2015-02-23T23:07:25Z Bicyclidine joined #sbcl
2015-02-23T23:17:23Z gingerale quit (Ping timeout: 240 seconds)
2015-02-23T23:26:25Z psy_ quit (Ping timeout: 256 seconds)
2015-02-23T23:26:48Z attila_lendvai quit (Quit: Leaving.)
2015-02-23T23:46:01Z stassats`: not really, the status part has accrued exceptions, doesn't it?
2015-02-23T23:47:38Z ASau quit (Remote host closed the connection)
2015-02-23T23:49:02Z attila_lendvai joined #sbcl
2015-02-23T23:50:09Z leuler: Yes it has, but the test as you wrote it explicitly ignores these.
2015-02-23T23:51:45Z leuler: In the meantime, I have looked through that a bit more. I guess the test should just compare the set of traps that are enabled. This would be possible portably using (sb-vm::get-floating-point-modes).
2015-02-23T23:52:46Z leuler: Operating directly with sb-vm:floating-point-modes seems to be more difficult, portably.
2015-02-23T23:53:13Z stassats`: i don't care about the test, i need floating point modes operating correctly
2015-02-23T23:53:19Z stassats`: are they or are they not?
2015-02-23T23:54:44Z stassats`: i don't have enough of an answer for that at the moment
2015-02-23T23:55:12Z leuler: I think they are. Also I think that the test fails despite that because it compares parts of the modes (e.g., condition codes) that vary unpredictably.
2015-02-23T23:55:24Z Bicyclidine quit (Quit: quit)
2015-02-23T23:57:23Z stassats`: the idea is to make them predictable
2015-02-23T23:59:24Z attila_lendvai quit (Quit: Leaving.)