| 01:28:28 | fowl_ | ok its been running for a while with no output so im considering it hung at this point, https://gist.github.com/3737049 |
| 01:35:12 | fowl_ | err more showed up after i interrupted it |
| 01:35:31 | fowl_ | i'll leave it running while i sleep tonight |
| 04:44:19 | reactormonk | dom96: hmm |
| 05:55:14 | reactormonk | Araq: http://www.cs.hmc.edu/~geoff/mtwist.html yet again |
| 05:58:21 | * | Trix[a]r_za is now known as Trixar_za |
| 07:43:14 | * | CodeBlock quit (*.net *.split) |
| 07:56:19 | * | CodeBlock joined #nimrod |
| 08:33:23 | * | Araq_ joined #nimrod |
| 08:37:55 | * | Boscop joined #nimrod |
| 09:29:08 | * | Trixar_za is now known as Trix[a]r_za |
| 09:45:25 | * | Araq_ quit (Read error: Connection timed out) |
| 09:46:38 | * | Araq_ joined #nimrod |
| 09:51:46 | * | Araq_ quit (Quit: ChatZilla 0.9.88.2 [Firefox 14.0.1/20120713134347]) |
| 11:03:07 | * | Boscop quit (Ping timeout: 240 seconds) |
| 11:46:02 | * | Boscop joined #nimrod |
| 12:12:48 | * | Boscop quit (Ping timeout: 244 seconds) |
| 12:20:55 | * | Boscop joined #nimrod |
| 12:43:08 | * | ccssnet quit (Read error: Connection reset by peer) |
| 12:46:02 | * | fowl joined #nimrod |
| 12:46:10 | * | Boscop quit (*.net *.split) |
| 12:46:10 | * | fowl_ quit (*.net *.split) |
| 12:46:10 | * | Reisen quit (*.net *.split) |
| 12:46:17 | * | Boscop joined #nimrod |
| 12:50:27 | * | zahary joined #nimrod |
| 13:00:27 | * | Reisen joined #nimrod |
| 13:06:39 | fowl | it was hanging at that point |
| 13:37:28 | * | q66 joined #nimrod |
| 14:02:33 | * | ccssnet joined #nimrod |
| 14:09:41 | * | ccssnet quit (Quit: http://atccss.net) |
| 14:27:07 | * | Boscop quit (Ping timeout: 252 seconds) |
| 14:31:38 | * | Boscop joined #nimrod |
| 15:18:00 | dom96 | reactormonk: ...hmm...? |
| 16:13:37 | * | Boscop quit (Ping timeout: 240 seconds) |
| 16:35:30 | * | fowl quit (Read error: Connection reset by peer) |
| 16:39:17 | * | shevy quit (Ping timeout: 256 seconds) |
| 16:52:01 | * | shevy joined #nimrod |
| 17:09:09 | dom96 | I'm starting to consider making the sockets module track whether a socket is blocking. |
| 17:09:51 | dom96 | or rather I am considering again. I think i've thought about this before. |
| 17:13:40 | dom96 | Araq: Is the overhead of a TSocket really bad even when it's unbuffered? |
| 17:14:06 | dom96 | I suppose the fact it's an object creates a lot of overhead? |
| 17:16:28 | Araq | it's a quite big object with its own buffer ... |
| 17:17:08 | Araq | but meh, it's still good enough |
| 17:17:31 | Araq | people hate that we use 'select' though |
| 17:17:41 | reactormonk | why so? |
| 17:17:41 | dom96 | Who does? |
| 17:18:13 | dom96 | Araq: How do object variants work? Even when 'isBuffered' is false, will the overhead of the buffer still be there? |
| 17:18:41 | Araq | dom96: yes, it will be there |
| 17:19:16 | Araq | but I have an 'unsafeNew' proc planned already |
| 17:20:00 | Araq | brb |
| 17:20:09 | dom96 | Well then I could either provide a TSocket and a PBufferedSocket, or you could just optimise this in the language somehow :P |
| 17:54:32 | Araq | well add the single bool to TSocket, it doesn't matter |
| 17:54:59 | Araq | but a TSocket with no abstraction and a PBufferedSocket sound good ... |
| 17:55:13 | dom96 | That would break the API yet again |
| 17:55:16 | Araq | however I think you said, it can't be done conveniently? |
| 17:55:20 | dom96 | mmm |
| 17:56:14 | Araq | I don't mind breaking the API again as it's incompatible with 0.8.14 already |
| 17:56:25 | Araq | and we had no release in the meantime |
| 17:56:36 | dom96 | Yeah, but that means more work for me :P |
| 17:56:48 | dom96 | And, it is compatible with 0.8.14 I think ;) |
| 17:56:57 | Araq | unfortunately getting this API right is quite important |
| 17:57:03 | Araq | as lots of modules depend on it |
| 17:57:28 | Araq | reactormonk: there are os specific alternatives to 'select' which are much more performant |
| 18:02:42 | reactormonk | Araq: all the notify stuff? |
| 18:03:20 | * | Boscop joined #nimrod |
| 18:03:25 | Araq | yeah |
| 18:07:04 | * | Boscop quit (Disconnected by services) |
| 18:07:06 | * | Boscop joined #nimrod |
| 18:48:18 | * | Boscop quit (Ping timeout: 252 seconds) |
| 18:59:06 | Araq | dom96: take a look at nimbuild please |
| 18:59:15 | Araq | no test results :-( |
| 18:59:32 | dom96 | 0_o |
| 18:59:40 | dom96 | tester must be frozen |
| 19:00:20 | Araq | maybe some test triggered an endless loop in the compiler |
| 19:00:42 | Araq | we should make tester.nim timeout |
| 19:01:55 | dom96 | tthreadheapviolation1.nim |
| 19:02:01 | dom96 | Seems to be the culprit |
| 19:02:05 | dom96 | Could you check locally? |
| 19:02:28 | Araq | later, I'm hunting some other bug |
| 19:56:57 | shevy | yeah |
| 19:57:01 | shevy | with bow and arrow |
| 19:57:04 | shevy | shoot that bug down! |
| 20:03:02 | Araq | it's hard :-/ |
| 20:50:31 | * | fowl joined #nimrod |
| 21:13:11 | reactormonk | Araq: whats the bug exactly? |
| 21:13:42 | Araq | some weird interaction of multiply instantiated generics using gensym'ed symbols |
| 21:14:11 | reactormonk | huh? |
| 21:14:22 | Araq | resulting in wrong C code but the types appear to be fine in the frontend... |
| 21:15:33 | reactormonk | so the types are wrong in the C code? |
| 21:15:39 | Araq | yes |
| 21:15:57 | reactormonk | (trace) ? ;-) |
| 21:36:17 | reactormonk | is it possible in a macro to tell wherever something is a literal/computed from literals or not? |
| 22:08:28 | * | Trix[a]r_za is now known as Trixar_za |
| 22:08:59 | Araq | x: expr{lit} |
| 22:09:28 | Araq | reactormonk: but currently only for tr macros |
| 22:11:12 | reactormonk | Araq: would be nice for tests, based on wherever stuff is literal or not, different explanation |
| 22:15:40 | Araq | well you can easily test for nnkIntLit etc within a macro |
| 22:16:41 | reactormonk | but what of check-within x, 2..3 ? |
| 22:19:03 | Araq | 'in'? |
| 22:19:22 | reactormonk | ehm, right. |
| 22:19:39 | reactormonk | My point was wherever the literal property is inherited or not |
| 22:20:36 | Araq | well I have to sleep now; I'm sure what you want is easy to accomplish ;-) |
| 22:20:45 | Araq | good night |
| 22:21:15 | reactormonk | but it's only 1700 ^^ |
| 23:10:42 | * | q66 quit (Quit: Quit) |
| 23:54:40 | * | zahary1 joined #nimrod |
| 23:55:05 | * | zahary quit (Ping timeout: 252 seconds) |