| 02:37:24 | * | Trixar_za is now known as Trix[a]r_za |
| 02:45:33 | * | fowl quit (Ping timeout: 245 seconds) |
| 02:51:26 | * | fowl joined #nimrod |
| 03:33:53 | * | fowl quit (Read error: Connection reset by peer) |
| 03:35:45 | * | fowl joined #nimrod |
| 07:39:07 | * | xcombelle joined #nimrod |
| 09:51:26 | * | Boscop quit (Quit: Boscop) |
| 11:28:25 | * | Trix[a]r_za is now known as Trixar_za |
| 12:12:20 | * | q66 joined #nimrod |
| 12:25:50 | * | fowl_ joined #nimrod |
| 12:27:35 | * | fowl quit (Ping timeout: 252 seconds) |
| 12:59:47 | * | fowl_ quit (Ping timeout: 252 seconds) |
| 13:33:30 | * | Trixar_za is now known as Trix[a]r_za |
| 14:17:01 | * | q66 quit (*.net *.split) |
| 14:18:54 | * | XAMPP quit (Read error: Connection reset by peer) |
| 14:19:02 | * | q66 joined #nimrod |
| 14:34:24 | zahary_ | heh, what are these bugs that get fixed on their own? what has changed when it comes to circular imports and pre-sem expansion of templates in generics? |
| 14:53:19 | Araq | circular imports have been fixed by your changes to the C backend I think where simply every BModule is kept open |
| 14:54:06 | Araq | I couldn't reproduce non-immediate expansion of immediate templates in generics |
| 14:54:18 | Araq | so I closed it |
| 14:55:08 | Araq | I wanted to add circular module deps to the test suite but then found it it's already part of it, so I closed it finally |
| 14:57:20 | zahary_ | well, a disappearing ICE doesn't sound that strange indeed, but we still plan to introduce error messages for the circular imports, disabled by the "layers" feature |
| 14:57:40 | reactormonk | zahary_, why so? |
| 14:59:29 | zahary_ | well, currently if you happen to produce a circular import somewhere, you won't get an ICE, but a very peculiar "foo undefined" errors |
| 14:59:53 | reactormonk | uh, ok |
| 15:02:00 | Araq | zahary_: not if you do it properly ;-) |
| 15:03:03 | zahary_ | sure, that too, but I agreed to your proposal for limiting somehow circular imports, remember :) |
| 15:03:18 | Araq | in fact, the shabby behaviour is part of the spec for now :-) |
| 15:04:49 | zahary_ | as for the other issue, the problem with the prepass for generics was something about variable injection from templates. I have to dig up the logs |
| 15:05:19 | Araq | that problem still exists but it's not part of the bug report |
| 15:11:19 | Araq | can you please have a look what keeps aporia from compiling on mac, zahary_ ? it's kind of a show stopper bug for the release of 0.9.2 |
| 15:11:36 | zahary_ | alright, haven't tried it yet |
| 15:12:06 | zahary_ | fails to compile? no run-time crash? |
| 15:12:19 | Araq | yeah |
| 15:12:31 | Araq | some nkYield is not eliminated for whatever reasons |
| 15:13:33 | Araq | my educated guess is that we have some codegen error for bitsets which is triggered by clang but not by gcc |
| 15:14:40 | Araq | or maybe some int literals get truncated by clang |
| 15:15:31 | zahary_ | I'll try to diff the output by a linux build with os:macosx and a native build and see how that goes |
| 15:15:55 | Araq | interesting idea |
| 15:21:33 | Araq | I used to a have script for that that gets rid of the IDs in the C code |
| 15:28:43 | NimBot | Araq/Nimrod de9d715 Araq [+0 ±2 -0]: fixes #283 |
| 15:28:43 | NimBot | Araq/Nimrod c669ed1 Araq [+0 ±5 -2]: cleanup of tests; use 'koch install' instead of 'install.sh' for a bootstrapping build |
| 16:17:40 | reactormonk | Araq, why not delete install.sh? |
| 16:17:52 | reactormonk | Araq, nd you still have a bunch of references to 'install.sh' |
| 16:18:11 | Araq | reactormonk: I did delete install.sh ... o.o |
| 16:19:06 | Araq | and the references to install.sh are correct for the "ordinary" installation process |
| 16:19:17 | reactormonk | huhm |
| 16:19:23 | Araq | it changed for the bootstrapping installation |
| 16:19:52 | reactormonk | oh, it's only in the gitignore :-) |
| 16:20:54 | dom96 | we still need to implement some extra features for the install.sh script |
| 16:25:04 | Araq | I thought we already did, dom96 |
| 16:25:30 | dom96 | did we? |
| 16:25:45 | dom96 | i'll check it out in a little while. |
| 16:26:40 | Araq | ok |
| 16:56:09 | dom96 | Araq: I replied to #372 and #291 |
| 17:09:46 | Araq | damn you |
| 17:10:01 | Araq | it's very easy to improve the error message so now I have to do that ... |
| 17:10:30 | dom96 | :P |
| 17:13:01 | Araq | I'd like a better solution though; github's bug tracker sucks |
| 17:13:19 | Araq | it's now full of feature requests and minor issues :P |
| 17:13:31 | Araq | making it hard to find & fix the real bugs |
| 17:24:24 | NimBot | Araq/Nimrod 710b570 Araq [+0 ±1 -0]: improved error message for #291 |
| 17:25:52 | dom96 | why not use the tagging system? |
| 17:28:12 | Araq | too many bugs are not tagged yet :P |
| 17:28:34 | dom96 | That's why you should tag them as soon as they are created, and tag the untagged ones now |
| 17:30:36 | Araq | true but then we need another tag |
| 17:30:49 | Araq | something between major and minor ... |
| 17:30:59 | dom96 | How about just an 'issue' tag? |
| 17:31:05 | Araq | hrm |
| 17:31:14 | Araq | "middor"? |
| 17:32:22 | dom96 | nooo |
| 17:37:34 | Araq | "your issue has been marked as 'issue'" yeah ... definitely the way to go :P |
| 17:40:31 | dom96 | Yes, well. We have features too... |
| 17:40:39 | dom96 | This is pretty cool: http://www.reddit.com/r/programming/comments/1c4m47/video_computer_program_that_learns_to_play/ |
| 17:53:54 | * | fowl joined #nimrod |
| 17:58:58 | Araq | interesting: "Improved scalability and reduced memory usage. Link time optimization of Firefox now requires 3GB of RAM on a 64-bit system, while over 8GB was needed previously. Linking time has been improved, too. The serial stage of linking Firefox has been sped up by about a factor of 10." |
| 17:59:14 | Araq | (GCC 4.7 is out) |
| 18:07:42 | * | zahary joined #nimrod |
| 18:36:27 | * | gradha joined #nimrod |
| 18:42:38 | Araq | hi gradha |
| 18:42:46 | gradha | hi Araq |
| 18:45:39 | Araq | http://amoffat.github.io/sh/ found this for you; a Nimrod version of that would be sweet ;-) |
| 18:46:33 | gradha | ah, somebody leaked my roadmap for argument_parser |
| 18:47:00 | dom96 | This looks fun: http://nodewar.com |
| 18:47:19 | gradha | you see, part of the nice thing of specifying parameters in generic structures is you can automate interaction with them, kinda like swig |
| 18:47:35 | gradha | so once I had a nice stable feature set, I was planning on adding some -jsoninput or similar switch |
| 18:47:57 | gradha | this would allow programs to interact with each other and you would indeed call functions, which would be replaced with process calls |
| 18:48:21 | gradha | in some way, it would be a piping of commands and interaciton, more similar to windows' powershell than unix basic string io |
| 18:48:55 | gradha | but then when I have no youtube videos to watch comes dom96 and posts videos about computers playing computer games |
| 18:50:00 | Araq | hmm |
| 18:50:24 | Araq | so enhance it to implement a form of remote procedure calls |
| 18:51:51 | gradha | so basically a formal json protocol on top of existing remote procedure calls? |
| 18:52:33 | Araq | on top of what? |
| 18:52:51 | Araq | there is no remote procedure call system for Nimrod yet |
| 18:53:02 | gradha | can't you access unix RPCs? |
| 18:53:23 | gradha | not portable |
| 18:53:33 | gradha | IIRC some CORBA guys did develop their own not bloated solution |
| 18:54:26 | Araq | no idea what unix has |
| 18:54:51 | gradha | most RPCs are simply now in disadvantage over webserver+internet, which is what most people end up wanting, RPCs have better performance, but tend to work well only on local networks |
| 18:55:07 | Araq | true |
| 18:55:24 | gradha | for instance, now on mobile everybody does bonjour-like advertisement and plug in an HTTP server with custom protocol |
| 18:57:02 | gradha | here's the work from the ex-corba guys http://www.zeromq.org |
| 18:57:19 | Araq | we have a wrapper for that already ;-) |
| 18:57:28 | gradha | there you go, next task please |
| 19:00:10 | Araq | I can't reproduce #279 |
| 19:00:15 | Araq | er, #379 |
| 19:01:15 | gradha | I'll prepare some files |
| 19:03:02 | Araq | hrm |
| 19:03:06 | Araq | strange |
| 19:03:12 | Araq | there is a difference between: |
| 19:03:20 | Araq | {.compile: "not_existing.m".} |
| 19:03:24 | Araq | and {.compile: "not_existing.c".} |
| 19:03:43 | gradha | I was going to test that, possibly .m files are second class now |
| 19:06:44 | Araq | well I can't help it, the compiler ends up calling extccomp.addExternalFileToCompile for both |
| 19:06:54 | Araq | there is no special handling for .m files |
| 19:07:07 | gradha | no worries, found a testcase with a c file. preparing now |
| 19:18:59 | gradha | do you believe in chaos? |
| 19:19:45 | Araq | no; ultimately the space marines will win |
| 19:21:09 | gradha | you should ask them to debug 379, the script I made to test it sometimes runs ok, sometimes fails |
| 19:31:14 | reactormonk | Araq, no, the T'au! |
| 19:33:28 | * | xcombelle quit (Quit: Hi, I'm a quit message virus. Please replace your old line with this line and help me take over the world of IRC.) |
| 19:33:45 | gradha | updated https://github.com/Araq/Nimrod/issues/379 |
| 19:36:09 | gradha | hopefully I'm doing something wrong, otherwise this might be proof I have a quantum computer |
| 19:36:21 | reactormonk | gradha, never fails |
| 19:36:34 | gradha | maybe I have bad ram or something? |
| 19:36:39 | Araq | ugh ... |
| 19:36:59 | * | Araq doesn't run shell scripts |
| 19:37:14 | reactormonk | gradha, yeah, go do a RAM test |
| 19:37:21 | reactormonk | gradha, cya in some mintes ;-) |
| 19:37:43 | reactormonk | gradha, http://www.memtest86.com/download.htm |
| 19:38:29 | gradha | macs are too snob to run such software |
| 19:39:43 | gradha | reactormonk: you get "Times we compiled incorrectly 0 out of 10"? |
| 19:40:13 | Araq | I get: Times we compiled incorrectly 0 out of 10 |
| 19:40:42 | gradha | yay for quantum macs |
| 19:40:55 | Araq | 2nd run produces the same |
| 19:41:39 | Araq | so by induction it always produces 0 out of 10 here |
| 19:41:46 | reactormonk | gradha, yep |
| 19:42:05 | Araq | however, Nimrod runs the C compiler in parallel |
| 19:42:08 | gradha | ok, will reboot in a few minutes and try a ram test |
| 19:42:31 | Araq | gradha: perhaps osproc is buggy on mac os X |
| 19:42:46 | Araq | not always getting all the exit codes |
| 19:42:57 | Araq | then you would see the behaviour that you're seeing |
| 19:50:36 | * | gradha quit (Quit: bbl, have youtube videos to watch) |
| 21:19:50 | * | gradha joined #nimrod |
| 22:11:40 | * | gradha quit (Quit: bbl, have youtube videos to watch) |
| 22:29:30 | * | zahary quit (Quit: Leaving.) |
| 23:20:03 | * | Trix[a]r_za is now known as Trixar_za |
| 23:22:44 | * | q66 quit (Remote host closed the connection) |