| 00:10:28 | * | tinAndi quit (Quit: ChatZilla 0.9.91 [Firefox 33.1/20141106120505]) |
| 00:12:07 | * | bouliiii joined #nimrod |
| 00:12:32 | bouliiii | I have a problem with type inference: https://gist.github.com/bsegovia/e6b4081e50238efb941b. It does not work with auto |
| 00:15:20 | Araq | bouliiii: 'auto' return types are broken |
| 00:15:29 | Araq | it's a showstopper bug even |
| 00:15:41 | Araq | can't do more than that |
| 00:15:50 | Araq | (except fixing it) |
| 00:16:23 | bouliiii | Araq: OK. not problem. I could look at it. Maybe too high to climb for a newbie but it can be an interesting exercise |
| 00:20:05 | * | uber quit (Quit: bye) |
| 00:21:24 | * | superfunc joined #nimrod |
| 00:22:10 | * | Matthias247 quit (Read error: Connection reset by peer) |
| 00:22:48 | * | uber joined #nimrod |
| 00:23:01 | bouliiii | Araq: is there any debug mode in the compiler ? I mean a way to make it spit out everything it does (each compilation pass, each transformation it does...) |
| 00:24:33 | Araq | internal.txt now contains info of how I debug the compiler |
| 00:24:58 | Araq | there is no spit out everything it does because that's unreasonable |
| 00:25:16 | Araq | you inject strategic echo renderTree(n) calls instead |
| 00:25:28 | bouliiii | Araq: not eveerything but selectively (what the programmer thinks it may be important to log) |
| 00:25:44 | Araq | and if you don't guard it, you're doing it wrong ;-) |
| 00:27:53 | bouliiii | sorry for the question: where is internals.txt? |
| 00:28:02 | NimBot | Araq/Nimrod devel fd532b6 Araq [+1 ±1 -0]: fixes #206 |
| 00:28:27 | Araq | doc/intern.txt, sorry |
| 00:28:40 | bouliiii | thanks! |
| 00:28:54 | Araq | "Debugging the compiler" |
| 00:28:57 | Araq | only in devel |
| 00:29:00 | Araq | good night |
| 00:29:13 | bouliiii | See you. |
| 00:37:02 | * | uber quit (Changing host) |
| 00:37:02 | * | uber joined #nimrod |
| 00:58:17 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 00:58:49 | * | bouliiii quit (Ping timeout: 246 seconds) |
| 01:00:21 | * | q66[lap] joined #nimrod |
| 01:04:16 | * | BitPuffin quit (Ping timeout: 265 seconds) |
| 01:09:58 | * | Varriount joined #nimrod |
| 01:14:27 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 01:15:10 | * | q66[lap] joined #nimrod |
| 01:31:18 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 01:32:55 | * | darkf joined #nimrod |
| 01:33:01 | * | q66[lap] joined #nimrod |
| 01:45:38 | * | brson quit (Quit: leaving) |
| 01:46:09 | * | brson joined #nimrod |
| 01:57:07 | * | johnsoft quit (Ping timeout: 255 seconds) |
| 02:06:56 | * | kniteli joined #nimrod |
| 02:28:58 | * | q66 quit (Quit: Leaving) |
| 02:35:21 | * | kniteli quit (Ping timeout: 258 seconds) |
| 02:42:40 | * | superfunc quit (Quit: Connection closed for inactivity) |
| 02:47:32 | * | kniteli joined #nimrod |
| 02:58:44 | * | kniteli quit (Ping timeout: 258 seconds) |
| 03:10:51 | * | kniteli joined #nimrod |
| 03:13:14 | * | flaviu quit (Ping timeout: 255 seconds) |
| 03:25:47 | * | brson quit (Quit: leaving) |
| 04:05:49 | * | kniteli quit (Ping timeout: 258 seconds) |
| 04:06:17 | * | kniteli joined #nimrod |
| 04:42:33 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 04:43:16 | * | q66[lap] joined #nimrod |
| 04:51:10 | * | boydgreenfield joined #nimrod |
| 05:20:03 | * | Demos joined #nimrod |
| 05:21:47 | * | gokr_ quit (Remote host closed the connection) |
| 05:49:41 | * | untitaker quit (Ping timeout: 264 seconds) |
| 05:53:31 | * | BlaXpirit joined #nimrod |
| 05:55:07 | * | untitaker joined #nimrod |
| 05:59:25 | * | ARCADIVS joined #nimrod |
| 06:07:33 | * | boydgreenfield quit (Quit: boydgreenfield) |
| 06:21:50 | * | Demos quit (Read error: Connection reset by peer) |
| 06:22:17 | * | boydgreenfield joined #nimrod |
| 06:32:46 | Varriount | Araq, dom96: Something went wonky with permissions while I was doing something, and now the buildbot won't start via initscripts. |
| 06:33:02 | Varriount | For now, I'm running it manually. I'll fix it as soon as I can. |
| 07:12:45 | * | gour joined #nimrod |
| 07:12:49 | * | gour quit (Client Quit) |
| 07:13:44 | * | gour joined #nimrod |
| 07:14:35 | * | vendethiel quit (Quit: q+) |
| 07:18:40 | * | BitPuffin joined #nimrod |
| 07:24:47 | * | boydgreenfield quit (Quit: boydgreenfield) |
| 07:38:06 | * | boydgreenfield joined #nimrod |
| 07:45:52 | * | BitPuffin quit (Ping timeout: 255 seconds) |
| 07:50:50 | gour | morning |
| 07:51:26 | gour | are all blog posts considered as 'secondary source' for wikipedia? |
| 07:54:54 | * | boydgreenfield quit (Quit: boydgreenfield) |
| 08:04:04 | * | BitPuffin joined #nimrod |
| 08:27:20 | * | BitPuffin quit (Ping timeout: 255 seconds) |
| 08:29:25 | NimBot | Araq/Nimrod devel 729e048 Araq [+1 ±5 -0]: fixes #940 |
| 08:29:25 | NimBot | Araq/Nimrod devel c5cc20d Araq [+0 ±5 -0]: fixes #1548 |
| 08:31:20 | * | gokr_ joined #nimrod |
| 08:36:05 | * | Trustable joined #nimrod |
| 08:39:15 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 08:41:40 | * | q66[lap] joined #nimrod |
| 08:50:02 | * | kemet joined #nimrod |
| 09:11:56 | * | gokr_ quit (Ping timeout: 264 seconds) |
| 09:16:51 | * | wan joined #nimrod |
| 09:40:09 | gokr1 | Morning! |
| 09:52:04 | * | Etheco joined #nimrod |
| 09:52:48 | wan | Araq: thanks for #940 / #1649, it compiles and seems fixed! |
| 09:58:21 | * | darkf_ joined #nimrod |
| 10:00:32 | * | darkf quit (Ping timeout: 256 seconds) |
| 10:01:04 | * | wan quit (Quit: WeeChat 1.0.1) |
| 10:09:02 | * | kemet quit (Quit: Instantbird 1.5 -- http://www.instantbird.com) |
| 10:11:01 | * | darkf_ is now known as darkf |
| 10:14:09 | * | darkf_ joined #nimrod |
| 10:15:57 | * | darkf quit (Ping timeout: 256 seconds) |
| 10:41:18 | * | darkf_ is now known as darkf |
| 11:01:03 | * | khmm joined #nimrod |
| 11:37:11 | * | flaviu joined #nimrod |
| 11:44:36 | * | kemet joined #nimrod |
| 11:52:01 | * | flaviu quit (Ping timeout: 255 seconds) |
| 11:56:48 | * | kemet quit (Quit: Instantbird 1.5 -- http://www.instantbird.com) |
| 12:29:15 | * | johnsoft joined #nimrod |
| 12:32:24 | * | johnsoft quit (Read error: Connection reset by peer) |
| 12:39:29 | * | khmm quit (Ping timeout: 258 seconds) |
| 12:40:07 | * | kemet joined #nimrod |
| 12:40:07 | * | scidom joined #nimrod |
| 12:40:56 | * | scidom quit (Client Quit) |
| 12:41:08 | * | scidom joined #nimrod |
| 12:41:31 | * | kokozedman just used template for the first time: another wicked cool feature! Nim is awesome!!! |
| 12:41:37 | scidom | Hi all, I wanted to ask if there are any bindings for BLAS and LAPACK |
| 12:45:17 | kokozedman | scidom: http://nimrod-lang.org/lib.html doesn't list any... might want to search on Github if any. But creating bindings in Nim isn't a hard thing to do like in other languages. |
| 12:46:47 | * | khmm joined #nimrod |
| 12:46:49 | kokozedman | scidom: I'm a newbie myself, and I can assert it was not hard... may be it took me 2,5 hours to bind my Protobuf |
| 12:47:56 | scidom | Thanks a lot kokozedman, this is helpful. I am a newbie too, was checking the link you sent me. Perhaps I could use libsvm or Protbuf as examples of how to write bindings? |
| 12:49:43 | kokozedman | scidom: there doesn't seem to be any protobuf library yet at this time, I created a binding for my protobuf-c project |
| 12:49:57 | EXetoC | there is a libsvm module, which might or might not be up-to-date etc |
| 12:50:12 | EXetoC | but it's very small so it's probably easy to maintain it |
| 12:50:41 | scidom | Yeap, I only wanted to use libsvm as an example-prototype of how to create a C binding in nimrod, not necessarily to use libsvm at the moment |
| 12:52:09 | EXetoC | there is a c2nim guide on some blog I think. someone else might have the link to it |
| 12:52:54 | * | khmm quit (Ping timeout: 258 seconds) |
| 12:54:48 | scidom | Maybe this link is good? http://nimrod-lang.org/c2nim.html |
| 12:55:06 | scidom | Thanks both for the useful info, it is a good start |
| 12:55:47 | * | khmm joined #nimrod |
| 12:59:25 | * | johnsoft joined #nimrod |
| 13:00:15 | EXetoC | yes, but I think the article complements it. haven't read it yet though |
| 13:13:11 | kokozedman | scidom: gokr1 has a made several Nim articles ... one of them covers c2nim extensively |
| 13:13:16 | kokozedman | scidom: goran.krampe.se/category/nim |
| 13:13:43 | gokr1 | yeah |
| 13:13:44 | kokozedman | I mean, most of the practical aspects of c2nim |
| 13:14:19 | scidom | Thanks, that's a great blog! |
| 13:14:29 | kokozedman | I agree! :) |
| 13:14:37 | gokr1 | Thanks, I am cranking out articles as I go :) |
| 13:15:15 | kokozedman | gokr1: keep it up |
| 13:15:20 | kokozedman | we love it |
| 13:15:20 | gokr1 | Which reminds me I should turn that beginning of a hyperdex wrapper into a project. |
| 13:15:44 | gokr1 | I have ... an almost done article about "seq", then Exceptions. |
| 13:17:13 | scidom | Thanks gokr1, your blog is now bookmarked :) Thanks for the link kokozedman! |
| 13:18:07 | * | gokr1 reminds self to setup a Nim planet... |
| 13:19:02 | * | kokozedman wonders, where is the "Like" button for IRC |
| 13:24:17 | * | kemet quit (Ping timeout: 240 seconds) |
| 13:32:04 | * | kemet joined #nimrod |
| 13:32:12 | * | kemet quit (Client Quit) |
| 13:32:38 | * | scidom quit (Quit: Page closed) |
| 13:42:53 | * | BitPuffin joined #nimrod |
| 13:45:21 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 13:47:25 | * | q66[lap] joined #nimrod |
| 13:51:18 | * | flaviu__ joined #nimrod |
| 13:51:54 | flaviu__ | I came across mandrillapp.com, perhaps it can be used to simplify mailing on the forum. |
| 13:51:59 | * | flaviu__ quit (Client Quit) |
| 14:00:29 | * | fowl quit (Ping timeout: 264 seconds) |
| 14:00:42 | * | ARCADIVS quit (Quit: ARCADIVS) |
| 14:02:34 | * | perturbation joined #nimrod |
| 14:08:18 | * | fowl joined #nimrod |
| 14:16:21 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 14:18:23 | * | q66[lap] joined #nimrod |
| 14:25:16 | * | xyproto left #nimrod ("WeeChat 1.0.1") |
| 14:30:09 | * | MyMind joined #nimrod |
| 14:31:49 | * | Sembei quit (Ping timeout: 272 seconds) |
| 14:48:23 | * | darkf quit (Quit: Leaving) |
| 14:55:12 | * | brson joined #nimrod |
| 15:00:12 | * | AMorpork is now known as AFKMorpork |
| 15:00:15 | * | MyMind quit (Read error: Connection reset by peer) |
| 15:00:23 | * | Pisuke joined #nimrod |
| 15:05:17 | * | dom96_ quit (Ping timeout: 264 seconds) |
| 15:09:32 | * | brson quit (Quit: leaving) |
| 15:13:11 | * | BlaXpirit quit (Read error: Connection reset by peer) |
| 15:13:19 | * | BlaXpirit joined #nimrod |
| 15:15:21 | * | saml joined #nimrod |
| 15:20:50 | gour | what is the status in regard to wikipedia article? |
| 15:31:46 | * | khmm quit (Read error: Connection reset by peer) |
| 15:35:29 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 15:37:52 | * | q66[lap] joined #nimrod |
| 15:43:57 | * | kemet joined #nimrod |
| 15:53:40 | Araq | gour: wikipedia doesn't like me and I don't like wikipedia :P |
| 15:54:11 | Araq | Varriount: what's the status of the builder? and how can I see which tests fail? |
| 15:55:42 | gour | Araq: you're not alone. i stopped contributing due to case like Nimrod...still many are checking there, so i wonder what can be done in regard |
| 15:57:14 | gour | *cases |
| 16:11:15 | * | Boscop_ joined #nimrod |
| 16:15:16 | * | Boscop quit (Ping timeout: 255 seconds) |
| 16:16:54 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 16:17:37 | * | q66[lap] joined #nimrod |
| 16:18:36 | * | mko joined #nimrod |
| 16:21:08 | * | brson joined #nimrod |
| 16:30:34 | * | fowl quit (Ping timeout: 255 seconds) |
| 16:33:40 | gokr1 | Araq: http://178.62.143.63:8010/waterfall |
| 16:42:14 | Etheco | gokr1, you done any more blog posts ? :) |
| 16:42:23 | EXetoC | neat |
| 16:42:42 | * | xcombelle joined #nimrod |
| 16:45:38 | * | irrequietus joined #nimrod |
| 16:47:17 | * | Matthias247 joined #nimrod |
| 16:57:57 | * | brson quit (Ping timeout: 240 seconds) |
| 17:05:27 | * | brson joined #nimrod |
| 17:06:12 | gokr1 | Etheco: Not recently |
| 17:06:30 | gokr1 | Etheco: Just check goran.krampe.se/category/nim and you see all I have. |
| 17:06:55 | Etheco | you only had 3-4 posts on it |
| 17:07:01 | Etheco | awesome will have a read next week |
| 17:07:12 | Etheco | holiday paused me learning nim so :) |
| 17:09:23 | * | gour is curious, what are the most popular linux distros used by nim users/devs... |
| 17:09:52 | EXetoC | I use Arch |
| 17:10:38 | perturbation | I use CrunchBang (Debian Wheezy-ish) |
| 17:14:58 | gour | i'm on debian (sid) and wonder if anyone used opensuse recently |
| 17:27:00 | * | flaviu__ joined #nimrod |
| 17:28:11 | flaviu__ | gour: I used opensuse before arch, and it was nice. It was difficult to set up encryption the way I wanted, but that's about all the issues I had with it. |
| 17:28:42 | gour | flaviu__: i used arch before debian :-) |
| 17:29:21 | flaviu__ | It has a public build service, so you still have stuff like PPAs and the AUR |
| 17:31:42 | flaviu__ | Poorly phrased, but hopefully it's clear what I'm trying to say |
| 17:33:03 | * | q66 joined #nimrod |
| 17:33:11 | gour | np, i read about their build service |
| 17:36:46 | * | Boscop__ joined #nimrod |
| 17:39:52 | * | Boscop_ quit (Ping timeout: 250 seconds) |
| 17:40:09 | irrequietus | gour: I use opensuse for some years now, side to side with debian and nixos |
| 17:40:44 | irrequietus | irrequietus: it is a hassle-free, low-maintenance and quite uptodate distro |
| 17:40:47 | irrequietus | err |
| 17:40:51 | irrequietus | gour: ^ |
| 17:41:14 | gour | irrequietus: would you use it for server as well? |
| 17:41:41 | gour | i did start with suse back in '99, but changed few distros in between |
| 17:41:42 | irrequietus | whenever I had that chance, things worked; yast2 is quite easy to use if you are tired of cli |
| 17:42:39 | gour | irrequietus: in the past i left suse due to rpm dep-hell, now i read that yast is more capable as well as zypp |
| 17:42:56 | gour | irrequietus: so, you do nim on opensuse? |
| 17:42:59 | irrequietus | it is the most point-n-click friendly distro I know to be honest, without bringing down the house. |
| 17:43:05 | irrequietus | gour: I nim everywhere |
| 17:43:07 | irrequietus | :P |
| 17:43:13 | gour | :-) |
| 17:43:42 | gour | well, will try it under vbox |
| 17:43:53 | irrequietus | gour: zypper is a SAT solver |
| 17:44:24 | irrequietus | and, never had any rpm hell problem since they stopped fooling around |
| 17:44:29 | irrequietus | like it must be years |
| 17:44:32 | irrequietus | eons |
| 17:44:43 | irrequietus | cosmic scale timescale |
| 17:44:45 | irrequietus | :P~ |
| 17:45:23 | gour | well, i used suse in 1999 ;) |
| 17:45:48 | gour | however, i want distro to do less admin work |
| 17:45:55 | irrequietus | I beat you, because I had it when they only had 3.5" floppies :P~ |
| 17:46:15 | irrequietus | my experience with it is that when I want zero-maintenance, I pick that. |
| 17:46:35 | gour | heh, i startdd with 4.x, iirc and cd box :-) |
| 17:46:36 | irrequietus | hope we are not going too into off-topic ness, pm for anything you want in relation :) |
| 17:46:58 | gour | irrequietus: sure. i'll ask in #opensuse |
| 17:48:04 | irrequietus | or pm, I don't mind if I can help :) |
| 17:54:13 | * | brson quit (Quit: leaving) |
| 17:54:28 | * | brson joined #nimrod |
| 17:56:14 | * | q66[lap] quit (Read error: Connection reset by peer) |
| 17:58:57 | * | q66[lap] joined #nimrod |
| 18:00:12 | kokozedman | what's the Nim way of doing memset() ? I want to zero-fill an array[0..5, char] |
| 18:00:36 | * | vendethiel joined #nimrod |
| 18:06:34 | flaviu__ | arrays are 0 initialized by default, but http://nim-lang.org/system.html#zeroMem,pointer,int exists |
| 18:06:53 | flaviu__ | It's unsafe though, it's probably easier to just use a loop |
| 18:07:13 | kokozedman | flaviu__: thanks for pointing that out |
| 18:15:13 | * | flaviu__ quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) |
| 18:21:12 | * | flaviu__ joined #nimrod |
| 18:22:12 | * | flaviu__ quit (Client Quit) |
| 18:26:42 | * | enurlyx joined #nimrod |
| 18:27:45 | kokozedman | kind of trivial, but is there a way to get only the first 2 chars of a string? |
| 18:27:58 | kokozedman | I mean, Nim way of doing that |
| 18:29:22 | enurlyx | a[1..2] ? |
| 18:33:21 | enurlyx | it should be a[0..1] |
| 18:33:36 | kokozedman | enurlyx: nice! thanks |
| 18:39:30 | * | perturbation quit (Quit: Leaving) |
| 18:41:01 | Araq | gokr1: thanks but meh. I now depend on nimbuild's nice output ... |
| 18:49:23 | * | gour quit (Disconnected by services) |
| 18:49:25 | * | gour_ joined #nimrod |
| 18:49:45 | * | kemet quit (Quit: Instantbird 1.5 -- http://www.instantbird.com) |
| 18:53:24 | * | flaviu__ joined #nimrod |
| 18:55:06 | * | flaviu__ quit (Client Quit) |
| 18:55:37 | * | brson quit (Ping timeout: 240 seconds) |
| 18:58:30 | * | brson joined #nimrod |
| 18:58:35 | Varriount | Araq: I'm working on things as fast as I can, however the semester is nearing its end, and I have exams to study for and projects to do. |
| 18:59:29 | Araq | Varriount: don't worry, I will manage to read log files to see what breaks |
| 19:00:21 | Etheco | i love the windows installer :) |
| 19:00:27 | Etheco | didnt realise it install Aporia too |
| 19:00:55 | * | dom96_ joined #nimrod |
| 19:05:52 | * | irrequietus quit () |
| 19:06:00 | gokr1 | Araq: Mmmm, not sure what you "declined". As an admin you can trigger builds, that's fairly useful. And the output is what Buildbot does, so... I have nothing to do with it :) |
| 19:07:08 | Araq | well with the old nimbuild I get a nice "these tests now fail, these now work" feedback |
| 19:07:58 | * | Etheco quit (Read error: Connection reset by peer) |
| 19:08:50 | gokr1 | gour_: Hehe, distros... I hardly think there is a pattern to that just because you use Nim. We use CentOS for deployment because its basically RHEL, and RHEL is what most corporate customers wants. And we use Ubuntu for "everything else" since its the most supported/uptodate one, especially in cloud hosting. |
| 19:10:11 | gokr1 | I also like fringe distros, used Lunar Linux for a long time "personally". But when you start maintaining 20-30 machines, then you want to streamline. We are probably going to get rid of CentOS too. |
| 19:11:10 | * | gour_ is now known as gour |
| 19:13:05 | * | flaviu__ joined #nimrod |
| 19:14:59 | gour | gokr1: i'll try opensuse again, after many years...maybe it will get nim in sooner than debian :-) |
| 19:15:37 | gokr1 | gour: Just git clone and build. |
| 19:15:57 | gour | gokr1: i know, just joking. |
| 19:16:07 | * | gour have trunk built |
| 19:18:08 | * | flaviu__ quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) |
| 19:22:01 | * | enurlyx quit (Quit: Leaving.) |
| 19:39:39 | dom96_ | def-: I may have asked this before but what did you use to create your slides for GPN14? |
| 19:40:26 | def- | dom96_: latex |
| 19:40:37 | def- | i can send you the sources |
| 19:40:56 | dom96_ | def-: That would be awesome. What software did you use to display the slides? |
| 19:41:11 | def- | just a regular pdf viewer, evince |
| 19:48:46 | * | gour recommends atril - probably less deps |
| 19:49:28 | def- | i usually use zathura, but for showing slides i prefer evince |
| 19:52:23 | dom96_ | I may need to display it in Adobe Reader :( |
| 20:03:18 | NimBot | Araq/Nimrod devel 8f89c30 Dominik Picheta [+0 ±1 -0]: Ensure Request.body is initialised. |
| 20:06:12 | * | enurlyx joined #nimrod |
| 20:09:17 | * | johnsoft quit (Ping timeout: 265 seconds) |
| 20:09:37 | * | johnsoft joined #nimrod |
| 20:09:50 | * | Joe_knock joined #nimrod |
| 20:09:50 | * | Joe_knock quit (Changing host) |
| 20:09:50 | * | Joe_knock joined #nimrod |
| 20:10:25 | Varriount | Araq: I find it interesting that the 64 bit builders run faster than the 32 bit builders. |
| 20:13:37 | * | bjz quit (Read error: Connection reset by peer) |
| 20:13:56 | * | bjz joined #nimrod |
| 20:19:20 | * | brson quit (Ping timeout: 250 seconds) |
| 20:21:12 | enurlyx | Araq: Just for the record. Seems gcc has a bug, so that 'packed' does not work if -mms-bitfields is specified. Seems be standard on windows. With -mno-ms-bitfields 'packed' works. |
| 20:21:32 | enurlyx | https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52991 |
| 20:22:07 | * | bouliiii joined #nimrod |
| 20:23:24 | bouliiii | Araq: I just fixed msvc build... I just realized that mmx is not supported in 64 bits with msvc. What a mess. Anyway, it works with vs2013 in 32 and 64 bits (well I was not able to run avx/avx2 since my machine does not support them) |
| 20:24:05 | Araq | bouliiii: great |
| 20:26:18 | bouliiii | Araq: I will install a windows at home in a VM so I can test avx/avx2 on windows so it should be fine. I am still a bit confused about avx state in win7. They need to support it in the kernel and I am still not sure it works. Following Intel instructions using cpuid, last time I checked, Win7 SP1 was not supposed to run any avx. Fucking hell |
| 20:30:42 | * | iamsumesh joined #nimrod |
| 20:31:08 | * | iamsumesh left #nimrod ("Quit") |
| 20:40:16 | * | flaviu joined #nimrod |
| 20:41:52 | * | Mat3|coding joined #nimrod |
| 20:41:57 | Mat3|coding | hello |
| 20:47:55 | Araq | ping Varriount |
| 20:48:08 | Varriount | Araq: Yes? |
| 20:48:52 | Araq | bootstrapping fails on our server, likely reason: out of memory |
| 20:49:05 | Araq | how come the buildbot runs? |
| 20:50:16 | Varriount | Araq: Wait, you mean, you're trying to bootstrap nimrod on the VPS, the same machine running the build master? |
| 20:50:42 | Araq | yes |
| 20:51:13 | Varriount | Araq: Uhm, well, the buildbot process memory usage is relatively constant. |
| 20:52:16 | Varriount | I don't know too much about linux internals, but I would hazard the guess that the OS is still keeping some memory in reserve for the other processes. |
| 20:53:00 | Varriount | After all, if a single process runs out of memory, it wouldn't do for the entire system to crash. |
| 20:54:29 | * | gour quit (Quit: WeeChat 1.0.1) |
| 20:56:35 | flaviu | Varriount: Please don't say anything negative about linux or Araq will rant for hours |
| 20:56:42 | Joe_knock | If the buildbot is constantly active, then it is probably eating most of the memory. |
| 20:56:56 | flaviu | Varriount: Run `free` |
| 20:57:07 | flaviu | or `top` |
| 20:57:22 | Varriount | Joe_knock: Uh, it isn't eating most of the memory. |
| 20:57:41 | Joe_knock | What size is the VPS? |
| 20:58:43 | flaviu | Buildslave is eating ~5MB for me, I can't imagine the master uses much more. |
| 21:00:04 | Araq | well it has 512 MB of RAM |
| 21:00:10 | Araq | unless I'm reading the specs wrong |
| 21:00:33 | Varriount | The build master is using about 11% |
| 21:00:39 | Araq | and Nim itself takes 300MB for bootstrapping |
| 21:00:59 | flaviu | Isn't gokr providing linux build slaves? |
| 21:01:13 | Araq | flaviu: he is. |
| 21:01:51 | Araq | Varriount: doesn't the build master bootstrap ? |
| 21:01:54 | flaviu | I'd say the best solution is to reserve the VPS to host stuff and not do any hard work on it. |
| 21:03:31 | * | gour joined #nimrod |
| 21:04:22 | Varriount | Araq: No, why should it? |
| 21:04:47 | Araq | Varriount: ok I see so that's why it works |
| 21:04:58 | Varriount | Araq: The build slaves do all the work. The build master just coordinates stuff. |
| 21:05:39 | Araq | well this sucks |
| 21:05:56 | flaviu | Just curious, what distro does the VPS run? |
| 21:06:00 | Araq | why don't we have a swap |
| 21:06:26 | * | nande joined #nimrod |
| 21:06:26 | Araq | flaviu: ubuntu |
| 21:07:30 | * | bjz quit (Ping timeout: 256 seconds) |
| 21:07:50 | Joe_knock | How many of you are running 0.10.0 ? |
| 21:08:24 | Joe_knock | How much RAM is the server itself eating? |
| 21:08:37 | Joe_knock | I take it its an Apache server |
| 21:09:11 | * | Pisuke quit (Ping timeout: 255 seconds) |
| 21:09:21 | Varriount | Joe_knock: Uh, no. |
| 21:09:43 | Varriount | Joe_knock: buildbot.net |
| 21:11:41 | Araq | Joe_knock: pretty much everybody is using devel (0.10.1) afaict |
| 21:11:59 | Joe_knock | Alright, time to do the mighty upgrade. |
| 21:16:14 | * | Pisuke joined #nimrod |
| 21:17:03 | * | superfunc joined #nimrod |
| 21:17:23 | flaviu | ps aux | awk '{print $4, $11}' | sort -nr | head |
| 21:19:48 | * | Mat3|coding left #nimrod (#nimrod) |
| 21:30:48 | * | xcombelle quit (Quit: Quitte) |
| 21:36:00 | * | johnsoft quit (Read error: Connection reset by peer) |
| 21:37:46 | * | superfunc quit (Ping timeout: 250 seconds) |
| 21:43:00 | * | johnsoft joined #nimrod |
| 21:55:21 | Araq | Varriount: just fyi I added a swap and now it works |
| 21:56:15 | Varriount | Araq: Ok... but why are we using the VPS to boot Nim? |
| 21:57:05 | Araq | because we can |
| 21:57:25 | Araq | well we like to run the forum on this machine |
| 21:58:41 | Joe_knock | Does Nim eat up over 300MBs on a server? I'm unclear on this. |
| 21:59:17 | * | lyro joined #nimrod |
| 22:00:19 | Araq | Joe_knock: (81021 lines compiled; 4.401 sec total; 261.267MB) |
| 22:01:08 | * | bouliiii_ joined #nimrod |
| 22:01:09 | gokr1 | Hey |
| 22:01:09 | Araq | it's a bit weird that this requires more than 500MB RAM but I dunno how much GCC takes |
| 22:01:21 | gokr1 | The linux32 is a 512Mb machine, yes. |
| 22:01:35 | gokr1 | Its an old laptop of mine actually :) |
| 22:01:47 | gokr1 | The Linux64 is at Cloudsigma. |
| 22:01:56 | Araq | yeah but your laptop has a swap |
| 22:02:11 | Araq | our VPS didn't |
| 22:02:20 | Joe_knock | The compiler can't be that bloated. or can it? |
| 22:02:34 | gokr1 | oh :) |
| 22:02:46 | flaviu | Araq: Then add swap. Super-easy to do |
| 22:02:59 | gokr1 | I am going to bed, I am dead tired |
| 22:03:05 | gokr1 | cya tomorrow |
| 22:03:33 | bouliiii_ | Araq: There is a problem with aligned elements and msvc. There is no way to pass anything aligned on >16 bytes by values with msvc. So, when encapsulating __m128 with a new type, there is an issue. I can use template instead of proc but I am not sure this is going to scale for a complete type declaration. In C++, the WA is to use const ref. I am not sure for nim |
| 22:03:59 | bouliiii_ | >16 bytes -> >= 16 bytes |
| 22:04:07 | Araq | flaviu: already did, already said it fixes the problem |
| 22:04:19 | flaviu | Oh, my bad then. |
| 22:04:22 | flaviu | sorry |
| 22:06:49 | bouliiii_ | Araq: I can use ptr types I guess if I need to. In case I do not want to inline (saying matrix inversion). A bit ugly syntax wise but well can be OK |
| 22:07:10 | Araq | bouliiii_: 16 byte stack alignment is part of the ABI for x86_64 |
| 22:07:35 | * | clone1018_ quit (Quit: No Ping reply in 180 seconds.) |
| 22:07:46 | * | clone1018_ joined #nimrod |
| 22:08:41 | Araq | Joe_knock: well the Nim compiler definitely could be optimized when it comes to RAM usage, but it's workable |
| 22:10:07 | * | AFKMorpork is now known as AMorpork |
| 22:12:16 | bouliiii_ | Araq: I have to recheck on x64. But it does not fix 32 bits builds and avx requires 32 bytes alignment anyway |
| 22:12:58 | Araq | bouliiii_: you can use .byref to make the compiler use ptr under the hood |
| 22:13:10 | Araq | no need to do it yourself |
| 22:13:50 | Araq | avx requires 32 bytes alignment? wtf |
| 22:16:40 | * | bouliiii_ quit (Ping timeout: 246 seconds) |
| 22:16:49 | * | bouliiii__ joined #nimrod |
| 22:17:38 | bouliiii__ | Araq: not *exactly*. Here: https://software.intel.com/en-us/node/514150 Aligned loads / stores yes. But unaligned (_mm256_loadu_ps) nope |
| 22:17:54 | bouliiii__ | but msvs is not happy anyway |
| 22:17:57 | bouliiii__ | msvc |
| 22:18:22 | bouliiii__ | Because __m256 is 32 bytes aligned and it does not want anything like this as function argument. Period. |
| 22:18:58 | dom96_ | Varriount: Does buildbot create nightly releases for us? |
| 22:19:50 | Araq | bouliiii__: btw what's the point of avx? I'd rather run these things on a GPU |
| 22:23:19 | * | bouliiii___ joined #nimrod |
| 22:23:30 | bouliiii___ | Araq: I am writing compilers for GPU and used to write drivers also. So, well, I'd rather not :-) CPU code is robust, GPU code, well... I guess it is debatable but beyond video games, I am still not convinced. But a GPU support for nim would be anyway cool for sure. I guess users have to decide what is best :-) |
| 22:24:28 | flaviu | dom96_: I'd rather it not, while my upward network connection isn't bad, I'd still like to avoid using it. |
| 22:24:43 | * | bouliiii__ quit (Ping timeout: 246 seconds) |
| 22:25:08 | dom96_ | flaviu: What architecture are you running a build bot on? |
| 22:25:16 | flaviu | ARM v5 |
| 22:26:46 | flaviu | It takes 11 hours to build on ARM, :O |
| 22:27:33 | Araq | flaviu: ha ha that'll teach you about "who cares about compiler efficiency" ;-) |
| 22:27:50 | Araq | "let's have MOAR passes" |
| 22:28:01 | flaviu | I wouldn't mind if it took 24h if it meant there weren't bugs. |
| 22:28:25 | Araq | welcome to reality where you get slower + bugs instead |
| 22:28:55 | EXetoC | Araq: must it not be a long-running task for it to pay off? |
| 22:32:03 | EXetoC | more parallelizable in other words |
| 22:32:30 | * | Trustable quit (Quit: Leaving) |
| 22:34:55 | * | gour quit (Quit: WeeChat 1.0.1) |
| 22:38:21 | * | nande quit (Remote host closed the connection) |
| 22:38:43 | * | nande joined #nimrod |
| 22:42:21 | * | bouliiii_ joined #nimrod |
| 22:42:34 | * | bouliiii___ quit (Ping timeout: 246 seconds) |
| 22:42:55 | * | brson joined #nimrod |
| 22:44:23 | * | BitPuffin quit (Ping timeout: 258 seconds) |
| 23:01:17 | * | _bouliiii_ joined #nimrod |
| 23:05:19 | * | _bouliiii_ quit (Ping timeout: 246 seconds) |
| 23:26:21 | Varriount | dom96: Not yet, though that's trivial to add. |
| 23:27:17 | Varriount | dom96: Again, there's lots to do, I just haven't had the time. I want to try and update the buildmaster config tonight, to match what is in the repository. |
| 23:33:29 | * | brson quit (Quit: leaving) |
| 23:33:43 | * | brson joined #nimrod |
| 23:40:04 | * | nande quit (Remote host closed the connection) |
| 23:40:08 | * | brson quit (Quit: leaving) |
| 23:40:30 | * | nande joined #nimrod |
| 23:41:29 | Joe_knock | Anybody have a recommended method to update Nim to 0.10.0?? |
| 23:42:18 | * | bouliiii_keeps_o joined #nimrod |
| 23:42:46 | Araq | Joe_knock: use the csources based installation |
| 23:42:55 | Araq | I don't think 0.9.6 can compile 0.10.0 |
| 23:43:03 | Araq | (as usual) |
| 23:43:07 | * | bouliiii_ quit (Ping timeout: 246 seconds) |
| 23:43:21 | Joe_knock | I'm on the 0.9.4 build? |
| 23:43:26 | Araq | I think there has not been a single release where the old version was able to compile the newer |
| 23:43:26 | Joe_knock | *. |
| 23:43:53 | * | _bouliiii_ joined #nimrod |
| 23:45:47 | * | Matthias247 quit (Read error: Connection reset by peer) |
| 23:46:37 | * | bouliiii_keeps_o quit (Ping timeout: 246 seconds) |
| 23:46:59 | Joe_knock | Araq: So I should follow the instructions here: https://github.com/nim-lang/csources ?? |
| 23:48:56 | * | enurlyx quit (Quit: Leaving.) |
| 23:50:21 | Araq | Joe_knock: http://nimrod-lang.org/download.html, but replace master with devel |
| 23:52:24 | Joe_knock | Araq: Do I just delete the old Nimrod folder? |
| 23:53:39 | Araq | Joe_knock: depends on how you installed it |
| 23:53:59 | Joe_knock | It looks like I did the same github installation as before. |
| 23:54:17 | Araq | but yeah, you can simply delete it or rename it |
| 23:56:10 | Joe_knock | Good idea. Let me keep the old one and re-do the install. |