| 00:00:00 | * | leorize quit (Ping timeout: 260 seconds) |
| 00:03:57 | * | krux02_ joined #nim |
| 00:06:43 | * | krux02 quit (Ping timeout: 250 seconds) |
| 00:24:49 | * | krux02_ quit (Remote host closed the connection) |
| 00:51:25 | FromDiscord | <exelotl> Kinda threw me that arr.len doesn't work for var arr: ptr array[5, int] |
| 00:52:15 | FromDiscord | <exelotl> I guess arr[].len would be ok |
| 00:58:57 | * | ng0 quit (Quit: Alexa, when is the end of world?) |
| 01:11:46 | madprops | made another thing https://github.com/madprops/lester |
| 01:13:41 | disruptek | madprops: checkout the cligen package on github. i think you'll like it. |
| 01:23:18 | skrylar[m] | well most of glib is dealt with. couple edge cases but meh |
| 01:29:42 | madprops | im trying to make the compiler know that all code is in /src |
| 01:29:46 | madprops | im doing srcDir = "src" |
| 01:29:51 | madprops | in the .nimble file |
| 01:29:54 | madprops | but that doesn't seem to work |
| 01:30:40 | mipri | that's what I do, and it finds code in src. why do you think it doesn't work? |
| 01:31:19 | madprops | for instance if i do: nim compile -d:release lester |
| 01:31:26 | madprops | it expects to find lester.nim |
| 01:31:45 | mipri | it's 'nimble' that cares about your .nimble file, not nim. |
| 01:32:21 | mipri | if you also have `bin = @["lester"]`, then you can 'nimble build' to build it. |
| 01:34:41 | madprops | is there an equivalent for nim compile --run with nimble? |
| 01:34:47 | madprops | nimble run seems to make a package |
| 01:35:19 | shashlick | nimble c |
| 01:36:12 | madprops | so im guessing it needs an argument |
| 01:36:19 | FromGitter | <kaushalmodi> madprops: yes, use config.nims |
| 01:36:37 | madprops | so far i have srcDir and bin in there |
| 01:36:51 | FromGitter | <kaushalmodi> E.g.: https://github.com/kaushalmodi/elnim/blob/master/tests/config.nims |
| 01:37:03 | mipri | nimble c doesn't respect srcDir settings annoyingly. |
| 01:37:25 | mipri | use 'nimble run lester' |
| 01:37:59 | FromGitter | <kaushalmodi> I use nimble only to store the nimble package metadata |
| 01:38:11 | FromGitter | <kaushalmodi> Everything else goes in config.nims |
| 01:39:30 | madprops | http://i.imgur.com/Dti1dtt.png |
| 01:39:45 | madprops | http://i.imgur.com/yNhQ1Nd.png |
| 01:40:31 | mipri | 'nimble c' takes flags, and since it doesn't respect your .nimble it isn't more useful than 'nim c' for this anyway. |
| 01:40:41 | mipri | 'nimble run' does what you want. you just need to pass the name of the binary. |
| 01:41:08 | mipri | that's an unfortunate error message from nimble though. |
| 01:41:23 | * | abm quit (Quit: Leaving) |
| 01:41:27 | madprops | btw, is nimble build a release bin? |
| 01:42:37 | madprops | or do i need -d:release |
| 01:42:44 | mipri | with --debug nimble will print out the exact commands it's running. |
| 01:43:01 | mipri | you need -d:release |
| 01:43:30 | madprops | can't see a reason to not use just the compiler lol |
| 01:43:39 | disruptek | i'm with you, buddy. |
| 01:44:40 | mipri | you mostly can. nimble gives you an alternative to makefiles for custom tasks, and it handles dependencies, and it has some reasonable tasks by default, like 'nimble test' |
| 01:45:18 | * | endragor joined #nim |
| 01:45:44 | * | endragor quit (Remote host closed the connection) |
| 01:46:22 | * | endragor joined #nim |
| 01:46:53 | FromGitter | <kaushalmodi> mipri: config.nims serves as more general purpose Makefile replacement |
| 01:47:10 | FromGitter | <kaushalmodi> nimble test doesn't run the tests in runnableExamples, for example |
| 01:48:09 | FromGitter | <kaushalmodi> config.nims also has the advantage that you define your favorite tasks once in your global config.nims, and then reuse them in all your projects |
| 01:48:30 | disruptek | yer blowin' my mind right now. |
| 01:48:47 | mipri | that sounds like an annoyance rather than an advantage. how do other people get your favorite tasks? |
| 01:48:52 | FromGitter | <kaushalmodi> disruptek: serious or sarcastic? |
| 01:49:14 | disruptek | there's an only directory where you can trade tasks. |
| 01:49:19 | disruptek | s/only/online/ |
| 01:49:49 | FromGitter | <kaushalmodi> mipri: I git clone my config on my Travis CIs |
| 01:50:26 | disruptek | mipri: he really does have a 300 line config.nims. |
| 01:50:54 | FromGitter | <kaushalmodi> disruptek: most of those is for the musl build support |
| 01:51:17 | disruptek | disbot: you awake, buddy? |
| 01:51:18 | disbot | yep. 😊 |
| 01:51:20 | FromGitter | <kaushalmodi> but at least now "nim musl foo.nim" just works for any of my projects ;) |
| 01:56:48 | * | endragor_ joined #nim |
| 01:57:41 | disruptek | i admit, that sounds like the right way to impl that. |
| 01:58:36 | * | endragor quit (Read error: Connection reset by peer) |
| 01:58:39 | * | endragor_ quit (Remote host closed the connection) |
| 02:05:46 | * | uu91 quit (Read error: Connection reset by peer) |
| 02:06:04 | * | uu91 joined #nim |
| 02:12:59 | madprops | how can i compile --run but create the binary somewhere specific? |
| 02:13:08 | madprops | or maybe i should just use nimble |
| 02:14:45 | mipri | just combine the flags |
| 02:14:59 | mipri | nim -c -o:wat -r src/lester.nim |
| 02:15:08 | mipri | without the first - |
| 02:16:06 | madprops | nice ty |
| 02:17:21 | madprops | is there a way to change the context of --run ? |
| 02:17:28 | mipri | what do you mean by 'context'? |
| 02:17:36 | madprops | i put it in bin/lester but i think it's running it from the root dir |
| 02:18:09 | madprops | nim c -o:bin/lester -r src/lester.nim |
| 02:18:41 | madprops | if i go to bin/ and run it, it works fine |
| 02:19:31 | madprops | i guess i could cd somewhere first |
| 02:20:03 | mipri | that's what I'd do, if the current directory matters to the command. |
| 02:20:20 | mipri | ( cd bin; nim c -r ../src/lester.nim ) |
| 02:21:27 | madprops | interestingly. if i omit -o lester it places it in src |
| 02:22:15 | * | Kevin5 joined #nim |
| 02:22:33 | * | Jjp137 quit (Ping timeout: 268 seconds) |
| 02:23:46 | madprops | now i have to figure out to do the same but with -d:release |
| 02:24:15 | madprops | it expects a filename. but a path doesn't work |
| 02:30:58 | disruptek | --outdir="somedirectory" sets the location to output binaries. |
| 02:31:48 | madprops | worked |
| 02:34:00 | * | ponyrider quit (Quit: WeeChat 2.6) |
| 02:34:19 | * | Jesin joined #nim |
| 02:37:32 | * | Kevin5 quit (Remote host closed the connection) |
| 02:43:52 | madprops | i want to make the working dir as the binary's dir |
| 02:43:55 | madprops | im doing setCurrentDir(newDir=".") |
| 02:43:58 | madprops | but that didn't work |
| 02:44:51 | disruptek | you need to operate on the path of the app. |
| 02:45:09 | madprops | thing is sometimes i call it as bin/lester |
| 02:45:14 | madprops | instead of going to bin first |
| 02:45:19 | madprops | it should work in both cases |
| 02:45:32 | disruptek | setCurrentDir(getAppDir()) |
| 02:45:56 | madprops | ace |
| 02:49:47 | * | leorize joined #nim |
| 02:50:18 | * | lritter quit (Ping timeout: 245 seconds) |
| 02:51:12 | * | lritter joined #nim |
| 02:51:42 | madprops | this is so weird |
| 02:51:56 | madprops | does echo auto inserts a newline before it starts? |
| 02:52:01 | madprops | somehow the start of the program is printing a newline |
| 02:52:13 | disruptek | nah. |
| 02:53:18 | madprops | oops :) |
| 02:53:36 | madprops | had a sneaky debugging echo |
| 03:06:22 | * | ponyrider joined #nim |
| 03:06:38 | * | rockcavera quit (Remote host closed the connection) |
| 03:09:36 | * | Jjp137 joined #nim |
| 03:15:51 | * | uu91 quit (Read error: Connection reset by peer) |
| 03:16:08 | * | uu91 joined #nim |
| 03:38:24 | * | Kevin5 joined #nim |
| 03:39:25 | * | endragor joined #nim |
| 03:51:42 | * | uu91 quit (Read error: Connection reset by peer) |
| 03:51:58 | * | uu91 joined #nim |
| 04:03:09 | * | bacterio quit (Read error: Connection reset by peer) |
| 04:10:02 | * | theelous3 quit (Ping timeout: 240 seconds) |
| 04:12:25 | * | uu91 quit (Read error: Connection reset by peer) |
| 04:12:38 | * | uu91 joined #nim |
| 04:14:26 | * | ponyrider quit (Quit: WeeChat 2.6) |
| 04:15:06 | * | Kevin5 quit (Remote host closed the connection) |
| 04:16:19 | * | theelous3_ joined #nim |
| 04:24:01 | * | ponyrider joined #nim |
| 04:45:47 | * | chemist69 quit (Ping timeout: 250 seconds) |
| 04:47:39 | * | chemist69 joined #nim |
| 04:50:28 | * | nsf joined #nim |
| 05:17:55 | FromGitter | <Vindaar> @deech: there's a package for clojure persistent vectors by PMunch here https://github.com/PMunch/nim-persistent-vector ⏎ I use it and its working fine for me. |
| 05:57:09 | * | ltriant quit (Quit: leaving) |
| 06:19:48 | * | uu91 quit (Read error: Connection reset by peer) |
| 06:20:09 | * | uu91 joined #nim |
| 06:25:19 | * | narimiran joined #nim |
| 06:27:57 | * | onionhammer quit (Ping timeout: 240 seconds) |
| 06:28:20 | * | onionhammer joined #nim |
| 06:33:08 | * | dddddd quit (Remote host closed the connection) |
| 06:34:23 | * | solitudesf joined #nim |
| 06:36:25 | Araq | https://arxiv.org/pdf/1902.05178.pdf |
| 06:58:37 | * | lritter quit (Ping timeout: 240 seconds) |
| 07:08:38 | madprops | so putting globals in their own module and importing the module works. wondering if it's possible to expose globals defined in the main file to other modules |
| 07:10:21 | Araq | it doesn't |
| 07:27:39 | * | PMunch joined #nim |
| 07:30:31 | * | Kevin5 joined #nim |
| 07:32:04 | Mister_Magister | how can i specify compiler binary with --cpu:arm --os:linux |
| 07:34:30 | PMunch | Just came up with a topic for a FOSDEM talk that someone might want to do: Functional benefits in an imperative language |
| 07:36:08 | PMunch | On how Nim can track side-effects, has a distinction between `proc` and `func`, and how things like zero-functional works |
| 07:37:28 | * | kevin joined #nim |
| 07:37:51 | * | kevin is now known as Guest11723 |
| 07:37:53 | * | Kevin4 joined #nim |
| 07:42:56 | * | tane_ joined #nim |
| 07:46:07 | Mister_Magister | well i got that but now i'm getting stdlib_assertions.nim.c:10:10: fatal error: sys/types.h: No such file or directory |
| 07:46:46 | * | filcuc joined #nim |
| 07:47:24 | PMunch | --gcc.exe? |
| 07:47:45 | PMunch | And you possibly also want to set --gcc.linkerexe |
| 07:48:36 | PMunch | For reference, this is how I cross-compile to Windows from Linux: nim c --os:windows --gcc.exe:x86_64-w64-mingw32-gcc --gcc.linkerexe:x86_64-w64-mingw32-gcc |
| 07:48:59 | Mister_Magister | PMunch thanks but i alrady got that now i have he problem above |
| 07:49:13 | Mister_Magister | well to be more exact |
| 07:49:25 | Mister_Magister | --arm.linux.gcc.exe:arm-suse-linux-gnueabi-gcc --arm.linux.gcc.linker:arm-suse-linux-gnueabi-gcc |
| 07:51:13 | PMunch | Might need to set a gcc.path? |
| 07:52:50 | Mister_Magister | PMunch but it's /usr/bin not anything connected with /usr/include |
| 07:53:09 | PMunch | Yeah I'm just guessing here.. |
| 07:53:24 | PMunch | Not really sure where it tries to find sys/types.h when cross compiling.. |
| 07:54:08 | PMunch | Might be a separate package from the one that ships arm-suse-linux-gnueabi-gcc |
| 07:54:15 | PMunch | Which distro are you running? |
| 07:55:13 | Mister_Magister | opensuse |
| 07:55:33 | Mister_Magister | btw if i compile for arm does that mean int64 is not available? |
| 07:56:53 | PMunch | I think int64 would be available, but possibly being rewritten by the compiler to something less efficient |
| 07:57:29 | Mister_Magister | my program parses json to int64 and i'm getting error about out of range… |
| 07:58:20 | Mister_Magister | Error: unhandled exception: Parsed integer outside of valid range |
| 07:58:31 | PMunch | The default `int` might be int32 |
| 07:59:15 | Mister_Magister | yeah but i think it's using int64 |
| 08:00:00 | * | gmpreussner quit (Quit: kthxbye) |
| 08:00:17 | Mister_Magister | maybe i will try building whole nim on arm and then compiling |
| 08:00:31 | PMunch | Hmm, which package did you install to get the cross compiler? |
| 08:00:52 | PMunch | Slight "warning", building Nim on an arm processor might take a little while |
| 08:01:33 | Mister_Magister | yeah i know but ubuntu doesn't have newer packages |
| 08:01:44 | Mister_Magister | i've updated nim on opensuse and made it available on arm but well |
| 08:04:27 | * | gmpreussner joined #nim |
| 08:14:38 | Kevin4 | 👑️ |
| 08:17:53 | * | livcd quit (Quit: Lost terminal) |
| 08:20:15 | * | livcd joined #nim |
| 08:30:17 | * | venk joined #nim |
| 08:30:39 | FromDiscord | <itmuckel> Hey guys! https://github.com/itmuckel/nimtellij <-- I started a repository for an intellij plugin. The only thing it supports right now is syntax highlighting and code completion, but it might be something to build upon. |
| 08:30:40 | FromDiscord | <itmuckel> |
| 08:30:40 | FromDiscord | <itmuckel> Someone in this chat was interested in building an intellij plugin, but I can't remember who it was. Just letting you know, if someone wants to dig into the IntelliJ plugin framework to make this plugin useable, just do it. 🙂 |
| 08:31:08 | Mister_Magister | @itmuckel OH GOD THANKS LORD TO YOU |
| 08:31:33 | Mister_Magister | jesus that's what i've been waiting for. some sane IDE for nim |
| 08:31:42 | FromDiscord | <itmuckel> 😄 |
| 08:32:22 | Kevin4 | thanks ! very helpful |
| 08:32:29 | * | Kevin5 quit (Remote host closed the connection) |
| 08:32:59 | FromDiscord | <itmuckel> I'll comment out the code a bit more to make it easier to get into it |
| 08:33:50 | Mister_Magister | do you plan on doing prebuilds? |
| 08:34:20 | FromDiscord | <itmuckel> Oh, yeah, wait, I'll build a current version and upload it |
| 08:35:28 | FromDiscord | <itmuckel> But you still need nimlsp in your path like written in the README |
| 08:36:04 | FromDiscord | <itmuckel> So in fact this plugin doesn't do much on it's own. It connects intellij to the nimlsp and even for that it uses a library. 😄 |
| 08:36:48 | Mister_Magister | that's plenty already |
| 08:37:04 | Mister_Magister | and you can make nim buildtarget without plugin even i believe |
| 08:42:03 | * | venk quit (Quit: ERC (IRC client for Emacs 27.0.50)) |
| 08:42:54 | FromDiscord | <itmuckel> https://github.com/itmuckel/nimtellij/releases/tag/0.1 Okay, here you go. |
| 08:43:16 | Mister_Magister | thanks! |
| 08:43:26 | FromDiscord | <itmuckel> There's still much to do and I wanted to actually program a bit in nim, before writing IDE support for it, but maybe I find people who also want to work on it. |
| 08:44:41 | FromGitter | <zacharycarter> vscode is still probably better |
| 08:45:04 | FromGitter | <zacharycarter> but cool either way |
| 08:45:08 | Mister_Magister | @zacharycarter let's not go into that :) |
| 08:45:22 | Mister_Magister | everyone has their preferences and that's fine |
| 08:45:29 | FromGitter | <zacharycarter> no it's not about preferences |
| 08:45:38 | FromGitter | <zacharycarter> what makes IntellIiJ at all decent is its refactoring capabilities |
| 08:45:42 | Mister_Magister | oh here we go again |
| 08:46:07 | Mister_Magister | PMunch it wasn't that long lol |
| 08:46:24 | FromGitter | <zacharycarter> it's not even arguable it's fact - it's like how Visual Studio is shit without resharper and resharper is what makes Rider at all decent |
| 08:46:54 | FromGitter | <zacharycarter> unless the runtime / tooling supports those features any IDE will be not so great |
| 08:47:31 | FromGitter | <zacharycarter> and an editor will be just as good if not better - and I'm saying VSCode is probably still better because the Nim plugin supports more than syntax highlighting and auto complete |
| 08:47:39 | FromGitter | <zacharycarter> I think you're misunderstanding what I'm saying |
| 08:48:05 | FromGitter | <zacharycarter> you can replace the name IntelliJ with any IDE that offers Nim support - and the facts will be the same |
| 08:48:23 | FromDiscord | <itmuckel> Yeah, without all the refactoring stuff, IntelliJ is just a big text editor with syntax highlighting. |
| 08:48:28 | FromGitter | <zacharycarter> exactly |
| 08:48:54 | PMunch | Mister_Magister, I was about to start it earlier :P |
| 08:49:08 | PMunch | Your statement "jesus that's what i've been waiting for. some sane IDE for nim" kinda invited it :P |
| 08:49:59 | FromGitter | <zacharycarter> in fact IntelliJ without that tooling is arguably worse than VSCode |
| 08:50:04 | FromGitter | <zacharycarter> all that mouse clicking to set up build configurations etc |
| 08:50:06 | PMunch | Personally I've been programming in Vim without anything but syntax highlighting for a long time |
| 08:50:19 | FromGitter | <zacharycarter> same but VSCode |
| 08:50:23 | PMunch | Now I use nimlsp mostly for it's LspDefinition feature |
| 08:50:26 | FromGitter | <zacharycarter> I guess I have click through |
| 08:50:33 | FromGitter | <zacharycarter> but that's about all I use |
| 08:50:58 | Mister_Magister | PMunch yeah kinda :P |
| 08:51:35 | FromDiscord | <itmuckel> @PMunch @zacharycarter |
| 08:51:36 | FromDiscord | <itmuckel> https://cdn.discordapp.com/attachments/371759389889003532/644097013578727424/7b2.png |
| 08:52:05 | FromGitter | <zacharycarter> well - program in Nim for long enough and you won't really need them |
| 08:52:07 | Mister_Magister | let's all use vim |
| 08:52:14 | FromGitter | <zacharycarter> vim is fine |
| 08:52:31 | FromGitter | <zacharycarter> modal editors are popular for good reason |
| 08:52:54 | FromDiscord | <itmuckel> I use ideavim in IntelliJ |
| 08:53:32 | FromGitter | <zacharycarter> I have to use IntelliJ at work because Java, but if I wasn't using Java I'd uninstall it immediately |
| 08:54:37 | FromDiscord | <itmuckel> Really? Why? |
| 08:55:27 | FromDiscord | <itmuckel> At work we use it too for angular/typescript/js and everything else and it's really awesome. |
| 08:55:54 | FromGitter | <zacharycarter> well at home I mostly use C/C++ and Nim |
| 08:56:03 | FromGitter | <zacharycarter> so it's of limited use there |
| 08:56:17 | FromGitter | <zacharycarter> and at work for any web stuff I use VSCode |
| 08:56:46 | FromDiscord | <itmuckel> What do you use for C/C++? |
| 08:57:02 | FromGitter | <zacharycarter> I don't like using an IDE when an editor will suffice - they tend to involve a lot more mouse clicking and project setup and large runtime memory footprints |
| 08:57:21 | PMunch | zacharycarter, totally agree |
| 08:57:32 | tane_ | vscode works really well for most stuff actually |
| 08:57:40 | FromGitter | <zacharycarter> VSCode if I'm on linux and if I'm on windows sometimes Visual Studio. If on Mac then XCode |
| 08:57:44 | tane_ | despite being an electron app :) |
| 08:57:44 | FromGitter | <zacharycarter> agreed |
| 08:57:51 | FromGitter | <zacharycarter> it's a pretty efficient electron app |
| 08:57:58 | FromGitter | <zacharycarter> they've done well with it in that regard |
| 08:58:08 | PMunch | I use Vim for everything. Keeping my tools separated, an editor for editing, build tools for project setup, git for version control, and the terminal to tie them all together |
| 08:58:18 | tane_ | PMunch, do you have nim completion in vim? |
| 08:58:30 | PMunch | I have Nim completion in my head |
| 08:58:44 | PMunch | But yes, my Vim does have nimlsp, so I can trigger completions |
| 08:59:04 | tane_ | ok |
| 08:59:29 | tane_ | I have neovim + CoC, but it does not work out of the box at the moment, not sure if not too lazy to set that up :) |
| 09:00:19 | PMunch | CoC? |
| 09:00:31 | tane_ | https://github.com/neoclide/coc.nvim |
| 09:00:33 | FromDiscord | <itmuckel> I can't remember the standard library and which package has what methods and the docs for anything. So I like my IDE to tell me.^^ |
| 09:01:20 | FromGitter | <zacharycarter> I needed it at first but eventually it got annoying, especially in VSCode |
| 09:01:28 | FromGitter | <zacharycarter> where the suggestions would appear over where you were trying to type |
| 09:01:48 | FromGitter | <zacharycarter> and there was something else funky like, you had to manually click out of the auto suggestion box, in all it was just terrible |
| 09:02:16 | FromDiscord | <itmuckel> Yeah, that's the current state of it.^^ |
| 09:03:24 | FromGitter | <zacharycarter> I'm going to be doing a lot of Nim programming this week / weekend. Starting to get motivated now that my network code is written and my map parsing code is coming along |
| 09:03:30 | FromGitter | <zacharycarter> networked rts incoming |
| 09:04:09 | PMunch | Woop woop! |
| 09:04:11 | FromGitter | <zacharycarter> https://assetstore.unity.com/packages/3d/characters/humanoids/toon-rts-units-67948 - going to use these units as placeholder art |
| 09:05:02 | FromGitter | <zacharycarter> which I guess will kind of suck because I can't release them FOSS - right now my games github repo is private |
| 09:05:34 | FromGitter | <zacharycarter> eventually I'll try to recruit a 3d artist from work to make me some |
| 09:05:53 | FromGitter | <zacharycarter> that or I'll spend a few weeks making some really bad programmer art |
| 09:10:52 | PMunch | That is ultimately what have slowed my game design to a crawl, the constant struggle for assets |
| 09:11:06 | * | ng0 joined #nim |
| 09:11:34 | FromGitter | <zacharycarter> yeah - it's mostly why I switched to 3d, I feel like 3d assets are way easier to come by |
| 09:11:48 | FromGitter | <zacharycarter> especially isometric 2d art |
| 09:11:58 | FromGitter | <zacharycarter> unless you want to make your own from 3d art which is a chore in of itself |
| 09:12:43 | FromGitter | <zacharycarter> at least with 3d I can buy decent looking stuff from the asset store / find open source art on blendswap and then release a game with it |
| 09:12:52 | FromGitter | <zacharycarter> but it is a major pain / struggle |
| 09:13:10 | FromGitter | <zacharycarter> why does art have to be so hard? :P |
| 09:16:49 | PMunch | I just normally do 2D, because at least then I can change stuff if I need to |
| 09:17:20 | PMunch | So I typically end up finding a spritesheet and then just changing some colours around to create more characters :P |
| 09:19:26 | FromGitter | <zacharycarter> yeah - unless you have the source files for 3d that's painful |
| 09:19:39 | FromGitter | <zacharycarter> but thankfully when you buy assets off the asset stores they usually come with them |
| 09:20:38 | PMunch | Hmm, I should really clean up my browser session. It's starting to chug pretty heavily now.. |
| 09:21:10 | PMunch | Well that would mean doing 3D model work, which I find way harder than 2D art |
| 09:22:31 | PMunch | 320+ tabs doesn't seem to be a supported scenario in Vivaldi |
| 09:23:12 | FromGitter | <zacharycarter> :D |
| 09:28:15 | PMunch | There, down to 20 tabs :P |
| 09:28:40 | FromGitter | <zacharycarter> this is probably a noob question - but is there any way to serialize a browser session and send it via email or something? |
| 09:28:50 | FromGitter | <zacharycarter> so I can just click a link and have chrome open with all of the same tabs? |
| 09:29:29 | PMunch | Hmm, I don't think so |
| 09:29:35 | PMunch | Not in a cross-browser way at least |
| 09:29:45 | FromGitter | <zacharycarter> would be sweet |
| 09:30:46 | PMunch | Yeah I really want to have a more session based browsing experience |
| 09:30:53 | PMunch | With less focus on clicking everywhere.. |
| 09:37:47 | Zevv | tridactyl! |
| 09:37:49 | Zevv | on firefox |
| 09:44:53 | lqdev[m] | gosh, how can you even have >10 tabs open at a time |
| 09:45:11 | Zevv | always makes me wonder |
| 09:45:33 | Zevv | my wife does that - I can't stand to even look in the general direction of her desktop |
| 09:45:39 | lqdev[m] | I'd go crazy if I had any more than that |
| 09:45:47 | FromGitter | <Vindaar> imo because bookmarks in browser suck |
| 09:45:54 | lqdev[m] | it'd trigger my natural garbage collection cycle ;) |
| 09:47:20 | PMunch | Yeah I don't use bookmarks |
| 09:47:27 | PMunch | I just leave tabs I might want to check out later open |
| 09:47:35 | FromGitter | <Vindaar> yep |
| 09:47:49 | PMunch | And then use the search across open tabs feature to find what I need |
| 09:48:05 | Zevv | "search across oven tabs" - does that even exist, wow |
| 09:48:15 | Zevv | s/opebn |
| 09:48:28 | Zevv | raaagh, nevermind |
| 09:48:54 | PMunch | Well kinda |
| 09:49:18 | PMunch | It's a "Quick panel" that searches history, open tabs, or falls through to Google |
| 09:51:05 | PMunch | It's pretty dumb though, doesn't default to highlighting open tabs, so I have to use the arrows keys to get to the open tabs instead of searching google.. |
| 10:07:39 | * | solitudesf quit (Remote host closed the connection) |
| 10:08:08 | * | solitudesf joined #nim |
| 10:28:57 | * | theelous3_ quit (Ping timeout: 240 seconds) |
| 10:38:17 | FromDiscord | <Kurolox> Hello! |
| 10:39:50 | FromDiscord | <Kurolox> So I'm picking Nim. It's a language that I find fascinating, but I have some questions about it |
| 10:40:18 | FromDiscord | <Rika> hi, ask away |
| 10:41:10 | FromDiscord | <Kurolox> I'm following `Nim by Example` since I already have some experience programming, and it says that some data types have to be initialized beforehand (like floats) or the compiler won't work |
| 10:41:22 | FromDiscord | <Kurolox> so you can't do something like `var a = float` |
| 10:41:43 | FromDiscord | <Kurolox> right? |
| 10:41:47 | FromDiscord | <Rika> its more of `var a: float` |
| 10:41:57 | FromDiscord | <Rika> !eval var a: float; a = 2.0 |
| 10:42:00 | NimBot | <no output> |
| 10:42:05 | FromDiscord | <Rika> it compiles |
| 10:42:10 | FromDiscord | <Kurolox> hm |
| 10:42:17 | FromDiscord | <Rika> you can do `var a: float` then set it in the future |
| 10:43:03 | FromDiscord | <Rika> `(const/var/let) {identifier}[: {type name}][= {value}]` |
| 10:43:19 | FromDiscord | <Kurolox> Oh, I see what was my issue. It was using `let` instead of `var` in the example, so since it wasn't mutable the compiler was complaining |
| 10:43:28 | FromDiscord | <Kurolox> I'm quite dumb, I'm sorry |
| 10:43:39 | FromDiscord | <Rika> also you put `var a = float` instead of `var a: float` |
| 10:43:40 | FromDiscord | <Rika> it's fine |
| 10:43:47 | FromDiscord | <Rika> it's different to most languages |
| 10:43:57 | PMunch | Nim will also auto-initialise all values to binary 0 |
| 10:44:08 | PMunch | !eval var x: float; echo x |
| 10:44:11 | NimBot | 0.0 |
| 10:44:18 | FromDiscord | <Rika> C-likes use `{type} {ident}` so like `float a` instead of what nim does |
| 10:44:22 | PMunch | So even if it isn't given a value it will be 0.0 |
| 10:44:39 | FromDiscord | <Kurolox> That's really neat |
| 10:44:47 | PMunch | Strings will be empty strings, sequences empty sequences, and I believe Tables are now also treated as empty tables. |
| 10:44:53 | FromDiscord | <Rika> wish `nil` goes away in the future 😛 |
| 10:45:10 | PMunch | Rika, it's already mostly gone if you don't work with C interop |
| 10:45:20 | FromDiscord | <Rika> i mean, i guess haha |
| 10:45:21 | FromDiscord | <Kurolox> I'm also quite fascinated by how procedures work in nim, I haven't seen anything like that before but it makes a lot of sense |
| 10:45:39 | FromDiscord | <Rika> procedures are just what most people call functions, but named properly |
| 10:45:50 | FromDiscord | <Rika> what's amazing? |
| 10:45:53 | FromDiscord | <Rika> im not sure i see it |
| 10:46:07 | FromDiscord | <Kurolox> the fact that they have an implied `result` variable already declared that's returned automatically |
| 10:46:42 | PMunch | Oh right, yeah that is borrowed from Pascal I believe |
| 10:46:44 | FromDiscord | <Rika> i think that exists in golang too 😛 |
| 10:47:01 | FromDiscord | <Kurolox> I'm not familiar with golang unfortunately, most of my experience comes from Python, JS and Java |
| 10:47:03 | PMunch | If you want more implicit return goodness: https://peterme.net/tips-and-tricks-with-implicit-return-in-nim.html |
| 10:47:19 | PMunch | Oh well, off to lunch |
| 10:47:30 | FromDiscord | <Rika> also, last expression returns if its "not used" and matches type |
| 10:47:53 | FromDiscord | <Rika> oooh python, do you know python's type hints |
| 10:47:57 | Araq | Rika: there is an RFC and a prototype implementation for getting rid of 'nil' |
| 10:48:01 | FromDiscord | <Rika> it's the same syntax in nim |
| 10:48:03 | * | solitudesf quit (Ping timeout: 250 seconds) |
| 10:48:17 | FromDiscord | <Rika> nice |
| 10:48:46 | FromDiscord | <Kurolox> yeah, I know them. They're somewhat useful, but I'm sad that you have to use some really speficic linters that "enforce" them |
| 10:49:02 | * | crem joined #nim |
| 10:49:09 | FromDiscord | <Rika> well yeah its the same syntax in nim, just with var before the varname |
| 10:49:16 | FromDiscord | <Rika> or let |
| 10:49:36 | FromDiscord | <Kurolox> I guess this isn't a nim specific thing, but more of a general programming question |
| 10:49:46 | FromDiscord | <Rika> ~~also the enforcing issue is why i moved to nim lmaooo~~ |
| 10:49:51 | FromDiscord | <Kurolox> what's the difference between an inmutable data type and a constant? |
| 10:49:58 | FromDiscord | <Kurolox> both appear to be functionally the same |
| 10:50:07 | FromDiscord | <Rika> which, let vs const> |
| 10:50:09 | FromDiscord | <Rika> ?* |
| 10:50:11 | FromDiscord | <Kurolox> yes |
| 10:50:22 | FromDiscord | <Rika> let is runtime, const is compiletime |
| 10:51:10 | FromDiscord | <Rika> if the variable is a `ref`, you can change its contents if it's in a `let` because the reference is made immutable, not the referenced data |
| 10:51:14 | FromDiscord | <Rika> (i think) |
| 10:52:22 | FromDiscord | <Rika> `const`s are evaluated in compiletime, then its result replaces all references to the variable (i think x2) |
| 10:53:07 | FromDiscord | <Rika> so `const a = 20; a * a` becomes `20 * 20` if my mind is performing correctly |
| 10:53:25 | FromDiscord | <Rika> whilst `let a = 20; a * a` doesnt change |
| 10:53:38 | FromDiscord | <Rika> there are better examples out there... |
| 10:54:00 | FromDiscord | <Rika> ooh ooh heres one more |
| 10:54:20 | FromDiscord | <Rika> `const` has to be computable at compilation, so no variables defined in runtime |
| 10:54:45 | FromDiscord | <Kurolox> I see. Thanks for the help! |
| 10:55:11 | FromDiscord | <Kurolox> Another thing I've never seen are nested multiline comments |
| 10:57:00 | FromDiscord | <Rika> well since the multiline begin and multiline end tokens look different, its possible to do |
| 10:57:07 | FromDiscord | <Rika> though i dont see the reason to do so haha |
| 10:58:15 | FromDiscord | <Kurolox> I mean, technically the same can be said for Java, but nested multiline comments won't work |
| 10:58:26 | FromDiscord | <Rika> hm? why not? |
| 10:58:30 | FromDiscord | <Kurolox> ```java |
| 10:58:31 | FromDiscord | <Kurolox> /* /* test */ */``` will fail |
| 10:58:37 | FromDiscord | <Rika> oh i see |
| 10:58:41 | FromDiscord | <Rika> now i see the reason to do so |
| 10:58:56 | FromDiscord | <Kurolox> hm, now I wonder |
| 10:59:04 | FromDiscord | <Kurolox> ```nim |
| 10:59:04 | FromDiscord | <Kurolox> proc getAlphabet(): string = |
| 10:59:04 | FromDiscord | <Kurolox> for letter in 'a'..'z': |
| 10:59:05 | FromDiscord | <Kurolox> result.add(letter)``` |
| 10:59:16 | FromDiscord | <Kurolox> there's syntax highlighting support for nim, that's neat |
| 10:59:24 | FromDiscord | <Rika> RIP IRC people |
| 10:59:29 | FromDiscord | <Kurolox> oh, sorry |
| 10:59:31 | * | letto quit (Quit: Konversation terminated!) |
| 10:59:38 | FromDiscord | <Rika> its recommended not to do multiline code inserts |
| 10:59:52 | FromDiscord | <Rika> at least, not here |
| 10:59:53 | FromDiscord | <Kurolox> Yeah, I just wanted to see if there was highlighting support for nim |
| 11:00:06 | FromDiscord | <kodkuce> i head to sorry for posting code from discord like 20times, alwes froget |
| 11:00:09 | FromDiscord | <Kurolox> should've done it outside of here, I guess |
| 11:00:14 | FromDiscord | <Rika> there's #discord-only |
| 11:00:15 | FromDiscord | <Rika> haha |
| 11:00:51 | FromDiscord | <Kurolox> Well, I want to end up making something in Nim |
| 11:01:07 | FromDiscord | <Kurolox> probably an API REST, I'll try to get comfortable with the language first |
| 11:01:12 | FromDiscord | <Rika> i've made a (pretty inefficient) image dithering program in Nim haha |
| 11:01:14 | * | letto joined #nim |
| 11:01:15 | narimiran | @Kurolox for questions like "what's the difference between an inmutable data type and a constant?" i recommend you https://narimiran.github.io/nim-basics/ |
| 11:01:39 | narimiran | e.g. see the answer here: https://narimiran.github.io/nim-basics/#_immutable_assignment |
| 11:02:08 | FromDiscord | <Rika> narimiran book when 😛 |
| 11:03:19 | narimiran | @Rika so you want to pay me for the stuff that you can read now freely? ok, no problem, i can send you an invoice |
| 11:03:32 | FromDiscord | <Rika> ;; no i was joking ;; |
| 11:03:35 | FromDiscord | <Kurolox> I mean, I'd pay for a physical release |
| 11:04:02 | FromDiscord | <Kurolox> I'll grab a copy of Nim in Action in the future |
| 11:04:14 | narimiran | @Kurolox there's a pdf and epub version, if that'll suit you better than .html? |
| 11:04:34 | narimiran | (note to self: make epub version more visible) |
| 11:04:50 | FromDiscord | <Kurolox> It's more about supporting the author and having a physical book to get in my shelves rather than the book content itself |
| 11:05:02 | FromDiscord | <Kurolox> do you have any donation address? |
| 11:05:13 | narimiran | i don't |
| 11:05:31 | FromDiscord | <kodkuce> nice work narimiran , it looks nice and easyer then default turtorial at least for me at first glance, hope you fill it more |
| 11:05:46 | narimiran | and i'm too much of an ecologist to go for a physical book :) |
| 11:05:57 | FromDiscord | <kodkuce> buy clearink 2020 |
| 11:06:01 | FromDiscord | <Kurolox> that's fair |
| 11:06:18 | FromDiscord | <Kurolox> if you end up being open for donations let me know, I always try to support this kind of stuff |
| 11:06:26 | narimiran | @kodkuce: more content would probably go into an "intermediate nim" tutorial, which is my long-term goal with no clear deadlines |
| 11:06:31 | FromDiscord | <Kurolox> In fact I just set up the bountysource for nim like an hour ago |
| 11:07:02 | FromDiscord | <Kurolox> and ordered some nim stickers too because my laptop needs it |
| 11:07:46 | FromDiscord | <Rika> whats bountysource |
| 11:07:56 | FromDiscord | <Rika> ratger |
| 11:07:58 | FromDiscord | <Rika> whats it for |
| 11:08:13 | FromDiscord | <Rika> whats the money gonna go into |
| 11:08:25 | FromDiscord | <kodkuce> lol i thinked first this real github link, thinked you self hosting gitea or something |
| 11:09:04 | * | Guest11723 quit (Quit: Leaving) |
| 11:09:04 | * | Kevin4 quit (Quit: Leaving) |
| 11:11:20 | FromDiscord | <Kurolox> well, as far as I understand the money goes to dom96 and Araq, and they use the money to pay for hosting for the nim website, and to place bounties on Nim priority bugs to incentive people to fix them |
| 11:11:38 | narimiran | dom doesn't work for Nim anymore |
| 11:11:56 | FromDiscord | <Kurolox> oh well, he's still listed in the bountysource page |
| 11:12:02 | FromDiscord | <Kurolox> maybe that should be fixed then |
| 11:12:20 | narimiran | he's still involved with nim, make no mistake |
| 11:12:59 | FromDiscord | <Kurolox> Dom96 is dominik, right? |
| 11:13:07 | narimiran | yep |
| 11:13:22 | FromDiscord | <Kurolox> I'm fine with it then. I hope I can grab his book soon |
| 11:13:29 | FromDiscord | <Kurolox> it looks like an interesting read |
| 11:15:07 | * | krux02 joined #nim |
| 11:15:31 | FromDiscord | <kodkuce> does anyone officaly work for Nim? |
| 11:15:39 | narimiran | yes, me :) |
| 11:15:40 | FromDiscord | <kodkuce> like a payed job? |
| 11:15:52 | FromDiscord | <Rika> a fulltime? |
| 11:16:10 | FromDiscord | <kodkuce> dident even know Nim head some sponsoring |
| 11:16:20 | FromDiscord | <Rika> nim is sponsored??? |
| 11:16:35 | FromDiscord | <kodkuce> i know only Signal i think gived some money |
| 11:16:54 | narimiran | status, not signal: https://nim-lang.org/blog/2018/08/07/nim-partners-with-status.html |
| 11:17:04 | FromDiscord | <kodkuce> ye my bad |
| 11:17:18 | FromDiscord | <kodkuce> yep there is on https://nim-lang.org/ on bot page 2 psonsore allready |
| 11:17:41 | FromGitter | <mratsim> @Kurolox, AFAIK hosting is kindly offered by Digital Ocean |
| 11:20:46 | FromGitter | <mratsim> Oh, I totally forgot but the sponsoring has been going on for more than a year now, you should definitely prepare a blog post on how being 3 instead of 1 core compiler developer helped ;) |
| 11:21:14 | * | tim17d joined #nim |
| 11:23:28 | FromDiscord | <Kurolox> Can anyone tell me why this code snippet isn't working? |
| 11:23:39 | FromDiscord | <Kurolox> https://ghostbin.co/paste/yox4u |
| 11:24:10 | FromDiscord | <Kurolox> it should be straightforward and it's even copied from Nim by Example, but it's giving me a "SIGSEGV: Illegal storage access" error |
| 11:25:50 | FromDiscord | <Rika> https://play.nim-lang.org/#ix=1Uh3 it works here |
| 11:25:59 | narimiran | @Kurolox are you sure you're on a recent version of nim? |
| 11:26:00 | FromDiscord | <Rika> are you sure its this proc thats the issue |
| 11:26:12 | narimiran | what does `nim -v` print for you? |
| 11:26:13 | lqdev[m] | pre-0.19 |
| 11:26:13 | lqdev[m] | @Kurolox you probably have an old version of Nim |
| 11:27:15 | FromDiscord | <Kurolox> Nim Compiler Version 0.18.0 [Linux: amd64] Copyright (c) 2006-2018 by Andreas Rumpf |
| 11:27:16 | * | solitudesf joined #nim |
| 11:27:18 | narimiran | wild guess: are you on Arch Linux and have used pacman to install Nim? |
| 11:27:25 | FromDiscord | <Kurolox> Spot on |
| 11:27:35 | narimiran | arch "bleeding edge" linux |
| 11:27:53 | lqdev[m] | use choosenim to install Nim |
| 11:28:01 | narimiran | v0.18 was released on 1st March 2018 |
| 11:28:11 | FromDiscord | <Kurolox> that doesn't sound right |
| 11:28:24 | FromDiscord | <Kurolox> as in, I can't really believe it hasn't been updated since then |
| 11:28:31 | narimiran | what doesn't sound right? that arch has severely outdated package? |
| 11:28:45 | narimiran | i try to notify them, but they either ignored me or ridiculed me |
| 11:28:45 | * | Hideki_ joined #nim |
| 11:28:50 | lqdev[m] | as narimiran said, arch "bleeding edge" linux |
| 11:28:53 | FromDiscord | <Kurolox> yeah, but I'm not aware of how main repository package management works |
| 11:29:02 | narimiran | we had 3 big releases since them, but they just don't care |
| 11:29:07 | lqdev[m] | narimiran: ridiculed? |
| 11:29:09 | narimiran | btw, manjaro has the latest nim |
| 11:29:17 | FromDiscord | <Kurolox> man, that sucks |
| 11:29:23 | narimiran | but MANJARO IS NOT ARCH!1!!!1111!!1 |
| 11:29:51 | federico3 | https://repology.org/project/nim/versions |
| 11:30:04 | FromDiscord | <Kurolox> installing choosenim then |
| 11:30:07 | narimiran | lqdev[m]: yeah, so i decided to not care anymore about arch and their teenage users |
| 11:30:09 | lqdev[m] | ^ lol |
| 11:30:27 | lqdev[m] | damn |
| 11:30:45 | Araq | so hard to resist... |
| 11:30:50 | narimiran | :D :D |
| 11:31:07 | FromDiscord | <Kurolox> I'm sorry about that, it's true that part of the arch community is like that |
| 11:31:44 | lqdev[m] | anyways, @Kurolox choosenim is the official way of installing Nim. try that. you can find it in the download section on nim-lang.org |
| 11:31:53 | lqdev[m] | or install, I don't remember |
| 11:32:03 | narimiran | but hey, you can use AUR, right? there you go, nim 0.19.4, fresh from february |
| 11:32:26 | FromDiscord | <Kurolox> pretty much |
| 11:32:44 | FromDiscord | <Kurolox> choosenim from the AUR doesn't even work, so I'm just installing it manually |
| 11:33:39 | FromDiscord | <Kurolox> this may be a good oportunity to learn about AUR packages, though. I may try to create the nim and choosenim packages later and try to keep them updated myself |
| 11:34:22 | federico3 | *sigh* |
| 11:38:55 | narimiran | Araq: https://i.imgflip.com/3g6z16.jpg :) |
| 11:40:39 | FromDiscord | <Kurolox> heh |
| 11:41:09 | FromDiscord | <Kurolox> Anyways, updating did fix the issue. I wonder if it was a bug or I was using a non-implemented feature |
| 11:41:36 | narimiran | the latter |
| 11:43:14 | narimiran | @Kurolox once again, Nim Basics to the rescue: https://narimiran.github.io/nim-basics/#_result_variable (see the big orange ! triangle) |
| 11:44:42 | FromDiscord | <Kurolox> Oh, I see. so it was `nil` rather than an empty string, so trying to append to said string would error out |
| 11:45:41 | narimiran | yeah |
| 11:47:37 | FromDiscord | <Kurolox> Now the error message makes way more sense than before |
| 11:49:53 | * | tklohna quit (Ping timeout: 268 seconds) |
| 11:50:45 | FromDiscord | <++x;> 😊 😊 👍 👍 💦 💦 |
| 11:51:41 | FromDiscord | <demotomohiro> Gentoo linux provides latest Nim. |
| 11:51:41 | FromDiscord | <demotomohiro> https://packages.gentoo.org/packages/dev-lang/nim |
| 11:51:48 | FromDiscord | <++x;> Are the discord users in this server becoming more active? |
| 11:52:06 | federico3 | demotomohiro: I already shared the link to repology |
| 11:52:09 | Araq | narimiran, lol |
| 11:53:30 | FromDiscord | <Kurolox> I've been asking in the arch linux IRC about this |
| 11:53:48 | FromDiscord | <Kurolox> the TL;DR is: "Blame the maintainer, and just don't use the repository" |
| 11:53:53 | * | tim17d left #nim (#nim) |
| 11:54:42 | Zevv | fair enough :) |
| 11:54:59 | narimiran | "we have the latest packages, in under an hour after they're released, and we're very proud of that" |
| 11:55:15 | narimiran | "what about this 1.5 year old outdated one? lalalalalalala, don't hear you" |
| 11:56:37 | FromDiscord | <Kurolox> to be honest the system sounds terribly made |
| 11:56:53 | FromDiscord | <Kurolox> as far as I understand, someone can flag the package out of date, and the maintainer gets a notification in his mail |
| 11:57:17 | FromDiscord | <Kurolox> but a package can only be flagged once, and if the maintainer misses that mail then nobody can report the package out of date again ever |
| 11:57:22 | Zevv | narimiran: joking aside, have you contacted the maintainer? |
| 11:57:31 | narimiran | Zevv: yep, that too |
| 11:58:01 | narimiran | i didn't raise the issue publicly until i waited for more than one month for his reply. (which i never got) |
| 11:58:08 | * | rockcavera joined #nim |
| 11:59:56 | FromDiscord | <Kurolox> to be honest that maintainer looks busy |
| 12:00:01 | FromDiscord | <Kurolox> as in, he's maintaining about 800 packages |
| 12:00:14 | FromDiscord | <Kurolox> I've sent him a mail too, maybe he'll answer |
| 12:00:28 | narimiran | maybe, just maybe, he might give part of those 800 to somebody else |
| 12:01:49 | FromDiscord | <Kurolox> Okay, it seems like someone in the IRC got direct contact with the maintainer and it's asking him about this directly |
| 12:03:37 | * | vsantana joined #nim |
| 12:06:54 | * | nc-x joined #nim |
| 12:07:24 | nc-x | Found this in the rust subreddit. haven't watched/read it yet: https://aardappel.github.io/lobster/memory_management.html https://www.youtube.com/watch?v=WUkYIdv9B8c |
| 12:07:58 | * | nc-x quit (Remote host closed the connection) |
| 12:10:10 | FromDiscord | <Kurolox> quick question |
| 12:10:20 | FromDiscord | <Kurolox> is csources still used in the latest versions? |
| 12:10:40 | FromDiscord | <Kurolox> since it has been archived and the package seemed to need it to build |
| 12:10:56 | narimiran | yes, it is archived so no new stuff will be pushed into it |
| 12:11:11 | narimiran | but it is used |
| 12:11:25 | narimiran | "This repository is archived because it's frozen, HEAD of csources can build Nim version 1 and any later version." |
| 12:13:35 | tane_ | is there a way to get a sequence like view into an array such that one can apply openArray[T] functions to a subrange of an array? |
| 12:14:01 | * | Hideki_ quit (Remote host closed the connection) |
| 12:14:14 | narimiran | tane_: `toOpenArray`? |
| 12:14:51 | * | NimBot joined #nim |
| 12:14:56 | tane_ | narimiran, looks promising, thanks |
| 12:15:26 | tane_ | perfect :) |
| 12:15:47 | * | Hideki_ joined #nim |
| 12:17:42 | narimiran | tane_: yep, i've just created a dumb example https://play.nim-lang.org/#ix=1UhR |
| 12:19:53 | * | Hideki_ quit (Ping timeout: 245 seconds) |
| 12:20:36 | tane_ | narimiran, yeah, nice |
| 12:26:11 | * | Hideki_ joined #nim |
| 12:27:26 | * | Hideki_ quit (Remote host closed the connection) |
| 12:27:41 | FromDiscord | <Kurolox> I wanted to learn nim and I've ended up learning to make PKGBUILDs and package repositories, huh |
| 12:28:25 | * | Hideki_ joined #nim |
| 12:28:28 | * | Hideki_ quit (Remote host closed the connection) |
| 12:28:45 | * | Hideki_ joined #nim |
| 12:36:01 | FromDiscord | <Kurolox> by the way, does anyone know if there's a lot of outdated stuff in the "nim in action" book? |
| 12:36:15 | FromDiscord | <Kurolox> since it was released two years ago, I'm not sure how much the language could've changed since then |
| 12:39:08 | narimiran | NiA is continuously tested with the latest nim versions, so it is not outdated. it is still worth the read |
| 12:39:44 | narimiran | but maybe we can push dom96 to release a second edition :) |
| 12:41:56 | FromDiscord | <Kurolox> call it Nim in Reaction |
| 12:41:59 | FromDiscord | <Kurolox> I'd buy it |
| 12:43:12 | Zevv | nah I'm not giving my money to dom again, he's spending it all on booze I heard |
| 12:44:13 | Zevv | kurolox: I believe the examples from doms book are part of the nim test suite to make sure the stuff from the book is kept in workig order |
| 12:44:39 | Zevv | so the book might not tell the latest stories, but what's in there is still relevant |
| 12:45:17 | FromDiscord | <Kurolox> that feels weird though, doesn't that mean that maybe there are some breaking changes you won't be able to make in the future due to keeping the book compatible? |
| 12:45:24 | FromDiscord | <Kurolox> that feels weird though, doesn't that mean that maybe there will be some breaking changes you won't be able to make in the future due to keeping the book compatible? |
| 12:46:10 | Zevv | at some time that promise will probably be broken, but I believe that at least for 1.0 the goal was to keep it relevant |
| 12:48:52 | Araq | well... |
| 12:48:53 | solitudesf | there were breaking changes, but they were introduced with compiler flags that allow for backwards compatibility |
| 12:49:39 | Araq | yeah and devel broke more examples by removing the unary < operator |
| 12:49:54 | Araq | we updated the tests :P |
| 12:53:25 | FromGitter | <Willyboar> @Kurolox there is a guy tracking and fixes errors in the book |
| 12:54:19 | FromGitter | <Willyboar> https://deepakg.github.io/nim/2019/09/28/nim-in-action-errata.html |
| 12:56:50 | Zevv | oh there is more in there then I expected, but it isnt too bad :) |
| 12:57:26 | FromGitter | <Willyboar> actually is very helpfull |
| 12:58:00 | FromDiscord | <Kurolox> I love the nim community |
| 12:58:06 | FromDiscord | <Kurolox> so many helpful resources coming from everywhere |
| 12:59:11 | narimiran | "future is deprecated" :) |
| 12:59:28 | Zevv | the future is overrated |
| 12:59:52 | FromDiscord | <kodkuce> the future is none |
| 13:02:31 | * | GordonBGood_ joined #nim |
| 13:02:32 | * | GordonBGood quit (Read error: Connection reset by peer) |
| 13:03:28 | FromDiscord | <Kurolox> the future is `nil` |
| 13:07:51 | dom96 | That article is amazing. Massive thanks to the person that compiled it. |
| 13:07:51 | dom96 | That article is amazing. Massive thanks to the person that compiled it. |
| 13:08:16 | FromDiscord | <mratsim> @Kurolox actually there is Nim 0.19 in Arch community-testing |
| 13:08:29 | FromDiscord | <mratsim> but this was also flagged out-of-date, only 1 year old though |
| 13:08:32 | dom96 | Note that all examples will compile, the only problems you might run into are deprecations warnings, which IMO aren't all that problematic when learning since Nim will tell you what to use instead |
| 13:08:53 | dom96 | And indeed from what I can see that article only lists deprecation warnings :D |
| 13:11:00 | * | endragor quit (Remote host closed the connection) |
| 13:12:45 | dom96 | (one exception is that `nil` is no longer possible for strings, but that only breaks one example and you get a pretty clear error message) |
| 13:13:19 | dom96 | So thank you to Araq et al. for not breaking my book (too much) :) |
| 13:13:45 | * | tklohna joined #nim |
| 13:16:01 | narimiran | dom96: `.. <` will also break in the latest nim ;) |
| 13:16:16 | narimiran | but it is written in that article how to fix it |
| 13:16:20 | FromGitter | <mratsim> that broke since a year or more though |
| 13:16:20 | livcd | that broke even when I bought the book almost a year ago |
| 13:16:37 | narimiran | livcd: then was deprecated, now is broken |
| 13:16:40 | livcd | see lots of unhappy customers |
| 13:17:13 | dom96 | narimiran you mean in 2.0? |
| 13:17:38 | narimiran | i mean in nim 1.1.1 (devel), so it will break in 1.2 when that becomes new stable |
| 13:17:56 | dom96 | But that's a breaking change, how can it be a part of 1.x :P |
| 13:18:13 | narimiran | dom96: it's the only way we can make you release 2nd edition :P |
| 13:18:43 | dom96 | Okay, it was deprecated so it's probably fair enough. But we need to get together and specify clearly that deprecated things are not covered under the promise not to break things. |
| 13:19:01 | narimiran | imagine all those people not buying NiA currently because they think it is outdated. $$$ |
| 13:19:07 | FromGitter | <mratsim> not to break things is for 1.0.X |
| 13:19:17 | dom96 | No it isn't |
| 13:19:20 | FromGitter | <mratsim> 1) 1.0 can change |
| 13:19:24 | * | narimiran grabs popcorn |
| 13:19:31 | * | Hideki_ quit (Remote host closed the connection) |
| 13:19:32 | dom96 | Are we really going to have this discussion again? |
| 13:19:57 | FromGitter | <mratsim> "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." - Saint Exupéry |
| 13:20:08 | dom96 | I'm going to write a post to outline what the expectations are. Read up on semver please. |
| 13:20:09 | narimiran | removing long-deprecated stuff is not breakage. |
| 13:20:15 | * | Hideki_ joined #nim |
| 13:20:26 | livcd | I only care if important libs break |
| 13:20:29 | FromGitter | <mratsim> I don't care about semver, I care about having a good language and a good standard library |
| 13:20:39 | FromGitter | <mratsim> and obviously good transitions |
| 13:21:03 | * | clyybber joined #nim |
| 13:21:09 | FromGitter | <mratsim> we have a lot of API debts in Nim |
| 13:21:10 | narimiran | @mratsim speaking of which: can we remove the outdated inplace `sequtils.map` or does arraymancer still depend on it? :P |
| 13:21:13 | FromGitter | <mratsim> we need to pay it |
| 13:21:40 | FromGitter | <mratsim> I never depended on it |
| 13:21:51 | FromGitter | <mratsim> all my maps are multithreaded |
| 13:22:00 | FromGitter | <mratsim> except for strings |
| 13:22:06 | dom96 | I'm fine with removing deprecated APIs |
| 13:22:07 | dom96 | I'm not fine with breaking the language or existing APIs in 1.x |
| 13:22:07 | narimiran | eeerrrrm, i remember re-introducing it because arraymancer was broken without it |
| 13:22:08 | dom96 | Why am I not fine? Because that's what we promised in our release article. |
| 13:22:14 | Araq | dom96, actually I'm considering to add back unary '<' for 1.2 |
| 13:22:30 | dom96 | And why's that? |
| 13:22:38 | Araq | because it's special |
| 13:22:47 | narimiran | @mratsim i'll make a PR with that `map` removed, maybe something changed in the mean time. if not, i'll ping you |
| 13:22:55 | Araq | I removed a bunch of deprecated stuff, '<' was among it |
| 13:23:06 | FromGitter | <mratsim> no problem @narimiran |
| 13:23:10 | Araq | but it turned out only '<' was still widely used |
| 13:23:20 | Araq | so we might add it back for a little longer |
| 13:23:45 | narimiran | (oh, btw, it was `mapIt`, not `map`) |
| 13:24:07 | Araq | ping clyybber |
| 13:24:17 | FromGitter | <mratsim> (https://files.gitter.im/nim-lang/Nim/R4AE/j3ea2qy3h8521.jpg.webp) |
| 13:24:19 | dom96 | cool. So there was no need to scare potential Nim in Action customers, as there? :) |
| 13:24:23 | dom96 | *was |
| 13:24:29 | * | uu91 quit (Ping timeout: 276 seconds) |
| 13:24:38 | FromGitter | <Willyboar> *Willyboar grabs narimiran popcorn |
| 13:24:51 | * | Hideki_ quit (Ping timeout: 268 seconds) |
| 13:25:20 | clyybber | pong Araq |
| 13:25:58 | Araq | why is it that nkClosure is *not* processed at all by injectdestructors.nim? |
| 13:26:12 | Araq | any ideas? it's frustrating |
| 13:27:07 | clyybber | Araq: Hmm, I have no idea, but I didn't change it during my refactoring. I saw you fixed it in your PR? |
| 13:27:13 | clyybber | Or is your fix not working? |
| 13:27:21 | Araq | not working |
| 13:27:28 | Araq | but at least conceptually clean now :P |
| 13:27:54 | Araq | I feel like we are 3 bugs away from --gc:arc working on async code |
| 13:28:21 | Araq | will reduce sleep now and won't eat until it's done |
| 13:28:38 | clyybber | good idea |
| 13:29:04 | * | tklohna quit (Ping timeout: 252 seconds) |
| 13:29:17 | Araq | also I still wonder whether what we do is sound |
| 13:29:35 | clyybber | <insert mratsims pic here> |
| 13:29:50 | * | Hideki_ joined #nim |
| 13:30:06 | clyybber | Doesn't really fit here, though, since we cut a lot of debt |
| 13:30:11 | FromGitter | <Vindaar> good thing to hear you guys have the same problems I do, just on more complicated stuff :P |
| 13:30:38 | FromGitter | <mratsim> Well that's because you are doing a PhD :p |
| 13:30:55 | clyybber | Araq: Are there no closures at all that go through p/pArg ? |
| 13:31:16 | clyybber | Vindaar: What are you doing an PhD on? |
| 13:31:17 | FromGitter | <mratsim> professional bump your head against the wall "why am I alone doing this" |
| 13:31:32 | Araq | I've only seen nkClosure(..., nkNilLit) |
| 13:31:34 | FromGitter | <mratsim> Nim is the same: no stack overflow with hundreds of people that had the same issue before |
| 13:32:08 | Araq | which are exactly the closures that we don't care about |
| 13:32:22 | Araq | all the important closures are not even processed... so weird |
| 13:32:36 | clyybber | Huh, that is indeed... weird |
| 13:33:38 | Araq | ah but the nkObjConstrs are looked at |
| 13:33:59 | * | tane_ quit (Remote host closed the connection) |
| 13:34:09 | FromGitter | <Vindaar> @mratsim well my comment wasn't mainly about my PhD though, because hey, my PhD is complicated too :D ⏎ @clybber doing a PhD in physics. search for axions (dark matter candidates) with CAST, a small experiment at CERN (sitting there right now) |
| 13:34:17 | * | tane joined #nim |
| 13:36:30 | clyybber | Araq: Why wouldn't they be looked at? |
| 13:36:46 | clyybber | Or are they related to closures in some way I am not aware of? |
| 13:37:00 | * | couven92 quit (Quit: Client disconnecting) |
| 13:37:35 | * | uu91 joined #nim |
| 13:37:39 | Araq | looks like some nkHiddenSubConv bullshit |
| 13:38:33 | FromGitter | <Vindaar> I always miss that second 'y' clyybber :P |
| 13:39:14 | clyybber | Vindaar: Oh, yeah, didin't notice you tagged me :D |
| 13:41:10 | clyybber | Vindaar: Oh, nice in CERN. So are you working with the accelerator? |
| 13:41:42 | clyybber | oh, well. I can't read apparently (don't mind me :D) |
| 13:42:46 | tane | clyybber, are you a phd student too? :) |
| 13:43:08 | clyybber | nope |
| 13:43:50 | clyybber | started doing a CS bachelor |
| 13:44:37 | tane | \o/, CS is great :P |
| 13:45:13 | FromGitter | <Vindaar> @clyybber: nope, nothing to do with the LHC (fortunately imo) except using a prototype dipole magnet of the LHC: https://en.wikipedia.org/wiki/CERN_Axion_Solar_Telescope |
| 13:45:26 | narimiran | clyybber: wait, how old are you? |
| 13:45:52 | clyybber | 19 |
| 13:46:08 | narimiran | sweet jesus!! i though you were some 40+ guy :D |
| 13:46:12 | clyybber | hahaha |
| 13:46:32 | clyybber | well, I acquired some music taste from my parents I guess |
| 13:46:51 | tane | clyybber, for instance? :) |
| 13:47:10 | disruptek | mom, dad, y'know, the usual... |
| 13:47:16 | narimiran | hahahahahaha |
| 13:47:36 | narimiran | disruptek: congrats, you are today's winner |
| 13:47:58 | tane | well, good music doesn't get old |
| 13:48:32 | disruptek | neither does clyybber, apparently. |
| 13:50:32 | * | xet7 quit (Remote host closed the connection) |
| 13:50:41 | clyybber | tane: I know for example that my dad listened to a lot of prodigy while I was still a fetus |
| 13:51:12 | Araq | clyybber, looks like a phase ordering problem |
| 13:51:26 | Araq | it's still a nkLambda when injectdestructors sees it |
| 13:51:36 | * | couven92 joined #nim |
| 13:51:51 | * | xet7 joined #nim |
| 13:52:02 | clyybber | Araq: Hmm, the closure stuff happens in lambdalifting, right? |
| 13:52:36 | tane | clyybber, yeah, good choice |
| 13:52:50 | clyybber | disruptek: I sleep(-1) regularily. |
| 13:53:23 | Araq | right |
| 13:53:37 | Araq | time for a break, see you later |
| 13:53:40 | clyybber | bb |
| 13:55:02 | clyybber | Araq: But lambdalifting is done from transf right? And AFAICT transf is done before injectdestr. |
| 13:55:12 | clyybber | narimiran: Do you know BOC? |
| 13:55:38 | narimiran | blue oyster cult? |
| 13:55:56 | clyybber | yeah |
| 13:56:02 | narimiran | i've never listened them |
| 13:56:14 | narimiran | but i'm into progressive rock |
| 13:56:31 | clyybber | https://www.youtube.com/watch?v=ClQcUyhoxTg |
| 13:57:29 | narimiran | clyybber: yeah, everybody knows that one. and the version with will ferrell, of course |
| 13:57:43 | * | narimiran needs more cowbell |
| 13:57:52 | clyybber | needs more cowble |
| 13:57:56 | clyybber | bell |
| 13:57:58 | clyybber | haha |
| 13:58:33 | narimiran | s/ferrell/walken |
| 13:58:49 | clyybber | narimiran: https://www.youtube.com/watch?v=ayCcCD_OrNM <- not sure if thats progressive rock, but damn I love it |
| 13:58:52 | narimiran | no wait, ferrell is there too |
| 13:59:32 | narimiran | clyybber: here's something for you to both listen and watch: https://www.youtube.com/watch?v=UM-yGcpaY_4 |
| 13:59:51 | narimiran | yep, they decide to completely switch their instruments in the middle of a song |
| 14:00:43 | narimiran | (and i'll be listening to your link) |
| 14:00:57 | clyybber | right at the beginning, from the first synth sound I was expecting oxygene |
| 14:02:40 | Zevv | oooh gentle giant |
| 14:02:48 | Zevv | that will brighten my afternoon! |
| 14:03:25 | narimiran | Zevv: i'm glad to hear that :) |
| 14:03:59 | * | Zevv puts on his bell bottoms |
| 14:04:29 | clyybber | how many keyboards is that, lol |
| 14:04:31 | Zevv | I heard these are coming back, so I'm wearing them provocatively these days so I can later say that was because of me |
| 14:04:51 | clyybber | amazing, 3 guitars and even mroe keyboards |
| 14:06:19 | narimiran | clyybber: wait until they bring out their violins, flutes, double bass, xylophone, etc. :D |
| 14:06:52 | FromGitter | <Lecale> Wow, I think I just gave Nim a really bad rep by opening an issue on stb_image https://github.com/nothings/stb/issues/833 |
| 14:10:41 | clyybber | Lecale: But it turned out to be a stb_image issue? |
| 14:11:49 | FromGitter | <Lecale> general shrug like activity |
| 14:12:27 | FromGitter | <brentp> fwiw, here's a manuscript for a tool written in nim: https://www.biorxiv.org/content/10.1101/839944v1 . it's already pretty widely used. |
| 14:12:29 | * | tklohna joined #nim |
| 14:12:52 | FromGitter | <Lecale> Should be the easiest thing in the world, doesn't work when i do it. Well, I had the same problem with nuclear fusion as well. |
| 14:18:52 | FromGitter | <Lecale> I don't like that you wrote "These data are used", but otherwise that looks totally great Brent. Hope the paper gets published. |
| 14:19:50 | FromGitter | <Vindaar> @Lecale I love it when people are immediately hostile for no reason. Don't worry about it. ⏎ @brentp nice! |
| 14:21:58 | FromGitter | <brentp> heh. I can take it. You have problems with the grammar? easily fixed, I suppose. |
| 14:23:13 | tane | I hope the source is reviewed too, before too much research depends on it :) |
| 14:25:33 | FromGitter | <brentp> link to source is in the abstract. happy to get your code review. it is tested on thousands of families for which we have known relationships to ensure they are correctly recovered. |
| 14:27:00 | tane | nice, I have no domain knowledge so I won't review anything :) |
| 14:27:05 | FromDiscord | <Rika> Lecale, wow that dev is aggressive |
| 14:27:28 | * | nsf quit (Quit: WeeChat 2.6) |
| 14:28:54 | clyybber | brentp: Amazing |
| 14:29:38 | tane | is there a way to reserve on a sequence (apart from newSeqOfCap)? |
| 14:29:56 | clyybber | you mean on an existing one? |
| 14:30:04 | tane | yes |
| 14:30:25 | clyybber | setLen |
| 14:30:50 | * | adeohluwa joined #nim |
| 14:30:53 | tane | that changes the size of the sequence |
| 14:31:02 | tane | I'm looking for a way to change the capacity |
| 14:31:02 | clyybber | Yeah, there is no setCap |
| 14:31:27 | clyybber | Why would you need it anyways? |
| 14:31:50 | FromGitter | <zetashift> @brentp wow great work!! |
| 14:32:00 | tane | sometimes you know an upper bound on the insertions you will perform |
| 14:32:15 | * | adeohluwa quit (Remote host closed the connection) |
| 14:32:17 | clyybber | Then it makes sense to use newSeqOfCap |
| 14:32:28 | clyybber | I meant, "why do you need another way" |
| 14:32:30 | disruptek | you'd have to copy it anyway in order to move it in memory. |
| 14:32:48 | clyybber | oh and that |
| 14:33:04 | tane | but this is hidden from me in case that I just .add stuff |
| 14:33:16 | disruptek | use a linked list instead. |
| 14:33:23 | tane | lol, no |
| 14:34:04 | tane | anyway, the existence of a setCap would have the same reason as the existence of a newSeqOfCap |
| 14:34:17 | disruptek | not really. |
| 14:34:30 | clyybber | tane: But it wouldn't prove more valuable |
| 14:34:42 | tane | disruptek, how so? |
| 14:34:48 | clyybber | Since setting the capacity would just do a newSeqOfCap internally |
| 14:34:56 | tane | clyybber, it might |
| 14:35:18 | disruptek | you can write it yourself and it will perform identically. |
| 14:35:20 | * | Hideki_ quit (Remote host closed the connection) |
| 14:35:28 | disruptek | maybe in writing it, you will discover the problem. |
| 14:35:48 | tane | disruptek, so you will not answer it? :) |
| 14:36:17 | disruptek | newSeqOfCap allocates sufficient memory for the sequence at instantiation, without regard to the size of the sequence. |
| 14:36:24 | * | Hideki_ joined #nim |
| 14:36:41 | tane | disruptek, ok, and? |
| 14:36:41 | disruptek | this allows subsequent additions to not demand a resize of the sequence. |
| 14:36:59 | disruptek | a resize of the sequence involves a new alloc and a copy of the contents to the new location. |
| 14:37:05 | disruptek | this is what you are proposing in setCap. |
| 14:37:17 | tane | well, setCap maybe, but not necessarily reserve |
| 14:37:21 | tane | my error |
| 14:37:29 | tane | let's call it ensureCap |
| 14:37:58 | disruptek | so implement it. |
| 14:38:06 | tane | yes, it will perform better |
| 14:38:19 | disruptek | how do you figure? |
| 14:39:43 | tane | assume a function that takes a sequence (possibly non-empty) and knows it will add N elements, then calling ensureCap(len+N) **might** allocate, depending on the capacity, while constructing a new sequence with newSeqOfCap(len + N) will definitely allocate |
| 14:40:25 | clyybber | tane: Hmm, sounds reasonable. There is an RFC for it https://github.com/nim-lang/RFCs/issues/97 |
| 14:40:29 | tane | then one might say: it's up to the user to ensure enough space if he wants to avoid allocations, but often that's not the approach you'd like |
| 14:40:48 | disruptek | you don't need to perform the alloc yourself, though; you can simply let the sequence grow automatically. |
| 14:40:51 | * | Hideki_ quit (Ping timeout: 246 seconds) |
| 14:41:15 | tane | disruptek, this requires knowledge about the allocation pattern |
| 14:41:26 | tane | if I add 10000 items, I want it to reallocate once, not an arbitrary number |
| 14:41:48 | tane | but I don't want to reallocate, when the capacity is already large enough |
| 14:41:58 | tane | and that's the whole point of "ensureCap", or "reserve" in C++ |
| 14:42:26 | disruptek | it won't realloc, but you're right that you cannot control the number of allocs. |
| 14:42:47 | disruptek | i think this is yak shaving, but take a look at the impl and see if it needs enhancing. |
| 14:42:49 | tane | exactly, and this is the reason an ensureCap function is desirable |
| 14:44:35 | tane | I mean, the RFC is obviously in place, so something along this line will probably be added at some point |
| 14:46:47 | tane | all languages I know have this in their growable list/seq/vector APIs, because it makes sense from a perf POV |
| 14:46:54 | clyybber | it should be pretty easy to implement |
| 14:47:12 | clyybber | I'll try :) |
| 14:47:16 | disruptek | of course; it has to do this anyway. |
| 14:47:36 | Zevv | narimiran: the dude can xylophone |
| 14:49:11 | clyybber | oh hell yeah |
| 14:49:28 | clyybber | hes giving the drummer a break :p |
| 14:50:03 | FromGitter | <alehander92> mypy |
| 14:50:08 | FromGitter | <alehander92> gets better |
| 14:50:14 | FromGitter | <alehander92> it seems it can model most of my nim types now |
| 14:50:40 | FromGitter | <alehander92> trying to annotate my python code again |
| 14:50:55 | FromGitter | <alehander92> i wonder if i should directly generate some annotations from my nim types |
| 14:51:03 | FromGitter | <alehander92> one day, not now |
| 14:51:13 | clyybber | then again, you can convert the code to nim while you're at it :) |
| 14:51:45 | narimiran | Zevv: crazy dudes, all of them |
| 14:57:44 | FromGitter | <alehander92> clyybber |
| 14:57:45 | FromGitter | <alehander92> i cant |
| 14:57:50 | FromGitter | <alehander92> its for gdb plugins |
| 14:57:53 | clyybber | ah |
| 14:58:28 | FromGitter | <alehander92> and its just a lot of code, no time for it now |
| 14:58:40 | FromGitter | <alehander92> but i was getting some embarrasing mistakes |
| 14:59:03 | FromGitter | <alehander92> like unknown variable and stuff and i just realized i have sloppy code there |
| 14:59:06 | FromGitter | <alehander92> and i need a type checker |
| 14:59:40 | FromGitter | <alehander92> the experience is kinda similar to the javascript backend, type some stubs for your third party, slowly annotate the other stuff |
| 15:06:23 | * | Hideki_ joined #nim |
| 15:07:22 | * | PMunch quit (Quit: Leaving) |
| 15:12:29 | narimiran | @mratsim: removing the deprecated `mapIt` seems to cause no trouble this time. merged into devel. |
| 15:12:55 | FromGitter | <mratsim> how do you test arraymancer? |
| 15:13:25 | FromGitter | <mratsim> because last time I look it only compiled arraymancer.nim, but since all procs are generic, none are instantiated |
| 15:13:36 | narimiran | nim c -r src/arraymancer.nim |
| 15:14:02 | FromGitter | <mratsim> but generic procs and templates cannot be tested like that even for just semcheck |
| 15:14:53 | narimiran | any other command we could run, which would test arraymancer better? (and won't take too much time) |
| 15:15:08 | FromGitter | <mratsim> you can try compiling this file: https://github.com/mratsim/Arraymancer/blob/master/tests/tests_cpu.nim . It shouldn't reach Travis 3GB limit anymore |
| 15:15:13 | FromGitter | <mratsim> try on your machine |
| 15:15:18 | narimiran | btw, that `mapIt` was deprecated since..... wait for it.... nim 0.12 |
| 15:15:56 | FromGitter | <mratsim> just compile, don't run. |
| 15:16:07 | FromGitter | <mratsim> unlike stint I don't rely on bit shifting ;) |
| 15:16:18 | narimiran | :) |
| 15:18:00 | narimiran | `nimble install` doesn't install test folder?! |
| 15:18:28 | narimiran | oh well, `git clone` it is |
| 15:18:50 | * | GordonBGood_ quit (Ping timeout: 240 seconds) |
| 15:19:53 | narimiran | @mratsim that test passes on my machine |
| 15:20:10 | FromGitter | <mratsim> check the memory it uses |
| 15:20:34 | FromGitter | <mratsim> I know I had to split it when complex were added because it reached Travis 3GB memory limit |
| 15:20:48 | FromGitter | <mratsim> but when I checked a month ago, I wasn't anymore (on Nim 1x) |
| 15:21:36 | FromGitter | <mratsim> see: https://github.com/mratsim/Arraymancer/issues/359#issuecomment-546595679 |
| 15:22:20 | narimiran | operation successful (191134 lines compiled; 15.216 sec total; 629.297MiB peakmem; Debug Build) |
| 15:22:27 | FromGitter | <mratsim> cool |
| 15:22:36 | FromGitter | <mratsim> before it was 4GB :P |
| 15:22:37 | narimiran | i have just sliiiiiightly slower machine than you :P |
| 15:22:50 | FromGitter | <mratsim> no that report was just a single test files |
| 15:22:56 | FromGitter | <mratsim> you ran 20 or so |
| 15:23:51 | narimiran | should i then put `nim c tests/tests_cpu.nim` in our important_packages.nim as a better way to test AM? |
| 15:24:26 | FromGitter | <mratsim> yep |
| 15:24:44 | FromGitter | <mratsim> that will instantiate Tensors of strings, floats, ints |
| 15:24:50 | * | jxy quit (Quit: leaving) |
| 15:24:52 | FromGitter | <mratsim> will avoid regression in generics especially |
| 15:25:36 | * | dddddd joined #nim |
| 15:26:10 | narimiran | done |
| 15:27:48 | * | Hideki_ quit (Remote host closed the connection) |
| 15:28:35 | * | Hideki_ joined #nim |
| 15:33:07 | * | Hideki_ quit (Ping timeout: 268 seconds) |
| 15:34:24 | * | letto quit (Quit: Konversation terminated!) |
| 15:35:52 | * | letto joined #nim |
| 15:39:21 | * | jxy joined #nim |
| 15:44:05 | dom96 | narimiran: FYI if you want to run tests in a package then you should be using `nimble develop` ;) |
| 15:45:13 | narimiran | dom96: i've heard recommendations for `nimble develop` so yesterday i wanted to use it, and — all i got was "Error: Cannot develop packages that are binaries only." yay |
| 15:45:51 | narimiran | so today i decided i won't be using it no more |
| 15:46:06 | dom96 | heh, fair enough. |
| 15:46:07 | narimiran | s/using/trying to use |
| 15:46:13 | dom96 | That's a bit harsh. This is something that we should fix. |
| 15:46:52 | dom96 | I was conservative when implementing this because I didn't have time to test how well `develop` worked for binary packages. |
| 15:47:08 | narimiran | while you're here: i'm trying to bring nimforum to 2019 by using the latest nim |
| 15:47:49 | narimiran | `nimble develop` didn't do the trick, so i manually fixed the dependencies |
| 15:48:15 | narimiran | and i've also fixed some outdated stuff in nimforum itself |
| 15:48:50 | narimiran | now when i run manually some of the tasks in nimforum.nimble on my machine — it seems to work fine |
| 15:49:35 | narimiran | but CI is red and i don't understand what went wrong |
| 15:49:54 | narimiran | Error: unhandled exception: FAILED: nimble c -y src/forum.nim |
| 15:50:06 | narimiran | The command "nimble -y test" exited with 1. |
| 15:53:17 | * | theelous3 joined #nim |
| 15:55:40 | * | tklohna quit (Quit: Leaving) |
| 15:55:57 | * | tklohna joined #nim |
| 15:57:08 | FromGitter | <zacharycarter> hrmmm I guess if I'm writing and reading from streams using MemMapFileStream sequences aren't going to work |
| 15:58:09 | Araq | clyybber, found one issue |
| 15:58:24 | * | tklohna quit (Client Quit) |
| 15:58:36 | Araq | it always bothered me, now I understand it :P |
| 15:58:53 | Araq | subconsciousness is great. |
| 15:59:30 | FromGitter | <zacharycarter> would it be possible to use unchecked arrays instead? |
| 16:00:52 | * | Hideki_ joined #nim |
| 16:01:31 | * | tklohna joined #nim |
| 16:02:11 | dom96 | narimiran: any other errors apart from that? |
| 16:02:47 | FromGitter | <zacharycarter> or is it possible to get the size of the data held in the sequence? I figured if I just read the number of bytes of the object stored in the sequence * the number of sequence elements it would work |
| 16:02:50 | FromGitter | <zacharycarter> but this seems not to be the case |
| 16:05:05 | narimiran | dom96: i've pushed a fix just now, i'll see tonight if the error is still there.... |
| 16:05:21 | dom96 | narimiran: have you tried running the same command locally? |
| 16:05:21 | FromGitter | <zacharycarter> oh nevermind I was doing something stupid |
| 16:06:20 | narimiran | dom96: yep, and now i figured i introduced one ambiguity with my previous quickfix |
| 16:10:29 | dom96 | narimiran what does your .nimble file look like? |
| 16:11:26 | FromGitter | <zacharycarter> could someone please help me to understand why this fails? https://play.nim-lang.org/#ix=1UUj |
| 16:11:45 | FromGitter | <zacharycarter> oh |
| 16:12:07 | FromGitter | <zacharycarter> nevermind I still don't know |
| 16:12:50 | * | Hideki_ quit (Ping timeout: 276 seconds) |
| 16:13:14 | FromGitter | <zacharycarter> ah this works |
| 16:13:32 | FromGitter | <zacharycarter> https://play.nim-lang.org/#ix=1UVj |
| 16:13:32 | Mister_Magister | @zacharycarter genious xd |
| 16:13:56 | FromGitter | <zacharycarter> what is genius? |
| 16:15:27 | FromGitter | <zacharycarter> something is wonky - it works in my test but the same code does not work in my app |
| 16:21:02 | clyybber | Araq: Oh, tell me about it |
| 16:22:25 | * | narimiran quit (Ping timeout: 265 seconds) |
| 16:25:54 | * | ng0 quit (Remote host closed the connection) |
| 16:27:22 | * | ng0 joined #nim |
| 16:30:56 | Araq | internalNew_43643(:env_143824) |
| 16:30:56 | Araq | `=_142637`(:env_143824.dep1_-142029, [1, 2, 3]) |
| 16:30:57 | Araq | `=sink_143567`(x_143535, LazyList_142003(c_142008: (:anonymous_142030, :env_143824))) |
| 16:31:16 | Araq | for some reason wasMoved(:env) is missing so we get a nice double free bug |
| 16:40:47 | clyybber | huh |
| 16:40:51 | clyybber | nice? |
| 16:41:56 | FromDiscord | <Rika> buy one take one |
| 16:43:04 | Araq | elif arg.kind in {nkBracket, nkObjConstr, nkTupleConstr, nkClosure} + nkLiterals: |
| 16:43:04 | Araq | # object construction to sink parameter: nothing to do |
| 16:43:04 | Araq | result = arg |
| 16:43:20 | Araq | ^ nope, wrong. you must still process 'arg' recursively anyway |
| 16:43:48 | Araq | so that 'destructiveMoveVar' is a thing for e.g. (a, 2) # a is moved |
| 16:45:30 | * | Vladar joined #nim |
| 16:51:04 | clyybber | Araq: Ah, that makes sense. |
| 16:51:23 | clyybber | good job in finding that |
| 16:53:18 | * | Hideki_ joined #nim |
| 16:57:37 | * | Hideki_ quit (Ping timeout: 240 seconds) |
| 16:57:53 | Araq | and... it makes some tests red, as usual |
| 17:02:23 | FromGitter | <zacharycarter> I know how to debug the compiler when no stack trace is printed, but what about a runtime crash with no stacktrace? |
| 17:03:49 | FromDiscord | <mratsim> gdb 😛 |
| 17:04:09 | FromDiscord | <mratsim> unless you mean a runtime crash of the compiler? |
| 17:04:23 | FromDiscord | <mratsim> Are you embedding the compiler as a library? |
| 17:04:42 | * | nsf joined #nim |
| 17:04:48 | FromGitter | <zacharycarter> I'm using vcc as my compieler though :/ |
| 17:04:56 | FromGitter | <zacharycarter> I meant a runtime crash of my program |
| 17:05:02 | FromDiscord | <mratsim> Ah I'm sorry for you |
| 17:05:08 | FromDiscord | <mratsim> 😛 |
| 17:05:22 | FromGitter | <zacharycarter> well - I can at least isolate the problem code and get that to compile with gdb support |
| 17:05:39 | FromDiscord | <mratsim> GDB cannot trace programs compiled with msvc? |
| 17:06:00 | FromDiscord | <mratsim> otherwise it should have a debugger as well |
| 17:06:28 | FromGitter | <alehander92> @zacharycarter the microsoft debugger should probably work? |
| 17:06:39 | FromGitter | <alehander92> if its compiled with debug info |
| 17:06:40 | FromGitter | <zacharycarter> err I meant use gcc to compile the example code |
| 17:07:04 | FromGitter | <alehander92> but i am not sure why would gdb not work with vcc |
| 17:07:07 | FromGitter | <zacharycarter> it's not too helpful anyway - ``` ⏎ Thread ⏎ 1 received signal SIGSEGV, Segmentation fault. ⏎ 0x0000000000000000 in ?? () ⏎ ... [https://gitter.im/nim-lang/Nim?at=5dcc383a35889012b1005f3d] |
| 17:07:26 | * | tklohna quit (Ping timeout: 276 seconds) |
| 17:07:38 | FromGitter | <alehander92> yeah, debug info? |
| 17:08:34 | FromGitter | <zacharycarter> I compiled with `--debugger:native` |
| 17:08:36 | FromGitter | <zacharycarter> https://gist.github.com/zacharycarter/a924a2c2ec0a3170cf20c2c63e503e09 |
| 17:08:47 | FromGitter | <zacharycarter> maybe someone can spot something stupid I'm doing |
| 17:09:40 | FromGitter | <alehander92> add --stacktrace:on --linetrace:on |
| 17:09:46 | FromGitter | <alehander92> they seem not implied |
| 17:10:04 | FromGitter | <zacharycarter> ah okay |
| 17:11:11 | FromGitter | <zacharycarter> same result :/ |
| 17:12:51 | Araq | addr materials |
| 17:12:54 | Araq | ---> |
| 17:12:57 | Araq | addr materials[0] |
| 17:13:24 | FromGitter | <zacharycarter> yeah - that was what I had at first but that causes run_first.nim to throw up |
| 17:13:25 | Araq | (probably should get some linting support, such a common mistake, apparently) |
| 17:14:29 | FromGitter | <zacharycarter> when trying to echo the seq |
| 17:15:20 | lqdev[m] | Araq: isn't it `materials[0].unsafeAddr`? |
| 17:15:25 | clyybber | Araq: Which tests? |
| 17:15:35 | FromGitter | <alehander92> huh i didnt realize we have memfiles |
| 17:15:37 | FromGitter | <alehander92> or i forgot |
| 17:16:48 | FromGitter | <zacharycarter> well - if I don't echo `materials` in `map.nim` things work but as soon as I echo I get a crash |
| 17:17:00 | Araq | tnewruntime_misc.nim(103, 17) Error: '=' is not available for type <tuple of (string, owned ref int)>; requires a copy because it's not the last read of '("a", new(int))'; routine: tnewruntime_misc |
| 17:17:13 | Araq | (silly, it IS the last read) |
| 17:17:22 | Araq | tmove_objconstr.nim(192, 19) Error: internal error: '=' operator not found for type tuple of (string, MySeqNonCopyable) |
| 17:17:29 | Araq | (same, no need to copy here) |
| 17:17:41 | FromGitter | <zacharycarter> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5dcc3ab5fd6fe41fc07a47e6] |
| 17:18:21 | FromGitter | <zacharycarter> `SIGSEGV: Illegal storage access. (Attempt to read from nil?)` |
| 17:23:20 | FromGitter | <zacharycarter> hmm I guess if I can avoid using sequences that'd be best |
| 17:23:25 | FromGitter | <zacharycarter> some fuckery is going on here |
| 17:24:17 | FromGitter | <alehander92> something is nil in the copied data |
| 17:32:29 | donpdonp | is it possible to specify a nimble "require '../otherpkg'" aka another local package in development |
| 17:32:30 | FromGitter | <zacharycarter> oh - Araq as always is right |
| 17:32:43 | FromGitter | <zacharycarter> it's in the write to the stream in run_first.nim |
| 17:33:02 | FromGitter | <zacharycarter> needs to be ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5dcc3e4dfd6fe41fc07a6291] |
| 17:34:12 | FromGitter | <zacharycarter> or maybe not, ugh going to take a break from this shaiba |
| 17:46:11 | FromGitter | <nixfreakz_twitter> Does anyone have detailed documentation for jester ? |
| 17:49:27 | clyybber | Araq: dfa bug? |
| 17:52:19 | FromDiscord | <spook> ello |
| 17:54:11 | donpdonp | is there a supertype of the nim types for string, float, etc. to make a seq[T] that can hold any primative? |
| 17:56:29 | clyybber | donpdonp: A seq can only hold elements of the same type |
| 17:56:38 | clyybber | Maybe what your are looking for is tuples? |
| 17:57:41 | FromGitter | <nixfreakz_twitter> Are there any other web frameworks for nim? |
| 17:57:51 | clyybber | Though tuples have a fixed size |
| 18:02:47 | donpdonp | clyybber: exactly. im wrapping a c-json lib and the return value is an arbitrary list of string/int |
| 18:03:17 | * | nsf quit (Quit: WeeChat 2.6) |
| 18:04:34 | * | donpdonp looks at the std json.nim for inspiration |
| 18:07:13 | clyybber | donpdonp: How can that be? Surely it has to store it somewhere what it is? |
| 18:07:37 | clyybber | If its a string or an int |
| 18:09:11 | donpdonp | well im prob not describing it very well. the 'object variant' setup in json.nim i think is leading me down the right path. |
| 18:10:03 | clyybber | ok, good luck :) |
| 18:10:31 | FromGitter | <zacharycarter> well - I tried writing the strings individually to the stream and still end up with the same problem |
| 18:10:32 | FromGitter | <zacharycarter> ugh |
| 18:10:52 | disruptek | 197 aws apis now supported. 🎉 |
| 18:11:03 | FromGitter | <zacharycarter> I guess I'll just use file streams and not a mem mapped stream |
| 18:11:49 | * | Hideki_ joined #nim |
| 18:12:29 | FromGitter | <zacharycarter> this is going to bother me though |
| 18:15:20 | clyybber | disruptek: You generate those wrappers? |
| 18:16:10 | * | Hideki_ quit (Ping timeout: 252 seconds) |
| 18:16:21 | disruptek | yeah. |
| 18:16:35 | clyybber | is the generator in your repo too? |
| 18:17:55 | disruptek | nah, i don't wanna support it. |
| 18:20:13 | clyybber | understandable |
| 18:20:50 | disruptek | it doesn't run with raising the vm limits. |
| 18:20:52 | clyybber | I also don't want to bother people with my unholy entanglement of sed awk and a broken c2nim (but hey it generates distinct cints) |
| 18:20:57 | disruptek | s/with/without/ |
| 18:21:28 | clyybber | disruptek: https://github.com/disruptek/cutelog/blob/master/cutelog.nim explain yourself!? |
| 18:21:51 | disruptek | i find it easier to read. |
| 18:23:23 | clyybber | I too 🔍 it easier to 📖 |
| 18:23:50 | clyybber | kidding |
| 18:27:37 | * | jjido joined #nim |
| 18:29:57 | lqdev[m] | I looking glass it easier to book as well. |
| 18:31:57 | * | uu91 quit (Ping timeout: 268 seconds) |
| 18:32:17 | * | uu91 joined #nim |
| 18:35:11 | * | jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 18:35:58 | clyybber | exactly |
| 18:37:39 | dom96 | Araq shashlick: you guys still up for the meeting in around 1.5 hours? |
| 18:38:29 | Araq | yup |
| 18:38:40 | shashlick | ditto - what's the medium |
| 18:41:23 | Araq | clyybber, why do we need p, pArg and moveOrCopy ? |
| 18:41:38 | Araq | the code duplication is still a pita even with the templates |
| 18:42:16 | dom96 | shashlick appear.in I guess |
| 18:44:40 | Araq | never mind, clyybber, it's fine |
| 18:45:15 | Araq | this test suite is a pita |
| 18:45:45 | mipri | TwitLogger: lvlError: "nobody:\n\nyour program:", lvlWarn: "this ain't it, chief:", lvlNotice: "yikes: ", lvlInfo: "this just happened:", lvlDebug: "@jack: " |
| 18:46:39 | clyybber | Araq: Cool, its as tight and non duplicated as I could get it :) |
| 18:52:03 | Araq | this was hard... tests are green again with my patch |
| 18:53:05 | clyybber | Nice, can I see it? |
| 18:53:31 | Araq | one sec |
| 18:55:26 | Araq | look at my araq-arc branch |
| 18:55:35 | * | jjido joined #nim |
| 18:57:40 | clyybber | Araq: Nice. |
| 18:57:43 | clyybber | Now... |
| 18:57:47 | clyybber | does async run? |
| 18:59:11 | Araq | I think the phase ordering bug still exists |
| 18:59:31 | Araq | it receives an nkLambda which must be an nkClosure instead |
| 18:59:52 | * | jjido quit (Client Quit) |
| 18:59:53 | Araq | there is also a transf caching bug and exception handling leaks memory |
| 18:59:59 | clyybber | ah |
| 19:00:12 | Araq | should probably keep a TODO somewhere |
| 19:00:26 | Araq | but IMO we need =trace |
| 19:00:53 | Araq | it's not only for cycle detection, it can also be used to mark all refs as 'shared' and only after that the RC ops must be atomic |
| 19:01:24 | Araq | or we can use it to check for deep uniqueness and require that for thread interop |
| 19:01:34 | Araq | so many possibilities when we can trace... |
| 19:04:06 | clyybber | hmm, that does indeed seem useful. |
| 19:04:29 | clyybber | I just wonder what signature =trace should have |
| 19:10:17 | * | rockcavera quit (Ping timeout: 240 seconds) |
| 19:11:28 | Araq | proc =trace(x: var T; env: pointer) |
| 19:11:39 | * | nsf joined #nim |
| 19:16:13 | clyybber | And env is? The proc thats called? |
| 19:16:24 | Araq | something |
| 19:16:39 | Araq | not of your business :P |
| 19:17:00 | Araq | you pass it along |
| 19:17:37 | clyybber | Ah, and trace also doesn't return anything? |
| 19:17:47 | Araq | yeah |
| 19:18:00 | * | Hideki_ joined #nim |
| 19:18:03 | clyybber | So how do I use it to, lets say =destroy all fields of a special object? |
| 19:18:09 | Araq | but you're right, it's a bit unclear how the =trace atom for 'ref T' would look like |
| 19:18:27 | Araq | well destruction stays as =destroy |
| 19:18:38 | clyybber | Araq: Yeah, I mean for cycles for example |
| 19:18:49 | * | narimiran joined #nim |
| 19:19:03 | * | seni joined #nim |
| 19:20:51 | clyybber | Ok, cycles are a dumb example. |
| 19:21:11 | Araq | I don't know, we'll figure it out |
| 19:21:19 | clyybber | But in general, some other code would use =trace to do something on the things it traces through, right? |
| 19:21:44 | clyybber | So =trace will need to take a closure |
| 19:21:58 | Araq | seems like it |
| 19:22:31 | * | Hideki_ quit (Ping timeout: 268 seconds) |
| 19:26:38 | Araq | er |
| 19:26:51 | Araq | transf imports lambdalifting with imports transf ... |
| 19:27:16 | Araq | no wonder we have phase ordering problems... |
| 19:27:31 | Araq | gah this is terrible |
| 19:29:12 | clyybber | Oof |
| 19:41:32 | * | ng0 quit (Remote host closed the connection) |
| 19:42:43 | * | ng0 joined #nim |
| 19:45:05 | * | filcuc quit (Quit: KVIrc 5.0.0 Aria http://www.kvirc.net/) |
| 19:48:36 | * | clyybber quit (Quit: WeeChat 2.6) |
| 19:48:46 | * | clyybber joined #nim |
| 19:58:47 | * | nsf quit (Quit: WeeChat 2.6) |
| 19:59:51 | dom96 | shashlick: Araq: so, we doing this? |
| 20:00:01 | * | lritter joined #nim |
| 20:01:11 | shashlick | Just getting out of a meeting |
| 20:03:10 | clyybber | Hey, lritter, I was wondering how you do "closures as zero-cost abstractions"? |
| 20:04:18 | * | vsantana quit (Remote host closed the connection) |
| 20:04:52 | * | ng0 quit (Remote host closed the connection) |
| 20:05:20 | * | ng0_ joined #nim |
| 20:05:22 | * | ng0_ is now known as ng0 |
| 20:06:43 | madprops | is there a quicker way to create an object, i mean not having to specify each key Cat(name, color) based on the variable names |
| 20:07:41 | madprops | instead of Cat(name:name, color:color) |
| 20:08:25 | * | jjido joined #nim |
| 20:08:56 | narimiran | madprops: you can make your own `initCat` proc to simplify things a bit |
| 20:09:28 | disruptek | i just use a 0 value for everything. |
| 20:09:44 | lritter | clyybber, they're compile time closures. all our functions are templates. when the function is first referenced, we capture its environment, and when it is typed, its environment is applied. |
| 20:10:25 | lritter | clyybber, unfortunately C-style functions don't allow accessing anything but globals outside the function scope, so we have two kinds of functions: |
| 20:11:01 | lritter | clyybber, the classic `fn`, which describes a C function with hard boundaries, which can only access constants defined in a parent function |
| 20:11:14 | * | Jesin quit (Quit: Leaving) |
| 20:11:36 | FromDiscord | <krab4t> madprops, just remember `isNil` is a user defined error. |
| 20:12:34 | lritter | clyybber, and `inline`, which uses the same syntax and describes an inline (as in "the inline"), which is a function that is inlined into the currently active C function. inlines can access variables in parent scopes and generate new inlines and C functions, but all accessed variables have to reside within the same C function boundaries. |
| 20:12:46 | FromDiscord | <krab4t> disruptek, and then 0 can't be a real value 👺 |
| 20:12:50 | lritter | effectively, an inline is a fancy kind of hygienic macro. |
| 20:13:25 | lritter | clyybber, hope that answers your question. |
| 20:13:41 | * | rockcavera joined #nim |
| 20:14:13 | lritter | in an earlier version, `fn` was used for both cases, and a heuristic found out when to inline a function, but it turned out to be difficult to control so i changed the design. |
| 20:15:08 | * | livcd quit (Quit: Lost terminal) |
| 20:15:52 | * | nif quit (Quit: ...) |
| 20:15:53 | clyybber | lritter: Yeah, thanks. Thats about what I expected. |
| 20:16:01 | * | nif joined #nim |
| 20:19:30 | * | Jesin joined #nim |
| 20:20:14 | * | jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 20:23:32 | FromDiscord | <Lantos> hey @shashlick the nimgraphql is just the ast parser right? Does macros/templates, graphqli need to be made to implement a graphql nim server ? |
| 20:38:01 | * | vsantana joined #nim |
| 20:47:13 | FromDiscord | <Lantos> ahh yes, its just a parser could you rename nimgraphql to nimgraphql-parser? or are there plans to implement it also? |
| 20:51:04 | * | filcuc joined #nim |
| 20:54:42 | * | tiorock joined #nim |
| 20:54:42 | * | tiorock quit (Changing host) |
| 20:54:42 | * | tiorock joined #nim |
| 20:54:42 | * | rockcavera is now known as Guest14892 |
| 20:54:43 | * | Guest14892 quit (Killed (tolkien.freenode.net (Nickname regained by services))) |
| 20:54:43 | * | tiorock is now known as rockcavera |
| 20:55:45 | * | tiorock joined #nim |
| 20:55:45 | * | tiorock quit (Changing host) |
| 20:55:45 | * | tiorock joined #nim |
| 20:55:45 | * | rockcavera quit (Killed (wolfe.freenode.net (Nickname regained by services))) |
| 20:55:45 | * | tiorock is now known as rockcavera |
| 20:59:52 | * | clyybber quit (Quit: WeeChat 2.6) |
| 21:02:27 | * | jjido joined #nim |
| 21:07:16 | madprops | wondering if argops has a way to expose argument flags |
| 21:07:23 | madprops | like for --help |
| 21:07:30 | madprops | parseopts* |
| 21:09:51 | madprops | gonna try https://github.com/iffy/nim-argparse |
| 21:10:44 | * | ltriant joined #nim |
| 21:15:03 | * | clyybber joined #nim |
| 21:30:08 | * | clyybber quit (Quit: WeeChat 2.6) |
| 21:31:42 | madprops | hmm wondering how i can make an argument optional with it |
| 21:33:09 | FromGitter | <mratsim> I use cligen for CLI arguments |
| 21:33:56 | * | uu91 quit (Ping timeout: 276 seconds) |
| 21:34:36 | * | uu91 joined #nim |
| 21:40:36 | Araq | hmm this phase ordering problem seems to be a non-issue after a little refactoring |
| 21:40:55 | Araq | but I should sleep now, good night |
| 21:44:47 | * | narimiran quit (Quit: Leaving) |
| 21:44:51 | * | azed joined #nim |
| 21:48:16 | * | sacredfrog quit (Quit: ZNC 1.7.5 - https://znc.in) |
| 21:49:16 | * | Vladar quit (Quit: Leaving) |
| 21:49:48 | * | sacredfrog joined #nim |
| 21:50:05 | madprops | mratsim do you have an example of cligen in action? |
| 21:50:17 | madprops | the documentation is very confusing |
| 21:50:55 | * | azed quit (Quit: WeeChat 2.6) |
| 21:52:27 | * | filcuc quit (Ping timeout: 240 seconds) |
| 21:56:32 | solitudesf | you can look at its tests, there are plenty of examples. what exactly is confusing? |
| 21:57:19 | madprops | im starting to understand it a bit |
| 22:00:39 | madprops | im calling get_config in the config module, from the main file |
| 22:00:46 | madprops | dispatch(get_config) |
| 22:00:56 | madprops | it just prints a bunch of gibberish |
| 22:01:35 | madprops | also im trying to make the function return an object |
| 22:01:45 | madprops | since it seems i have to do all the processing inside of it |
| 22:02:48 | * | solitudesf quit (Ping timeout: 245 seconds) |
| 22:02:57 | * | jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 22:08:41 | madprops | is the c++ backend better in some ways? |
| 22:09:34 | disruptek | it can produce faster crashes if you're into that sort of thing. |
| 22:12:55 | * | jjido joined #nim |
| 22:15:25 | FromDiscord | <mratsim> lol |
| 22:15:34 | madprops | not sure what backend means here |
| 22:15:39 | FromDiscord | <mratsim> they take less space in the binary because they use native C++ exceptions |
| 22:15:41 | madprops | are there different implementations of the compiler? |
| 22:16:04 | * | ng0 quit (Ping timeout: 260 seconds) |
| 22:16:04 | FromDiscord | <mratsim> Exceptions in C are handwritten, while in C++ they map to native compiler support |
| 22:16:18 | madprops | is there a widely preferred backend? |
| 22:16:40 | FromDiscord | <mratsim> That uses cligen: https://github.com/status-im/nimbus-launch/blob/master/src/nimbus_launch.nim#L136 |
| 22:17:11 | disruptek | the c backend gets the most testing. |
| 22:17:16 | FromDiscord | <mratsim> I prefer C personally. |
| 22:17:44 | FromDiscord | <mratsim> I do use C++ when I have to. |
| 22:18:17 | FromDiscord | <mratsim> Now for FFI, C++ is a bit of a pain due to the potential destructors that people add to their types which means you can't put them in Nim sequences |
| 22:18:28 | FromDiscord | <mratsim> but for pure Nim code, C or C++ doesn't really matter |
| 22:18:49 | FromDiscord | <mratsim> oh, sometimes you can get mysterious x2 to x5 performance by switching backend |
| 22:19:12 | FromDiscord | <mratsim> See: https://github.com/drujensen/fib#natively-compiled-statically-typed |
| 22:19:44 | FromDiscord | <mratsim> and I think @Vindaar had an example where C was 2x faster than C++ for the same code ... |
| 22:21:56 | madprops | how can i dispatch a function from another module and get a result? |
| 22:22:04 | madprops | or at least within the same module |
| 22:22:10 | madprops | with cligen i mean |
| 22:22:47 | FromDiscord | <mratsim> what do you mean by result? |
| 22:23:37 | FromDiscord | <mratsim> dispatch will make it that when calling your program from th CLI, every input will be passed to that function |
| 22:24:37 | madprops | but i need to pass a Config object somewhere, which i create inside the dispatched function |
| 22:25:09 | madprops | or is dispatch the new main function now? |
| 22:25:22 | FromDiscord | <mratsim> dispatch creates a main function for you |
| 22:25:54 | madprops | so i can't simply have arguments processed in their own module and then returned to main |
| 22:26:18 | madprops | i've been getting args in config.nim and sending a Config object to the main func |
| 22:27:13 | FromDiscord | <mratsim> the dispatch function can call other functions or access globals |
| 22:27:18 | FromDiscord | <mratsim> dispatched* |
| 22:30:31 | * | ltriant_ joined #nim |
| 22:31:15 | * | ltriant quit (Ping timeout: 240 seconds) |
| 22:44:24 | * | ng0 joined #nim |
| 22:46:15 | * | tane quit (Quit: Leaving) |
| 22:48:20 | * | abm joined #nim |
| 22:51:13 | * | sealmove joined #nim |
| 22:55:08 | sealmove | so what's new in 1.1.1? |
| 22:57:55 | Araq | sealmove: I'm working on https://github.com/nim-lang/RFCs/issues/177 but soon need to work on bugfixes again |
| 22:59:57 | * | nisstyre quit (Quit: WeeChat 2.5) |
| 23:01:19 | * | nisstyre joined #nim |
| 23:04:02 | sealmove | intresting |
| 23:11:18 | FromGitter | <alehander92> https://archiveprogram.github.com/ |
| 23:11:39 | FromGitter | <alehander92> this is kinda weird imo |
| 23:12:13 | FromGitter | <alehander92> but our code is going to svalbard |
| 23:12:29 | nisstyre | is this any good? https://github.com/BitR/ycm-nimsuggest I'm looking for something similar to clang / dcd in vim where it can, e.g. autocomplete field names of objects for me |
| 23:12:43 | nisstyre | I've already got ycm installed |
| 23:13:49 | FromGitter | <zetashift> @nisstyre this should take care of your nim needs in vim: https://github.com/alaviss/nim.nvim |
| 23:14:07 | nisstyre | hmm I'm not using nvim but I think there might be a compatibility thing |
| 23:14:28 | FromGitter | <zetashift> yea I'm not sure you can use it with ycm |
| 23:14:39 | nisstyre | I'll try it out though thanks |
| 23:15:36 | FromGitter | <zetashift> leorize(the wonderful maker of that great plugin) is AFK I think, you could ask and I think you'll get an answer ASAP |
| 23:28:36 | * | jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 23:29:18 | madprops | says this is deprecated: ref object {. inheritable .} |
| 23:29:24 | madprops | what's the proper way? |
| 23:31:03 | * | norm joined #nim |
| 23:31:26 | * | norm is now known as Guest72127 |
| 23:32:03 | * | Guest72127 quit (Client Quit) |
| 23:32:06 | madprops | nvm got it |
| 23:32:27 | * | norm_ joined #nim |
| 23:32:42 | madprops | not sure if what im doing is correct http://i.imgur.com/26HcAa3.png |
| 23:33:56 | FromGitter | <zetashift> @madprops you might be better of with an object variant: https://nim-lang.org/docs/tut2.html#object-oriented-programming-object-variants |
| 23:34:15 | FromGitter | <zetashift> Also https://play.nim-lang.org/ for all your nim related code sharing goodiness |
| 23:36:36 | * | krux02 quit (Remote host closed the connection) |
| 23:45:49 | FromGitter | <zetashift> off* |
| 23:54:40 | madprops | can table be accessed through.dot notation? or what should i be looking for? |
| 23:54:54 | madprops | obj.name instead of obj["name"] |
| 23:55:30 | FromDiscord | <treeform> has any one setup the new github workflow for their nim library? |
| 23:58:38 | * | vsantana quit (Remote host closed the connection) |
){kind=link}
