| 00:09:18 | * | zachk quit (Quit: Leaving) |
| 00:09:56 | * | darithorn quit () |
| 00:11:45 | * | floppydh quit (Ping timeout: 252 seconds) |
| 00:13:57 | * | floppydh joined #nim |
| 00:37:08 | * | banc quit (Ping timeout: 245 seconds) |
| 00:37:23 | * | banc joined #nim |
| 01:00:39 | * | druonysus quit (Read error: Connection reset by peer) |
| 01:14:15 | FromGitter | <zacharycarter> https://github.com/maiavictor/formality - Rust project specifically mentioning the EVM |
| 01:14:28 | FromGitter | <zacharycarter> isn't ethereum moving to eWASM or something? |
| 01:14:49 | FromGitter | <zacharycarter> the whole crypto space is a giant ball of confusion |
| 01:17:03 | FromGitter | <arnetheduck> ewasm is years away |
| 01:21:16 | FromGitter | <zacharycarter> ah |
| 01:22:52 | FromGitter | <zacharycarter> @arnetheduck - I'm leaving my current position soon, and taking a sabatical to do some further theoretical / DS and algo learning - I'm hoping to re-apply at status in the not-so distant future |
| 01:23:36 | FromGitter | <zacharycarter> but I also need to up my crypto knowledge, because I'm also very lacking in that area |
| 01:24:34 | FromGitter | <zacharycarter> in regards to both cryptography and ethereum |
| 01:25:25 | FromGitter | <zacharycarter> it's very cool to have a bridge to learn about all these concepts and topics via Nim |
| 01:25:44 | FromGitter | <zacharycarter> or a gateway - I guess is a better analogy |
| 01:28:43 | FromGitter | <rayman22201> @zacharycarter you should blog / stream your progress as you do this. It will give you some good "street cred" and it will benefit people wanting to learn Nim and people wanting to learn more about crypto. |
| 01:28:48 | FromGitter | <zacharycarter> I also want to write my own forth interpreter and write my own compiler, and then of course work on the next iteration of my hobby game engine |
| 01:29:01 | FromGitter | <zacharycarter> so who knows wtf I will end up actually working on / learning :P |
| 01:29:31 | FromGitter | <zacharycarter> @rayman22201 - that's a really good idea - I guess whatever I end up pursuing (I have a feeling it will be the game engine / graphics programming stuff) - I'll definitely try to find a way to incorporate that |
| 01:29:43 | FromGitter | <zacharycarter> I'm going to have a ton of free time to figure things like that out - which I'm very much excited about having |
| 01:29:50 | FromGitter | <rayman22201> :D awsome |
| 01:30:27 | FromGitter | <zacharycarter> It's funny - I shared an article in my work's slack channel earlier about algos / data structures |
| 01:30:41 | FromGitter | <zacharycarter> and a guy that recently joined was like - why the hell are you sharing this? doesn't everyone know this crap already? |
| 01:31:05 | FromGitter | <zacharycarter> and then I had to figure out a polite way to say no - especially the people I work with on a consistent basis |
| 01:31:17 | FromGitter | <rayman22201> Well, that's rude. Even though I know a lot of CS, I like to read articles and brush up my knowledge. |
| 01:31:36 | FromGitter | <zacharycarter> yeah - i have a feeling he's fresh out of a comp-sci or masters degree |
| 01:31:43 | FromGitter | <zacharycarter> and just kind of didn't think about what he was typing |
| 01:32:10 | FromGitter | <rayman22201> Your point about politeness is valid regardless. |
| 01:32:31 | FromGitter | <zacharycarter> but - he has a good point - like everyone SHOULD know this stuff if they want to work in software, and if they don't have that passion, then there needs to be some kind of distinction b/w those that do and those that don't |
| 01:32:39 | FromGitter | <zacharycarter> unfortunately, we don't have that / have never had that |
| 01:33:12 | FromGitter | <zacharycarter> so we talked 1 on 1 and kind of compared / contrasted our learning experiences - like he started with circuitry / breadboards / assembly language / C |
| 01:33:23 | FromGitter | <zacharycarter> and I started with HTML / JavaScript, then CSS then VB.NET then C# etc |
| 01:33:32 | FromGitter | <zacharycarter> so it was very cool / insightful to compare and contrats |
| 01:33:44 | FromGitter | <rayman22201> yeah. definitely. |
| 01:34:08 | FromGitter | <arnetheduck> the rigor in thinking will help you craft better code that needs less refactoring and behaves in a more predictable way.. that said, not all useful or great looking products have that - the road to great outcomes can take many turns. |
| 01:34:32 | FromGitter | <zacharycarter> yeah - I think simplicity and iteration are so important |
| 01:34:42 | FromGitter | <zacharycarter> don't seek out complexity and don't be afraid to fail fast |
| 01:35:40 | FromGitter | <zacharycarter> that's another frustration I have - like my job emphasizes pair programming, but I work with a dev that only cares about HTML / JS / CSS |
| 01:35:43 | FromGitter | <rayman22201> Also people take many paths to coding. I've seen amazing programmers come out of "code bootcamps" and shitty programmers with masters degrees. |
| 01:35:59 | FromGitter | <arnetheduck> there's also the domain to consider - sometimes it simply doesn't matter (ie a web ui or game) - but in the case of a project like nimbus, the difference between one and the other will be costly |
| 01:36:45 | FromGitter | <zacharycarter> it's so frustrating pairing with someone that doesn't have a drive or desire to learn, and just wants to write code |
| 01:36:58 | FromGitter | <zacharycarter> forced pair programming is the worst IMO |
| 01:37:38 | FromGitter | <rayman22201> Ironic that HN just had a thread on this: https://news.ycombinator.com/item?id=18245120 |
| 01:37:42 | FromGitter | <zacharycarter> @rayman22201 - yeah - that's true as well, although I think it just really depends on how passionate the person is about programming |
| 01:37:54 | FromGitter | <zacharycarter> :P |
| 01:38:19 | FromGitter | <zacharycarter> @arnetheduck - well, it sucks that web dev and other dev gets relegated to that domain |
| 01:38:29 | FromGitter | <zacharycarter> I'd love to write performant web apps at work - but it's not possible |
| 01:38:55 | FromGitter | <zacharycarter> my company will spend 3+ months optimizing a JavaScript frotnend, and spend zero effort on backend optimziation |
| 01:39:15 | FromGitter | <zacharycarter> also - they pass JSON between backend services everywhere |
| 01:39:20 | FromGitter | <zacharycarter> rarely utilize caching |
| 01:39:34 | FromGitter | <zacharycarter> just, rookie level stuff - that even myself, not having a CS degree, can find major faults with |
| 01:39:48 | FromGitter | <zacharycarter> and when you point out these issues - but don't solve them yourself - you're a malcontent |
| 01:40:45 | FromGitter | <zacharycarter> and when you want to introduce a new idea or performant technology - you face a mountain of criticism / have to defend yourself to a soul-crushing degree |
| 01:40:54 | FromGitter | <zacharycarter> to the point that it's not worth it, and you either leave or adapt |
| 01:41:24 | FromGitter | <rayman22201> https://en.wikipedia.org/wiki/Peter_principle |
| 01:41:55 | FromGitter | <zacharycarter> yup! and then they flatten the engineering org |
| 01:42:05 | FromGitter | <zacharycarter> and just randomly people will rise to positions of autonomy |
| 01:42:08 | FromGitter | <arnetheduck> no, it doesn't suck - it's a reasonable tradeoff. the user experience of a web site is more important than its performance - whether you use a linked list or an array really doesn't matter in that context as long as it's not blatantly and obviously bad - it is rare that you even need to know the difference in that context. also, if a single web page on the internet is poor, it's really not the end of the world. all |
| 01:42:08 | FromGitter | ... the other web pages keep working. |
| 01:42:09 | FromGitter | <zacharycarter> with zero transparency |
| 01:42:51 | FromGitter | <zacharycarter> I'd argue that the modern web experience does holistically suck |
| 01:43:23 | FromGitter | <zacharycarter> and a lot of that is due to software bloat / not paying attention to efficient and performant coding practices |
| 01:43:53 | * | vlad1777d__ quit (Remote host closed the connection) |
| 01:43:54 | FromGitter | <zacharycarter> and with the focus on SEO these days, at least in my experience with the company I work for - having a performant web stack is becoming increasingly important |
| 01:44:10 | FromGitter | <zacharycarter> which directly contradicts all the attention on aesthetics that the modern web dictates |
| 01:44:30 | FromGitter | <zacharycarter> a visit to a modern news website will take several megs of bandwidth for 250k of text |
| 01:44:39 | FromGitter | <zacharycarter> if you ask me - it's because it's not a reasonable tradeoff |
| 01:45:40 | * | vlad1777d joined #nim |
| 01:47:13 | FromGitter | <rayman22201> It's economics. I did webdev for a long time. I got sick of it. The economics just dictate that it's not worth companies putting money into hiring better people. At the end of the day, google is still showing the links in their results, and showing their ads, and "joe schmoe" is still using the website, and paying them money. Consumers don't know any better, and don't really care. You have to remember that you |
| 01:47:13 | FromGitter | ... aren't the target market. Your parents and less-computer literate peers are. |
| 01:47:15 | FromGitter | <arnetheduck> beyond "good enough performance", in the majority of applications you're just wasting time. it's a bit like optimizing a parser for command line arguments - sure, it's a fun intellectual exercise, but for all intents and purposes, it doesn't matter the least. |
| 01:47:50 | FromGitter | <zacharycarter> well - please explain that to my org, because they're wasting millions of dollars in that case |
| 01:47:56 | FromGitter | <zacharycarter> trying to outrank cargurus in SEO |
| 01:48:22 | FromGitter | <arnetheduck> seo matters in that case - not compsci 101 |
| 01:48:33 | FromGitter | <zacharycarter> well if you have people writing bad javascript |
| 01:48:35 | FromGitter | <rayman22201> There are industries where this is different. Banking, crypto, those have higher standards because they loose a lot of money if they don't. Thus they put out money to hire the kind of people that excel in those area. |
| 01:48:36 | FromGitter | <zacharycarter> I mean |
| 01:48:38 | FromGitter | <zacharycarter> ... |
| 01:49:16 | FromGitter | <rayman22201> Gaming is interesting, because they care about performance, but not necessarily reliability. So different kinds of patterns appear there. |
| 01:49:25 | FromGitter | <zacharycarter> I'm not implying the web stack is as critical as what you work on |
| 01:50:01 | FromGitter | <rayman22201> I'm not saying one is better / more critical than the other. Just that the people with the money have different ideas about what is "good" |
| 01:50:03 | FromGitter | <zacharycarter> I'm saying if someone is allocating inside of a loop or writing terrible nested conditional logic because they dont' know any better |
| 01:50:03 | FromGitter | <rayman22201> or good enough |
| 01:50:32 | FromGitter | <zacharycarter> it's not helping in any scenario - and the more folks understand about programming / computer science in general and how to measure the performance of the code they're authoring, the better off they will be |
| 01:51:07 | FromGitter | <zacharycarter> I know that I've operated for over a decade at a high level of understanding - and I've regretted my lack of low level knowledge |
| 01:51:09 | * | vlad1777d quit (Ping timeout: 268 seconds) |
| 01:51:32 | FromGitter | <rayman22201> Unless, they don't care, b/c they are just looking for a paycheck.... which is sadly a lot of people |
| 01:51:35 | FromGitter | <zacharycarter> I'm sure it's not essential to things - but I also know that software is very bloated these days and slow and it's layered and I think that's a result of programmers getting lazy and not caring |
| 01:51:44 | FromGitter | <zacharycarter> > Unless, they don't care, b/c they are just looking for a paycheck.... which is sadly a lot of people |
| 01:52:25 | FromGitter | <rayman22201> and it seem like a lot of those people end up in webdev... because that industry can tolerate that kind of crap. again, unfortunately. |
| 01:52:38 | FromGitter | <zacharycarter> yeah - that's why I'm out |
| 01:52:46 | FromGitter | <zacharycarter> it's really enterprise dev |
| 01:52:51 | FromGitter | <zacharycarter> just don't get into that scene IMO |
| 01:53:10 | FromGitter | <zacharycarter> as soon as you start hearing those buzz words |
| 01:53:13 | FromGitter | <rayman22201> lol... true... the amount of bad enterprise Java I've seen would give you nightmares |
| 01:54:47 | FromGitter | <rayman22201> anyway. Despite the personal / philosophical stuff, I hope you get to play with Nim more. I wish I could. viva la Nim :-) |
| 01:56:37 | FromGitter | <yyyc514> is having a method/proc named the same as my module not possible? |
| 01:56:46 | FromGitter | <yyyc514> "Error: expression 'runtime' has no type (or is ambiguous)" |
| 01:56:50 | FromGitter | <yyyc514> even if i try runtime.runtime |
| 01:56:54 | FromGitter | <yyyc514> renaming it it works fine though |
| 01:57:10 | FromGitter | <zacharycarter> @rayman22201 thank you! me too! |
| 01:57:49 | FromGitter | <zacharycarter> and I hope you get to as well! |
| 01:59:21 | FromGitter | <zacharycarter> @yyyc514 - I don't have a definitive answer for you, but maybe I can help provide some insight |
| 01:59:35 | FromGitter | <zacharycarter> https://nim-lang.org/docs/manual.html#modules-scope-rules & https://github.com/nim-lang/Nim/issues/4971 |
| 02:00:08 | FromGitter | <zacharycarter> going to sign off for the night - have a good one all |
| 02:00:17 | FromGitter | <yyyc514> i’d think this might fall under qualification |
| 02:00:47 | FromGitter | <yyyc514> thanks |
| 02:01:29 | FromGitter | <zacharycarter> 👍 |
| 02:04:53 | * | banc quit (Quit: ZNC - http://znc.in) |
| 02:13:46 | * | darithorn joined #nim |
| 02:20:24 | * | banc joined #nim |
| 02:41:37 | * | dddddd quit (Remote host closed the connection) |
| 02:43:23 | * | arecaceae quit (Remote host closed the connection) |
| 02:43:41 | * | arecaceae joined #nim |
| 02:44:06 | * | endragor joined #nim |
| 03:08:19 | * | chemist69_ joined #nim |
| 03:11:56 | * | chemist69 quit (Ping timeout: 268 seconds) |
| 03:35:02 | FromGitter | <yyyc514> does nim have anything like elixir’s |> for chains of things? |
| 03:36:28 | FromGitter | <kaushalmodi> can you give an example? |
| 03:36:39 | FromGitter | <kaushalmodi> do you know of the Nim UFCS syntax? |
| 03:36:59 | FromGitter | <yyyc514> lol |
| 03:37:23 | FromGitter | <yyyc514> (https://files.gitter.im/nim-lang/Nim/CWnp/Screen-Shot-2018-10-17-at-11.37.00-PM.png) |
| 03:37:42 | FromGitter | <yyyc514> pipes output into the first parameter of the next call |
| 03:37:58 | FromGitter | <kaushalmodi> why lol? just asking asked because I wasn't sure what you meant |
| 03:38:54 | FromGitter | <kaushalmodi> depending on if the proc signatures are designed for use with UFCS and in which combination you are using those procs, that is possible. |
| 03:41:04 | FromGitter | <yyyc514> in that case you could just chain with ., yes? |
| 03:41:24 | FromGitter | <kaushalmodi> yep |
| 03:41:49 | FromGitter | <yyyc514> well isn’t that a neat trick :) |
| 03:42:50 | FromGitter | <kaushalmodi> Nim calls it *Command Invocation Syntax* for some reason: https://nim-lang.github.io/Nim/manual.html#procedures-command-invocation-syntax |
| 03:43:10 | FromGitter | <kaushalmodi> but I think this is more commonly known as UFCS: https://en.wikipedia.org/wiki/Uniform_Function_Call_Syntax |
| 03:43:45 | FromGitter | <yyyc514> teaching me new acronyms |
| 03:45:05 | * | smt quit (Read error: Connection reset by peer) |
| 03:46:04 | * | smt joined #nim |
| 03:52:23 | leorize | kaushalmodi: *Command Invocation Syntax* is **not** UFCS... |
| 03:52:46 | FromGitter | <kaushalmodi> oops |
| 03:52:50 | leorize | in Nim it's actually called the *Method call syntax* |
| 03:53:29 | FromGitter | <kaushalmodi> This example in Command Invocation Syntax uses UFCS: `let y = 1.optarg optarg 8` |
| 03:54:08 | leorize | you do realise it was trying to demonstrate the syntax, right? |
| 03:54:23 | leorize | in UFCS you'd have to: `let y = 1.optarg(8)` |
| 03:55:09 | leorize | but we can omit the parentheses since there's command call syntax |
| 03:55:16 | FromGitter | <kaushalmodi> thanks for correcting me |
| 03:55:32 | FromGitter | <kaushalmodi> so optional parens is command invocation syntax? |
| 03:55:41 | leorize | yep |
| 03:55:49 | leorize | quite unique I'd say |
| 03:55:57 | FromGitter | <kaushalmodi> and dot-chaining aka UFCS is method call syntax? |
| 03:56:03 | FromGitter | <kaushalmodi> *time to fix notes* |
| 04:18:13 | FromGitter | <yyyc514> hmmmm can i not pass the name of a proc in place of a inline function? |
| 04:19:10 | FromGitter | <yyyc514> but expression 'router' is of type: void |
| 04:28:48 | FromGitter | <yyyc514> the exact example in first class functions works |
| 04:28:55 | FromGitter | <yyyc514> mabye it’s because i’m inside another proc and the scope is different? |
| 04:29:35 | FromGitter | <yyyc514> nope |
| 04:32:32 | * | darithorn quit () |
| 04:33:10 | FromGitter | <yyyc514> maybe i’m missing some pragma? |
| 04:34:05 | leorize | what are you trying to do? |
| 04:34:19 | FromGitter | <yyyc514> pass a proc |
| 04:34:48 | FromGitter | <kaushalmodi> can you paste a code snippet? |
| 04:35:30 | FromGitter | <yyyc514> https://gist.github.com/yyyc514/b91b59fdd6bb52d0f29eb42952819e8b |
| 04:37:11 | FromGitter | <kaushalmodi> what is `router` in that example? may be a minimal all-inclusive example would help |
| 04:37:33 | FromGitter | <yyyc514> added |
| 04:37:46 | FromGitter | <yyyc514> it’s just a proc |
| 04:37:55 | FromGitter | <yyyc514> that matches the call protocol |
| 04:38:47 | FromGitter | <yyyc514> oh grrrr it’s probably the name of the improt killing me again |
| 04:38:54 | FromGitter | <yyyc514> import router and then also a router rpco |
| 04:38:59 | FromGitter | <yyyc514> let me change that :) |
| 04:39:02 | FromGitter | <kaushalmodi> that would do it |
| 04:39:22 | FromGitter | <kaushalmodi> /because I did a simpler version of what you are doing yest and it works./ |
| 04:39:36 | leorize | same here |
| 04:39:52 | FromGitter | <yyyc514> yep that fixes it :) |
| 04:48:17 | FromGitter | <yyyc514> @kaushalmodi thanks |
| 05:19:32 | * | floppydh quit (Quit: WeeChat 2.2) |
| 05:31:31 | * | nsf joined #nim |
| 05:36:51 | * | Cthalupa quit (Ping timeout: 268 seconds) |
| 05:38:59 | * | Cthalupa joined #nim |
| 05:52:28 | * | krux02 joined #nim |
| 05:55:14 | FromGitter | <yyyc514> how can i have a proc that optimzies differently depending on if it’s given a constant or not? |
| 05:55:21 | FromGitter | <yyyc514> if it’s constant it can be {.compileTime} |
| 05:56:37 | leorize | the compiler can usually infer {.noSideEffects.} functions and optimize them at compile time |
| 05:57:08 | FromGitter | <yyyc514> oh so this is built in? |
| 05:57:27 | FromGitter | <yyyc514> i was imagining like string helpers etc |
| 05:57:41 | FromGitter | <yyyc514> “constant”.toUpperAscii could be done at comile time |
| 05:57:43 | FromGitter | <yyyc514> for example |
| 05:58:01 | leorize | `const someConst = "constant".toUpperAscii` |
| 05:58:05 | leorize | yep, done at compile time |
| 05:58:16 | FromGitter | <yyyc514> sure if you use const, but if not will it figure it out? |
| 05:58:29 | leorize | one way to know, try it |
| 05:59:27 | FromGitter | <yyyc514> nope seems dynamic |
| 05:59:38 | FromGitter | <yyyc514> even if the proc only takes constant input |
| 06:00:57 | FromGitter | <yyyc514> wait, how would i test that properly? |
| 06:01:15 | leorize | read the generated C code, ofc |
| 06:01:30 | leorize | but the C compiler might optimize it |
| 06:01:36 | leorize | so read the assembly |
| 06:01:43 | FromGitter | <yyyc514> ok another day :) |
| 06:02:02 | FromGitter | <yyyc514> might be getting into premature optimization |
| 06:04:38 | leorize | yea compile time eval only stop at `const` & `static` section |
| 06:05:12 | leorize | so if you want your `proc` to only work in VM context, use `{.compileTime.}` |
| 06:05:38 | FromGitter | <yyyc514> yeah but then if you give it a variable it throws an obscure compiler error :) |
| 06:05:51 | FromGitter | <yyyc514> i was wanting best of both worlds |
| 06:06:15 | leorize | https://nim-lang.github.io/Nim/manual.html#special-types-static-t |
| 06:08:09 | FromGitter | <yyyc514> i think i was hoping static would be a part of the overloading so i could just write both functions, one static and one not |
| 06:08:47 | FromGitter | <yyyc514> probably could do it with macros |
| 06:09:29 | leorize | `static` is a part of the overloading resolution |
| 06:15:01 | * | smt quit (Read error: Connection reset by peer) |
| 06:15:04 | FromGitter | <yyyc514> well then maybe it would work :) |
| 06:15:31 | leorize | and static param is pretty weird |
| 06:15:53 | leorize | for example: `echo toUpperAscii toLowerAscii "constant"` |
| 06:16:15 | leorize | toUpperAscii here takes a `static string`, but there's also an overload for `string` |
| 06:16:29 | leorize | toLowerAscii is the stock one from strutils |
| 06:17:04 | leorize | instead of running in runtime (like it does if you don't have the `static` version) |
| 06:17:23 | leorize | `toLowerAscii` is evaluated during compile time |
| 06:19:13 | leorize | https://ptpb.pw/8cFg/nim |
| 06:22:09 | FromGitter | <yyyc514> nice |
| 06:23:29 | FromGitter | <yyyc514> seems like a pretty verbose way to get a small speedup though |
| 06:23:41 | leorize | the sad part is that `toUpperAscii` is still evaluated at runtime |
| 06:23:51 | leorize | yea, this is literally premature optimization :P |
| 06:24:03 | leorize | instead, use `const` whenever you can |
| 06:24:35 | FromGitter | <yyyc514> wait what? why would toLower be evaluted at compile time but not too upper? |
| 06:25:23 | leorize | yes, it compiles into a special `proc` that takes nothing and operate on the string directly |
| 06:25:35 | leorize | which is how `static[T]` works |
| 06:26:08 | FromGitter | <yyyc514> and you’re seeing this from the C output? |
| 06:26:12 | leorize | yep |
| 06:26:28 | leorize | https://ptpb.pw/pHw9/nim |
| 06:26:41 | leorize | and here is how to make something `{.compileTime.}` |
| 06:27:10 | FromGitter | <yyyc514> yeah i tried that earlier |
| 06:27:26 | FromGitter | <yyyc514> i was playing with variables or not though not layers of functions |
| 06:41:59 | FromGitter | <yyyc514> name params with macors is a bit funny |
| 06:42:14 | FromGitter | <yyyc514> in that if you just break the syntax slightly now your named params gets whatever |
| 06:55:00 | FromGitter | <yyyc514> doe a macro have access to the name of the current file context it’s running inside of? |
| 06:56:34 | FromGitter | <yyyc514> what could lineInfoObj be :) |
| 07:01:12 | FromGitter | <yyyc514> though it needs a node to work with |
| 07:01:18 | FromGitter | <yyyc514> just just call a macro with no params |
| 07:03:18 | FromGitter | <yyyc514> ok nope you can just use the empty resul’ts AST :) |
| 07:08:33 | FromGitter | <yyyc514> macros may not work so well with filters though |
| 07:12:13 | FromGitter | <yyyc514> has anyone played with stdtmpl and macros before? |
| 07:18:41 | * | PMunch joined #nim |
| 07:20:01 | * | sagax joined #nim |
| 07:21:05 | FromGitter | <technicallyagd> You mean generating macros with SCF? |
| 07:21:25 | FromGitter | <yyyc514> trying to use a macro inside stdtmpl to write a template :) |
| 07:21:32 | FromGitter | <yyyc514> not even sure that’s supported in general |
| 07:22:45 | FromGitter | <technicallyagd> Gonna try that now |
| 07:23:29 | FromGitter | <yyyc514> the source for it is matching keywords in order to do the proper spacing |
| 07:23:40 | FromGitter | <yyyc514> it doesn’t have any code to match on “:” that i see so i think the spcing would forever be wrong |
| 07:24:05 | FromGitter | <yyyc514> which kind of makes sense with what i’m seeing |
| 07:27:06 | FromGitter | <yyyc514> https://gist.github.com/yyyc514/302ebd7341b7ecb6f64eb9d75a975ce7 |
| 07:27:15 | FromGitter | <yyyc514> was trying to make a macro to write that boiler plate |
| 07:29:32 | FromGitter | <alehander42> @Araq I saw that, what I did was a helper to actually load karaxdsl templates from different files on compile time (e.g. let page = render("main_view")) |
| 07:29:55 | FromGitter | <alehander42> so I wondered if this should be included somewhere |
| 07:30:28 | FromGitter | <yyyc514> are you talking to me at all? |
| 07:30:33 | FromGitter | <yyyc514> looks related even if it’s not lol |
| 07:31:51 | FromGitter | <technicallyagd> @yyyc514 the gist you provided works though? |
| 07:32:43 | FromGitter | <yyyc514> yes what i provided works |
| 07:32:56 | FromGitter | <yyyc514> i wanted to remove the ‘result’ boiler plate and auto-generate the name of the template based on the filename |
| 07:33:22 | FromGitter | <yyyc514> but i’m not sure source code filters is advanced enough for that |
| 07:34:13 | FromGitter | <technicallyagd> So you want a macro that creates template for you with different names |
| 07:34:30 | FromGitter | <yyyc514> inside a code filter file :) |
| 07:34:47 | FromGitter | <yyyc514> i’m not sure if macros can write templates in general, i’ve never tried |
| 07:35:15 | FromGitter | <yyyc514> but i think the larger issue is there seems to be no way to run a macro inside a code filter and also pass it a block |
| 07:36:52 | FromGitter | <technicallyagd> I am still unsure of how you want to use this macro |
| 07:37:14 | FromGitter | <yyyc514> `# view:` LOL |
| 07:37:37 | FromGitter | <yyyc514> then it spits out the boilter plate proper template, etc |
| 07:37:53 | FromGitter | <yyyc514> and i’d have a 3 line header |
| 07:38:03 | FromGitter | <yyyc514> #stdtmpl, import view, and view: |
| 07:38:06 | FromGitter | <yyyc514> whoa |
| 07:38:42 | FromGitter | <technicallyagd> what is `view`? |
| 07:38:43 | FromGitter | <yyyc514> having to include the name right now is the most annoying aprt |
| 07:38:55 | FromGitter | <yyyc514> it would be a macro that writes the view template |
| 07:39:02 | FromGitter | <yyyc514> which is what the whole gist is |
| 07:39:35 | FromGitter | <yyyc514> updated gist |
| 07:40:03 | FromGitter | <technicallyagd> Ah I see |
| 07:41:08 | FromGitter | <yyyc514> i’ll probably end up havign to reimplement the stdtmpl stuff myself to just work on “raw” files without any header boilerplate |
| 07:46:01 | FromGitter | <alehander42> @yyyc514 I wasn't addressing you, but you might also find it interesting |
| 07:46:11 | sagax | hi all! |
| 07:46:17 | sagax | we have some web framework? |
| 07:46:22 | FromGitter | <alehander42> I am using dsl similar to https://github.com/pragmagic/karax/blob/master/examples/mediaplayer/mediaplayer.nim#L29 without any boilerplate |
| 07:46:50 | leorize | sagax: there're plenty |
| 07:46:53 | FromGitter | <yyyc514> sagax: i’m working on one :) there is also Jester |
| 07:46:58 | FromGitter | <yyyc514> oh, there are plenty? |
| 07:47:04 | FromGitter | <yyyc514> what else is big other than jester? |
| 07:47:33 | FromGitter | <yyyc514> @alehander42 yeah i want/need RAW html |
| 07:47:40 | FromGitter | <yyyc514> i don’t want to write in some psuedo html thingy |
| 07:47:41 | leorize | mofuw, or Meccha hayai Asynchronous I/O no super Fast de Ultra minimal na Web |
| 07:48:00 | leorize | yes, it's a thing :P https://github.com/2vg/mofuw |
| 07:48:15 | sagax | well |
| 07:48:36 | sagax | good |
| 07:48:42 | sagax | will se this framework |
| 07:49:03 | FromGitter | <yyyc514> cool might be good to look at for inspiration but i think i’m wanting to build something a bit larger :) |
| 07:49:25 | leorize | there's also httpbeast by dom96 |
| 07:49:33 | FromGitter | <yyyc514> sagax: it’s probably a few days away from actually doing much, though i think i’ve figured out a lot of foundational stuff |
| 07:49:39 | FromGitter | <yyyc514> i’d call beast a server not a framework |
| 07:49:49 | FromGitter | <yyyc514> like asynchttpserver is a server |
| 07:49:53 | FromGitter | <alehander42> @yyyc514 I hate raw html, xml-like syntaxes take so much space |
| 07:50:04 | FromGitter | <alehander42> yeah httpbeast is a server |
| 07:50:13 | FromGitter | <alehander42> a bigger framework would be very cool to see |
| 07:50:27 | FromGitter | <yyyc514> my background is Rails and Phoenix |
| 07:51:04 | FromGitter | <yyyc514> so far i have simple routing, before and after processors, middleware, and working on controllers and templates now |
| 07:52:27 | FromGitter | <yyyc514> well controllers work actually it’s just templating and bulking them up a bit more |
| 07:53:13 | FromGitter | <yyyc514> so this is aiming to be less minimal than Jester, although at first i’m just going to build on top of the Jester foundation (ripping out the router)… that’ll get something running fast |
| 07:53:41 | FromGitter | <yyyc514> i’m curious about the value of having some standard HTTP wrapper like Rack that would make working with different back ends simpler |
| 07:54:03 | FromGitter | <yyyc514> so i’m building that in the middle and then maybe later we can push it out |
| 07:56:37 | FromGitter | <yyyc514> (https://files.gitter.im/nim-lang/Nim/ArQ6/Screen-Shot-2018-10-18-at-3.56.18-AM.png) |
| 07:56:44 | FromGitter | <yyyc514> ie, much more app like structure vs “everything in one big file" |
| 08:08:56 | FromGitter | <mratsim> ugh, so many folders |
| 08:09:44 | PMunch | Hmm, unsafeAddr tells me it has no address and says that I should try to use unsafeAddr |
| 08:09:52 | FromGitter | <mratsim> I’m curious, so many people from Ruby try Elixir and Crystal, what brought you to Nim? |
| 08:10:16 | FromGitter | <mratsim> @yyyc514 ^ |
| 08:12:38 | FromGitter | <yyyc514> i’ve tried Elixir and Crystal :) is it either or ? :) |
| 08:13:15 | FromGitter | <yyyc514> i found: https://totallywearingpants.com/posts/nim-underdog/ |
| 08:13:36 | FromGitter | <yyyc514> i like that it compiles to C (and JS) and might even be usable (for me) in an arduino context one day |
| 08:13:38 | FromGitter | <mratsim> not really more like curious of what brought you here |
| 08:14:03 | FromGitter | <yyyc514> i liked that it’s small and fast, that the compiler can compile itself so quickly is hugely impressive |
| 08:14:09 | PMunch | yyyc514, Nim does run on Arduino |
| 08:14:10 | FromGitter | <mratsim> For embedded: https://github.com/VPashkov/awesome-nim#embedded |
| 08:14:14 | FromGitter | <yyyc514> if you’ve ever tried to build rust or crystal :) |
| 08:14:31 | FromGitter | <mratsim> I did try Rust for a couple of months before Nim, didn’t try crystal |
| 08:14:58 | FromGitter | <yyyc514> i tried crystal a year or two ago, been a while |
| 08:15:31 | FromGitter | <mratsim> Crystal main issu I would say is the syntax with `end` that I just don’t like and the lack of Windows support. |
| 08:15:49 | FromGitter | <yyyc514> i don’t mind end, coming from Ruby of course :) |
| 08:15:51 | FromGitter | <mratsim> Ah, the OOP focus. |
| 08:15:58 | FromGitter | <yyyc514> the python way is kind of blah, but i can get used to it |
| 08:16:09 | FromGitter | <mratsim> I suppose I can get used to it as well |
| 08:16:26 | FromGitter | <yyyc514> yeah i havent’ looked at hacking objects with macros yet, but my middleware stuff comes awefully close |
| 08:17:17 | FromGitter | <alehander42> @yyyc514 yeah something more unified like rack would be nice, currently httpserver and httpbeast's server api-s are not too different (something like jester's request.nim can be also expected) |
| 08:18:58 | FromGitter | <yyyc514> well you’d start with very low-leverl (like rack, ENV and 3 piece typle responses) and then build on top of that |
| 08:19:04 | FromGitter | <yyyc514> like a generic request object just to make things nicer |
| 08:19:17 | FromGitter | <yyyc514> proc toTuple*(r: Response) : RackOut = |
| 08:19:19 | FromGitter | <yyyc514> lol :) |
| 08:20:08 | FromGitter | <yyyc514> and request would wrap ENV and just make it nicer |
| 08:20:15 | FromGitter | <yyyc514> right now both async and httpbeast are trying to do too much |
| 08:20:27 | FromGitter | <yyyc514> trying to build the whole abstraction because no one else has |
| 08:20:40 | FromGitter | <alehander42> also, a way to use a custom template would be cool (I guess you can staticRead a config file and import `<current_template_lib>` based on it) |
| 08:20:54 | FromGitter | <yyyc514> custom template? |
| 08:21:21 | FromGitter | <yyyc514> eventually you’d want the ability to have different types of template files i presume |
| 08:21:31 | FromGitter | <yyyc514> but i’m not sure how that works with the compiler workflow |
| 08:21:36 | FromGitter | <alehander42> yeah, this would make sense, especially if it's zero overhead (no unnecesarry object alocations) |
| 08:21:45 | FromGitter | <alehander42> I described a possible way above |
| 08:22:04 | FromGitter | <yyyc514> no i mean how to parse and get all the individual templates into the binary |
| 08:22:05 | FromGitter | <mratsim> I vote for renaming Alexander to alehanderZero instead of Alehander42 |
| 08:22:05 | FromGitter | <alehander42> where you can generate the import statement in a macro |
| 08:22:27 | FromGitter | <yyyc514> no cross-platform way to walk the file tree with compiler |
| 08:23:05 | FromGitter | <alehander42> and e.g. expect that a template system defines `render(source: string): Stuff` |
| 08:23:41 | FromGitter | <alehander42> ah well I do it also with a macro currently |
| 08:23:51 | FromGitter | <mratsim> Ah I’d like to embed html templates and JS into a Nim binary like what is possible in go https://github.com/zserge/webview#how-to-serve-or-inject-the-initial-htmlcssjavascript-into-the-webview |
| 08:23:55 | FromGitter | <yyyc514> that presumes you arent’ passing render dynamic values i think |
| 08:24:17 | FromGitter | <yyyc514> render(variable) kind of falls on it’s face |
| 08:24:23 | FromGitter | <yyyc514> since variable isn’t known at compile time |
| 08:24:26 | FromGitter | <yyyc514> but maybe you just don’t allow that |
| 08:24:27 | FromGitter | <technicallyagd> @yyyc514 LOL I got it to work. |
| 08:24:27 | FromGitter | <alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc8433bc08b8b306732186a] |
| 08:24:44 | FromGitter | <yyyc514> that’s a constant |
| 08:24:45 | FromGitter | <yyyc514> not a variable :) |
| 08:25:03 | FromGitter | <technicallyagd> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc8435fc08b8b30673218bc] |
| 08:25:04 | FromGitter | <yyyc514> view_name : string; render(view_name) is a whle other thing |
| 08:25:34 | FromGitter | <alehander42> it is a string, I just used `untyped` out of habit |
| 08:25:40 | FromGitter | <yyyc514> @technicallyagd what the hell? is the trick your emit or what |
| 08:25:48 | FromGitter | <technicallyagd> yeah |
| 08:25:51 | FromGitter | <alehander42> I don't understand what you mean by constant |
| 08:25:57 | FromGitter | <yyyc514> @alehander42 it’s a constant at compile time |
| 08:26:00 | FromGitter | <yyyc514> KNOWN |
| 08:26:05 | FromGitter | <technicallyagd> I manually put 2 spaces before result.add |
| 08:26:05 | FromGitter | <alehander42> what is a constant? |
| 08:26:11 | FromGitter | <technicallyagd> so the indentation will be correct |
| 08:26:16 | FromGitter | <alehander42> the path name? |
| 08:26:22 | FromGitter | <yyyc514> the filename |
| 08:26:34 | FromGitter | <yyyc514> @technicallyagd damn is that all it is? |
| 08:26:49 | FromGitter | <yyyc514> @technicallyagd is that generating the template also? |
| 08:26:52 | FromGitter | <technicallyagd> the `view` macro is super ugly though |
| 08:27:09 | FromGitter | <technicallyagd> I will update a gist, hold on |
| 08:27:58 | FromGitter | <alehander42> well, I very rarely need to pass a dynamic path honestly |
| 08:29:14 | FromGitter | <technicallyagd> @yyyc514 https://gist.github.com/technicallyagd/27e5e51f6f3cc5778f8e873304c57e7a |
| 08:29:24 | FromGitter | <alehander42> @yyyc514 but still that's not a big problem: you can do something else, load all the templates on compile time and assign each one to a function, this way you would be able to render them dynamically |
| 08:29:34 | FromGitter | <technicallyagd> All three files included |
| 08:30:44 | FromGitter | <yyyc514> can’t you quote any of that? |
| 08:31:03 | FromGitter | <yyyc514> yeah the loading them all at compile time seaptely was what i was wondering about |
| 08:32:49 | FromGitter | <yyyc514> (https://files.gitter.im/nim-lang/Nim/RDmx/Screen-Shot-2018-10-18-at-4.32.33-AM.png) |
| 08:33:25 | FromGitter | <alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc84555271506518d2e75ed] |
| 08:33:39 | FromGitter | <alehander42> you basically need to generate something like this in a macro |
| 08:33:54 | FromGitter | <yyyc514> but you need all the paths :) |
| 08:34:05 | FromGitter | <yyyc514> so now you’re getting into a small build process :) |
| 08:34:07 | FromGitter | <yyyc514> not impossible |
| 08:34:12 | FromGitter | <yyyc514> but i don’t know how to do it all in one pass |
| 08:34:42 | FromGitter | <alehander42> well the framework probably expects the views to be in certain directories |
| 08:34:42 | FromGitter | <yyyc514> @technicallyagd any way to get rid of import view without glocally importing it everywhere? (which i think the compiler can do?) |
| 08:34:59 | FromGitter | <yyyc514> @alehander42 lol it would but you still can’t iterate over the filesystem inside the VM |
| 08:35:07 | FromGitter | <yyyc514> you can’t find out which files are present |
| 08:35:11 | FromGitter | <yyyc514> so you don’t known which ones to compile |
| 08:35:21 | FromGitter | <alehander42> ah |
| 08:35:56 | FromGitter | <yyyc514> all the file stuff (other than static read) is locked up inside `os` which is OS dependent and not accessible from the VM |
| 08:36:15 | FromGitter | <technicallyagd> @yyyc514 If you import it in `main.nim` before including the template then it would probably work |
| 08:36:21 | FromGitter | <yyyc514> you can drop to shell and do something like `find` but now you’re back to if that works on windows or not :) |
| 08:36:42 | FromGitter | <yyyc514> oh |
| 08:36:45 | FromGitter | <yyyc514> the template was imported |
| 08:36:51 | FromGitter | <yyyc514> but yeah if in inculded it you’re right i think |
| 08:37:17 | FromGitter | <yyyc514> include doesn’t support the nice syntax |
| 08:37:27 | FromGitter | <technicallyagd> @yyyc514 `quote` is awesome, I didn't know it exists lol. I will update the gist, hold on. |
| 08:37:32 | FromGitter | <yyyc514> import views/pastes / [index, show] |
| 08:37:44 | FromGitter | <yyyc514> @technicallyagd ok i have it working though :) |
| 08:38:01 | FromGitter | <yyyc514> yeah quote is super helpful if you don’t need to insert too much dynamically |
| 08:39:00 | FromGitter | <yyyc514> now if i could just do a global include for the view macro itsel then it imporst the other view stuff…. |
| 08:39:23 | FromGitter | <technicallyagd> Yeah I think you need to include it instead of importing it |
| 08:39:35 | FromGitter | <technicallyagd> I meant the macro |
| 08:39:38 | FromGitter | <yyyc514> would --import:PATH get the job done? |
| 08:40:22 | FromGitter | <yyyc514> yep |
| 08:41:28 | FromGitter | <yyyc514> SWEET |
| 08:41:32 | FromGitter | <yyyc514> 2 line header |
| 08:41:50 | FromGitter | <technicallyagd> import for the `macrosTemp.tmpl`? |
| 08:42:13 | FromGitter | <yyyc514> i just imported the view.nim which has the view macro globally |
| 08:42:23 | FromGitter | <yyyc514> if i add more stuff to it theni’ll split it out and have it do it’s own imports |
| 08:43:14 | FromGitter | <yyyc514> grrrrr |
| 08:43:17 | FromGitter | <yyyc514> only works on empty files |
| 08:43:25 | FromGitter | <yyyc514> because the padding for the COMMANDS is now wrong |
| 08:43:29 | FromGitter | <yyyc514> only HTML works |
| 08:44:19 | FromGitter | <yyyc514> well damn |
| 08:45:16 | FromGitter | <technicallyagd> COMMANDS? |
| 08:45:24 | FromGitter | <yyyc514> embedded code |
| 08:45:31 | FromGitter | <yyyc514> `# if true == true: ⏎ <h1>hello world</h1>` |
| 08:45:39 | FromGitter | <yyyc514> the if needs two extra spaces |
| 08:46:03 | FromGitter | <yyyc514> you fixed the string concat for HTML for not the inline code |
| 08:46:12 | FromGitter | <technicallyagd> Ah damn |
| 08:46:40 | FromGitter | <yyyc514> the compiler needs to be extended to support “do" |
| 08:47:07 | FromGitter | <yyyc514> but wait can’t you use do randomly with other things? |
| 08:48:26 | FromGitter | <yyyc514> @alehander42 but yes you’re completely right if your view name are known at compile time a simple macro will get it done |
| 08:49:46 | FromGitter | <yyyc514> actually if ou’re using a macro you can’t truly pass a variable to it |
| 08:49:55 | FromGitter | <yyyc514> without turning the whole thing into a runtime call |
| 08:50:14 | FromGitter | <yyyc514> but i’d say for 99% of uses it’d be static :) |
| 08:51:50 | FromGitter | <yyyc514> wait |
| 08:51:56 | FromGitter | <yyyc514> what is this strong spaces parser? |
| 08:53:33 | FromGitter | <alehander42> @yyyc514 yeah, but you can still have the best of both worlds |
| 08:53:56 | FromGitter | <alehander42> a macro for compile-time constant paths |
| 08:54:05 | FromGitter | <alehander42> and a proc for variables |
| 08:54:27 | FromGitter | <yyyc514> how would your proc work? :) i was assumign the templates included code that needed to be compiled |
| 08:54:50 | FromGitter | <alehander42> I am not sure about staticReadDir: maybe we can add it to the vm, but in the worst case one can also use staticExec |
| 08:55:08 | FromGitter | <yyyc514> yeah that’s the hack that’s not cross platform safe :) |
| 08:55:45 | FromGitter | <alehander42> @yyyc514 well it doesn't have to be ;) |
| 08:55:47 | FromGitter | <yyyc514> oh for sure if the VM includes a few things to walk the dir tree that’d be awesome |
| 08:56:05 | FromGitter | <yyyc514> how many people are using nim on windows nayways? |
| 08:56:12 | FromGitter | <alehander42> you can exec a helper nim program which will itself use the platform-compatible stdlib |
| 08:56:26 | FromGitter | <alehander42> not the best possible solution, but it's possible |
| 08:58:49 | FromGitter | <technicallyagd> @yyyc514 LOL, I think there is a even simpler solution |
| 08:59:20 | FromGitter | <technicallyagd> Use `view(code)` instead of `view: code` syntax |
| 08:59:35 | FromGitter | <technicallyagd> then you don't even need to tweak the `emit` param |
| 09:00:11 | FromGitter | <yyyc514> but how does thatwork in the template then? |
| 09:00:28 | FromGitter | <yyyc514> you need the : to get the block |
| 09:00:36 | FromGitter | <yyyc514> afaik |
| 09:03:26 | * | gmpreussner quit (Ping timeout: 268 seconds) |
| 09:03:31 | * | gmpreussner_ joined #nim |
| 09:04:24 | FromGitter | <technicallyagd> Well I think you could, I have updated the gist |
| 09:04:29 | FromGitter | <technicallyagd> https://gist.github.com/technicallyagd/27e5e51f6f3cc5778f8e873304c57e7a |
| 09:05:05 | FromGitter | <yyyc514> whoa does that work? |
| 09:05:15 | FromGitter | <technicallyagd> Yeah, it works on my machine |
| 09:06:38 | FromGitter | <yyyc514> hmmm |
| 09:06:42 | FromGitter | <yyyc514> not sure what i think of that syntax :) |
| 09:07:47 | FromGitter | <technicallyagd> lol |
| 09:08:11 | * | platoff joined #nim |
| 09:08:55 | * | craigger joined #nim |
| 09:09:34 | livcd | yyyc514: i do |
| 09:09:56 | FromGitter | <yyyc514> the dangling ) is just annoying :) |
| 09:10:17 | * | craigger_ quit (Ping timeout: 244 seconds) |
| 09:11:35 | FromGitter | <yyyc514> i’m trying to hack it but then i end up with: Error: 'export' is only allowed at top level |
| 09:12:21 | FromGitter | <yyyc514> ah because my way puts the sapces back |
| 09:12:27 | FromGitter | <yyyc514> but now the template export is no longer valid |
| 09:12:33 | FromGitter | <yyyc514> it has to be at the top level |
| 09:13:13 | FromGitter | <yyyc514> seems by the time all is said and done it would be, but i guess the compiler isn’t liking it |
| 09:13:39 | FromGitter | <yyyc514> you can use a trick like `# block: discard; view:` to get the compiler to shift the indent |
| 09:13:46 | FromGitter | <yyyc514> oh wait |
| 09:15:43 | FromGitter | <alehander42> @yyyc514 I managed to write a loadViews macro (which I'll probably reuse for my project too) |
| 09:15:44 | * | MyMind quit (Read error: Connection reset by peer) |
| 09:17:23 | * | Senketsu quit (Ping timeout: 252 seconds) |
| 09:18:31 | * | Senketsu joined #nim |
| 09:18:37 | FromGitter | <yyyc514> so weird |
| 09:18:52 | * | MyMind joined #nim |
| 09:20:17 | FromGitter | <yyyc514> WAI |
| 09:20:31 | FromGitter | <yyyc514> is the spacing really what’s doing it? isn’t code just outputting an AST one way or the other? |
| 09:26:55 | FromGitter | <alehander42> what is the problem with spacing |
| 09:27:27 | FromGitter | <yyyc514> i tricked it into adding the two missing spaces but now it thinks i’m trying to define a templte at a non top level |
| 09:30:03 | FromGitter | <alehander42> well you should be able to define a template at non-top level |
| 09:30:14 | FromGitter | <yyyc514> sure but you can’t export it |
| 09:30:19 | FromGitter | <yyyc514> and i want to import my templates not include them |
| 09:31:34 | FromGitter | <yyyc514> feels buggy though |
| 09:31:41 | FromGitter | <yyyc514> just being isinde a macro isn’t non top level |
| 09:31:51 | FromGitter | <yyyc514> the macro (if it’s top level) is writing things into the top level |
| 09:32:58 | FromGitter | <yyyc514> the problem isn’t with the genrate AST though it starts at TemplateDef |
| 09:33:34 | FromGitter | <alehander42> well you have to give some kind of example(e.g. generated code) |
| 09:34:05 | FromGitter | <yyyc514> it blows up first :) |
| 09:34:20 | FromGitter | <yyyc514> but exports in a simple macro don’t work either |
| 09:34:22 | FromGitter | <alehander42> hm, actually I can iterate through directories in a macro |
| 09:34:32 | FromGitter | <yyyc514> @alehander42 how? |
| 09:45:20 | FromGitter | <yyyc514> so wow view( vs view: puts the compiler in an entirely different frame of mind |
| 09:45:30 | FromGitter | <yyyc514> one increases the depth one doesn't |
| 09:45:44 | FromGitter | <yyyc514> i wonder if there is a pragma to change that |
| 09:48:36 | * | sagax quit (Ping timeout: 252 seconds) |
| 09:49:15 | FromGitter | <alehander42> @yyyc514 well it seems os is actually available in macros |
| 09:49:27 | FromGitter | <alehander42> so I can `walkDir` succesfully |
| 09:49:41 | FromGitter | <yyyc514> oh now that didn’t work for me at all |
| 09:49:59 | FromGitter | <yyyc514> share your example? |
| 09:50:15 | FromGitter | <yyyc514> i specifically asked on Git and was told it can’t work :) |
| 09:50:15 | PMunch | Hmm, casting doesn't work in macros? |
| 09:50:15 | FromGitter | <gogolxdong> what's the equivalent of this snippet in Nim ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc8575782893a2f3bfe0b86] |
| 09:50:37 | FromGitter | <yyyc514> pmunch: why wouldn’t it? |
| 09:50:45 | PMunch | I tried to create a distinct type and some procedures that worked on it but it seems to copy the data |
| 09:50:46 | FromGitter | <yyyc514> OH |
| 09:50:50 | FromGitter | <yyyc514> you mean in the code |
| 09:50:56 | FromGitter | <yyyc514> compiled code or VM code? |
| 09:51:02 | leorize | gogolxong: once `=sink` is a thing you can use `system.move`... |
| 09:51:11 | PMunch | yyyc514, VM code |
| 09:51:23 | PMunch | Just a sec, I'll create an example |
| 09:51:29 | FromGitter | <yyyc514> yeah that’s probably the problem: that’s it’s VM code |
| 09:51:30 | FromGitter | <alehander42> @yyyc514 well it works for me, I am not sure why |
| 09:51:36 | FromGitter | <yyyc514> i bet the typcasting has to do with C semantics |
| 09:51:38 | leorize | @gogolxdong see my prev msg |
| 09:51:58 | FromGitter | <gogolxdong> How long ago? |
| 09:52:14 | FromGitter | <alehander42> it shouldn't, casting works in the js backend |
| 09:52:21 | FromGitter | <alehander42> works well* |
| 09:52:46 | FromGitter | <yyyc514> JS also isn’t VM :) |
| 09:52:53 | FromGitter | <yyyc514> either the VM supports it or it doesn't |
| 09:52:54 | FromGitter | <gogolxdong> ah, it is just like a flash. |
| 09:54:04 | FromGitter | <alehander42> i mean that it isn't closely tied to C semantics. |
| 09:54:50 | FromGitter | <yyyc514> s/C/backend compiler/ :-) |
| 09:55:05 | FromGitter | <yyyc514> but lets see that example :) |
| 09:55:41 | PMunch | Well not casting, but conversion |
| 09:55:57 | FromGitter | <gogolxdong> Does it mean I have to implement `proc `=sink`()` |
| 09:57:20 | FromGitter | <gogolxdong> of specified type |
| 09:57:23 | leorize | Pretty much |
| 09:57:55 | leorize | once destructor support is mature that is |
| 10:01:37 | leorize | gogolxdong: you could try `system.move` directly, I haven't tested that |
| 10:04:03 | PMunch | This is what I was trying to do: http://ix.io/1psu/ |
| 10:04:51 | * | craigger quit (Ping timeout: 260 seconds) |
| 10:05:10 | * | craigger joined #nim |
| 10:13:16 | FromGitter | <yyyc514> thinking of how to do layouts and sub-templates now |
| 10:13:23 | FromGitter | <yyyc514> not sure could make it atll static at compile tile |
| 10:13:37 | FromGitter | <yyyc514> without making that an explciit goal |
| 10:14:59 | FromGitter | <yyyc514> take so much for granted in a fully dynamic language :) |
| 10:19:45 | FromGitter | <yyyc514> @alehander42 you going to share that file traversal code? please? |
| 10:25:29 | FromGitter | <alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc85f99c08b8b306732c5af] |
| 10:26:28 | FromGitter | <alehander42> this assumes a `render(path: static[string]): untyped` for constant paths is implemented |
| 10:28:16 | FromGitter | <alehander42> i am going to put this in a lib, because I need it for my Jester projects(and with some support for passing a hierarchy of view folders it can be reused internally by all nim web frameworks) |
| 10:28:47 | FromGitter | <alehander42> but feel free to tweak it for your code if it seems useful |
| 10:41:33 | * | elrood joined #nim |
| 10:48:22 | PMunch | poopBot, https://github.com/PMunch/gtkgenui is now updated to work on 0.19.0 |
| 10:50:30 | * | kapil____ joined #nim |
| 10:52:11 | FromGitter | <vegai> huh, interesting. The nim layer on my home spacemacs works fine, while my work machine’s spacemacs freezes when I try to use it |
| 10:52:21 | FromGitter | <vegai> almost identical configurations…. but I suppose not |
| 10:59:48 | poopBot | PMunch, ty will try |
| 11:01:03 | * | Vladar joined #nim |
| 11:06:49 | * | smt joined #nim |
| 11:07:57 | * | smt quit (Read error: Connection reset by peer) |
| 11:21:59 | FromGitter | <kaushalmodi> @vegai it's probably nimsuggest. There's a variable for setting nimsuggest path; set that to "". |
| 11:23:37 | * | dddddd joined #nim |
| 11:24:38 | * | vlad1777d joined #nim |
| 11:28:16 | PMunch | poopBot, cool. Let me know if you get stuck on anything |
| 11:28:18 | * | couven92 joined #nim |
| 11:28:37 | poopBot | PMunch, dont wory i will xD |
| 11:31:06 | * | hwallann joined #nim |
| 11:40:31 | * | hwallann quit (Quit: Leaving) |
| 11:40:43 | * | hwallann joined #nim |
| 11:43:42 | * | pwntus quit (Changing host) |
| 11:43:42 | * | pwntus joined #nim |
| 11:44:28 | * | poopBot quit (Remote host closed the connection) |
| 11:46:19 | * | poopBot joined #nim |
| 11:52:35 | poopBot | PMunch, this only Gtk2 or suports GTK3 too? |
| 11:54:21 | PMunch | I have a version that supports Gtk3, but that's not yet updated to 0.19.0 |
| 11:54:26 | PMunch | I have the fix for it at home though |
| 11:54:32 | PMunch | So I'll fix it tonight |
| 11:54:38 | poopBot | nice |
| 11:58:58 | poopBot | ../../../../.nimble/pkgs/gtkgenui-0.1.0/gtkgenui.nim(1, 16) Error: cannot open file: stack |
| 11:59:21 | poopBot | hmm, clang issue or what, i just copy pasted your example |
| 12:01:29 | PMunch | Hmm, stack.nim should be in the genui package |
| 12:01:52 | PMunch | Might be a Nimble issue |
| 12:02:43 | poopBot | i dide nimble install gtkgenui agian, i try manul delete then reisntall again |
| 12:03:09 | PMunch | Nah, looks like the Nimble project is wrong |
| 12:03:10 | PMunch | Hold on |
| 12:03:19 | poopBot | ok |
| 12:15:00 | PMunch | There |
| 12:15:09 | PMunch | nimble install gtkgenui should now work |
| 12:15:38 | poopBot | yep |
| 12:15:42 | poopBot | i did it |
| 12:16:58 | poopBot | it works now |
| 12:16:59 | poopBot | yey |
| 12:18:59 | PMunch | Woo |
| 12:24:09 | * | adeohluwa joined #nim |
| 12:53:29 | * | vlad1777d quit (Ping timeout: 244 seconds) |
| 13:07:31 | * | endragor quit (Remote host closed the connection) |
| 13:12:10 | * | stefanos82 joined #nim |
| 13:17:54 | * | kapil____ quit (Quit: Connection closed for inactivity) |
| 13:27:17 | PMunch | poopBot, getting anywhere? |
| 13:31:46 | * | endragor joined #nim |
| 13:36:13 | * | endragor quit (Ping timeout: 250 seconds) |
| 13:52:22 | * | stephan___ joined #nim |
| 13:54:26 | * | platoff quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 13:57:52 | * | shadowbane joined #nim |
| 14:08:30 | poopBot | PMunch, not really, head to go out just came back, first plan is to read all stuff you writed to get to know how it works |
| 14:09:18 | PMunch | Ah right |
| 14:09:55 | * | kapil____ joined #nim |
| 14:10:19 | poopBot | dont worry will try for sure |
| 14:10:24 | PMunch | :) |
| 14:11:43 | PMunch | It's basically just a DSL for Gtk, so you should be able to use Gtk documentation to figure out which widgets you want and then use the DSL to create and layout the widgets |
| 14:12:20 | poopBot | yep, i think i get how it works, tought i dont know what DSL stands for |
| 14:12:23 | poopBot | xD |
| 14:12:46 | PMunch | Domain Specific language |
| 14:12:59 | PMunch | Basically a custom syntax for a particular application |
| 14:14:14 | poopBot | yep i saw that, i was hmm, how can he do that like that :) , guess it has something to do with nim macros ,right? |
| 14:15:12 | * | Snircle joined #nim |
| 14:21:46 | * | stephan___ quit (Quit: Page closed) |
| 14:22:35 | PMunch | Yup |
| 14:23:39 | PMunch | Basically a Nim macro takes in the code you write as a complex tree structure, and return a similar tree structure. The genui macro takes the tree you give it, and rewrites it to a bunch of Gtk statements in it's output tree |
| 14:24:23 | PMunch | If you compile it with -d:debug it will show you what the genui macro generates |
| 14:32:10 | * | enthus1ast joined #nim |
| 14:36:44 | * | platoff joined #nim |
| 14:39:10 | poopBot | ok ty |
| 14:43:15 | poopBot | am reading now explanation, and i can say you really nicly wirited it so nowbie like me can undurstend it |
| 14:44:44 | PMunch | Good to hear :) |
| 14:44:48 | * | PrimHelios joined #nim |
| 14:45:03 | * | PrimHelios quit (Client Quit) |
| 14:53:55 | * | adeohluwa quit (Quit: Connection closed for inactivity) |
| 14:56:40 | * | craigger_ joined #nim |
| 14:56:54 | * | craigger quit (Ping timeout: 264 seconds) |
| 15:00:12 | * | PMunch quit (Quit: Leaving) |
| 15:02:37 | poopBot | now i remember why i hated gtk, just installed Glade to make mokup, working with GameEngines is so easyer |
| 15:09:13 | FromGitter | <kaushalmodi> @skellock I happened to visit your profile and saw that you have forked https://github.com/mlichvar/newt. Please tell me you are building a Nim lib to wrap that :) (For other folks not familiar, newt is a TUI library, looks like this: https://www.youtube.com/watch?v=HZLt2KT4aLA). |
| 15:14:04 | poopBot | i dident even know CLI uis like this exists |
| 15:15:11 | FromGitter | <kaushalmodi> poopBot: These are very old school (which I love). I love apps like ncdu and htop which use similar TUI elements but from the ncurses lib |
| 15:15:40 | * | craigger joined #nim |
| 15:15:55 | * | craigger_ quit (Ping timeout: 246 seconds) |
| 15:16:42 | poopBot | i have htop tought it doset look so pop like that 1 in video |
| 15:18:28 | * | Summertime joined #nim |
| 15:20:12 | * | narimiran joined #nim |
| 15:20:19 | * | Trustable joined #nim |
| 15:32:30 | Summertime | is it possible to have nimble create a binary with a dash in it? e.g. word-word instead of word_word |
| 15:33:14 | FromGitter | <kaushalmodi> No, I remember nimble throwing an error on detecting dash |
| 15:33:42 | leorize | the compiler will throw an error before you do :P |
| 15:35:35 | Summertime | I can pass -o:word-word to the compiler though :n (well, to nim c, at least) |
| 15:36:18 | leorize | oh, totally forgot about that flag :P |
| 15:36:28 | leorize | I don't think nimble have it though |
| 15:36:55 | FromGitter | <kaushalmodi> you can always create a NimScript task that does mvFile at the end |
| 15:37:51 | leorize | can you still do `nimble install` after that? |
| 15:38:06 | FromGitter | <kaushalmodi> unfortunately not |
| 15:38:19 | FromGitter | <kaushalmodi> I had opened an issue on nimble regarding that |
| 15:40:30 | Summertime | the obvious solution is to do the mv in the .nimble directory, and also edit the metadata for the package to reflect it / allow uninstalling |
| 15:41:09 | FromGitter | <kaushalmodi> that's too much of a hack; better to fix this in nimble by submitting a PR |
| 15:45:35 | * | Vladar quit (Remote host closed the connection) |
| 15:46:12 | planetis[m] | Iam confused with macros, what does `==`(a,b: NimNode) check differently than eqIdent(a, b: NimNode)? |
| 15:46:30 | planetis[m] | and when to use which? |
| 15:46:59 | * | nsf quit (Quit: WeeChat 2.2) |
| 15:47:42 | leorize | `==` is, uh when you want to check whether two NimNode is equivalent |
| 15:48:02 | leorize | eqIdent when you just want to compare their identifier |
| 15:50:08 | planetis[m] | i was comparing the return types of two procs and eqIdent returns false if they're both void |
| 15:50:16 | planetis[m] | but == works |
| 15:50:31 | planetis[m] | is it a bug? |
| 15:51:06 | leorize | no |
| 15:51:47 | planetis[m] | i see their both nnkEmpty |
| 15:51:54 | planetis[m] | not nnkIdent |
| 15:52:35 | * | darithorn joined #nim |
| 15:54:27 | planetis[m] | so does == check their childern, right? |
| 15:54:48 | planetis[m] | *children Nodes |
| 15:55:52 | planetis[m] | so I can just write proc1.params == proc2.params |
| 16:00:27 | planetis[m] | ? |
| 16:01:00 | leorize | try it yourself... |
| 16:02:35 | planetis[m] | I can trust a test, man I usally read the code I run but this is "magic" :p |
| 16:06:00 | * | d10n-work joined #nim |
| 16:35:28 | * | PMunch joined #nim |
| 16:51:39 | * | vlad1777d joined #nim |
| 17:10:45 | * | Trustable quit (Remote host closed the connection) |
| 17:16:47 | * | enthus1ast quit (Ping timeout: 250 seconds) |
| 17:19:52 | * | krux02 quit (Remote host closed the connection) |
| 17:27:24 | FromGitter | <vegai> why’s the repl a secret? |
| 17:34:02 | FromDiscord | <treeform> hey @dom96 I have figured out that SSL puzzle, but I don't know how to make a PR to fix the actual SSL so that it can be submitted? https://github.com/nim-lang/Nim/issues/9419#issuecomment-430809742 |
| 17:37:25 | FromGitter | <kaushalmodi> Talking about the RFC repo, Araq, @narimiran , what if the GitHub Projects feature is used for that? https://github.com/nim-lang/Nim/projects |
| 17:37:55 | narimiran | i've never used it, but maybe this could work |
| 17:38:09 | FromGitter | <kaushalmodi> In Projects, you can create filtered lists of RFCs, vs everything else |
| 17:38:19 | FromDiscord | <treeform> I though the RFC tag did that? https://github.com/nim-lang/Nim/issues?q=is%3Aissue+is%3Aopen+RFC+label%3ARFC |
| 17:38:27 | narimiran | but isn't that something like trello? |
| 17:38:42 | FromGitter | <kaushalmodi> The everything else can be further broken down based on tags like nimpretty, etc |
| 17:39:18 | FromGitter | <kaushalmodi> treeform: the projects will give a more organized v |
| 17:39:22 | FromGitter | <kaushalmodi> view |
| 17:39:39 | FromGitter | <kaushalmodi> narimiran: yes, like Trello |
| 17:40:39 | FromGitter | <kaushalmodi> Projects is not being used at all, so doesn't hurt creating a bunch of filters and seeing what it looks like. :). |
| 17:42:23 | FromGitter | <rayman22201> @vegai repl is a secret because it is buggy and not very well supported. |
| 17:43:07 | FromGitter | <rayman22201> This needs to be put in a FAQ somewhere lol. The question gets asked here at least once every few weeks. |
| 17:43:31 | narimiran | @kaushalmodi i'm not sure how RFCs fit in with trello-like workflow :/ |
| 17:49:01 | FromDiscord | <hotdog> "nim check" (used by vscode extension) is stuck, using 100% CPU every time I open my project. How can I debug what's causing it? |
| 17:55:13 | FromGitter | <vegai> @rayman22201 or perhaps make it “nim i” again since all that secrecy is more confusing :) |
| 17:55:23 | FromGitter | <kaushalmodi> nirimiran: OK, I misunderstand, looks like issue cards can only be moved one at a time |
| 17:55:32 | FromGitter | <kaushalmodi> .. about GitHub Projects |
| 17:59:22 | narimiran | @hotdog are you maybe using macros? |
| 18:00:03 | narimiran | i'm not sure it's the macros, but in some cases, i know that nimsuggest (used by that extension) uses 100% cpu and i must manually kill it |
| 18:02:09 | FromDiscord | <hotdog> @narimiran not using any macros at the moment (apart from some stuff in stdlib). I've been killing the process manually, but every time I save a file it starts up again |
| 18:05:18 | narimiran | restart vscode maybe? |
| 18:05:31 | FromDiscord | <hotdog> It happens after restart too |
| 18:05:58 | FromGitter | <Vindaar> does it also happen if you run `nim check` manually on the file in the terminal? |
| 18:06:18 | FromDiscord | <hotdog> Yeah, tried running it myself and it hangs too |
| 18:06:19 | FromDiscord | <hotdog> Running "nim check" myself I see a lot of "Error: This module only works on the JavaScript platform" (I'm outputting JS in this project), how can I let nim check know that? |
| 18:06:31 | narimiran | eh, such is that plugin. when it works, it works....okayish. when it doesn't, maybe kill/restart helps (it usually does for me), maybe not |
| 18:07:30 | FromDiscord | <hotdog> Yeah |
| 18:08:13 | FromDiscord | <hotdog> "Hint: <module> [Processing]" seems to get printed _after_ each module is processed, so I can't even tell which module it's stuck on |
| 18:09:06 | FromGitter | <kaushalmodi> hotdog: I don't use VSCode, but don't have a good suggestion related to that. But what if you simply remove nimsuggest from PATH? |
| 18:09:35 | narimiran | well, you can easily turn off nim check on save, in vscode settings |
| 18:09:45 | leorize | nimsuggest still has a long way to go... |
| 18:10:23 | FromDiscord | <hotdog> @kaushalmodi yeah I can turn it off, but I quite like having it if it works 😃 |
| 18:11:02 | FromGitter | <kaushalmodi> hotdog: oh OK. Probably, like many others, I just don't use it and life is still good |
| 18:13:21 | leorize | nimsuggest gives you that "auto-completion" thing tho... |
| 18:13:40 | narimiran | ...until it stops working ;) |
| 18:13:40 | leorize | although I've to admit that I never use it |
| 18:14:26 | FromDiscord | <hotdog> I'm sure everyone's workflow is a bit different 😃 |
| 18:14:48 | FromDiscord | <hotdog> I'm not complaining that it's not working, just wanted to know if anyone had any tips on how to debug it |
| 18:15:02 | FromGitter | <kaushalmodi> I normally tend to auto-complete from whatever is in existing open file buffers. Emacs hippie-expand just works wonders |
| 18:15:13 | FromGitter | <kaushalmodi> it's lang agnostic and Just Works(TM) |
| 18:15:23 | FromGitter | <Vindaar> ^this |
| 18:15:51 | leorize | @hotdog: attach a debugger, maybe? |
| 18:16:10 | leorize | kaushalmodi: same, but for vim |
| 18:21:51 | * | nsf joined #nim |
| 18:27:40 | * | Trustable joined #nim |
| 18:34:57 | * | zachk joined #nim |
| 18:38:58 | * | kapil____ quit (Quit: Connection closed for inactivity) |
| 18:43:57 | * | hoijui joined #nim |
| 18:49:08 | * | Trustable quit (Remote host closed the connection) |
| 18:54:27 | * | zachk quit (Changing host) |
| 18:54:27 | * | zachk joined #nim |
| 18:59:48 | PMunch | poopBot, fixed gtk3genui now |
| 19:06:56 | masterm | hm... I have the most peculiar Nim compilation error (compiler bug perhaps?) |
| 19:07:19 | masterm | I'm working on C bindings and I have a type definition like this: https://pastebin.com/j5zDTKHN |
| 19:07:38 | masterm | this gets me: |
| 19:08:25 | masterm | ../nim-amigaos/amigaos/graphics/gels.nim(34, 3) Error: redefinition of 'VSprite'; previous declaration here: ../nim-amigaos/amigaos/graphics/gels.nim(34, 2) |
| 19:09:06 | masterm | so nim claims that previous declaration is *one character* to the left, which doesn't make sense |
| 19:09:23 | masterm | it's the same line |
| 19:09:28 | PMunch | Hmm, that is indeed particular |
| 19:10:11 | masterm | the same type definition works when pasted to Nim playground so it's probably something else in my code but the msg is somehow wrong... I guess... |
| 19:10:31 | PMunch | Yeah, that snippet looks good to me |
| 19:19:43 | * | tiorock joined #nim |
| 19:19:43 | * | tiorock quit (Changing host) |
| 19:19:43 | * | tiorock joined #nim |
| 19:19:43 | * | rockcavera is now known as Guest59631 |
| 19:19:43 | * | tiorock is now known as rockcavera |
| 19:22:54 | * | Guest59631 quit (Ping timeout: 252 seconds) |
| 19:32:46 | poopBot | PMunch, can gtk3genui use CSS ? |
| 19:32:53 | PMunch | Yes |
| 19:32:54 | * | glassofethanol joined #nim |
| 19:33:01 | poopBot | nice |
| 19:33:16 | PMunch | It just generates Gtk3 calls |
| 19:33:40 | PMunch | So you should be able to apply CSS after it has created the GUI |
| 19:34:35 | poopBot | nice, i am still trying to figure out how to create layout using Glade, then will just implament same in gtk3genui |
| 19:36:03 | * | glassofethanol left #nim (#nim) |
| 19:37:09 | FromGitter | <yyyc514> is ShallowCOpy not consistent? |
| 19:37:15 | FromGitter | <yyyc514> i read that shallow might or might not be honored by the compiler |
| 19:37:27 | * | glassofethanol joined #nim |
| 19:38:06 | FromGitter | <yyyc514> i can’t even get the simple example working |
| 19:39:43 | * | glassofethanol left #nim (#nim) |
| 19:39:50 | * | hoijui quit (Ping timeout: 268 seconds) |
| 19:44:27 | * | stefanos82 quit (Quit: Quitting for now...) |
| 19:44:45 | * | dddddd quit (Ping timeout: 244 seconds) |
| 19:45:50 | FromGitter | <yyyc514> ok is there no newString ? i just assign it to a variable then assign it to a pointer if i want a pointer to a string? |
| 19:46:30 | FromGitter | <yyyc514> ah there is but it takes an int |
| 19:49:11 | PMunch | yyyc514, what are you trying to do? |
| 19:49:40 | FromGitter | <yyyc514> trying to have a buffer (pointer to string) that i can easily point to different strings and then have things write to them and easily swap the “destination" |
| 19:49:49 | FromGitter | <yyyc514> still hacking the code filter stuff |
| 19:50:10 | FromGitter | <yyyc514> i have it working but the only way to get a string seems to be to assign it to a variable then use addr |
| 19:50:22 | FromGitter | <yyyc514> which actually is fine but i was just wondering if there is another way to init a new string |
| 19:50:28 | FromGitter | <yyyc514> like new(newString(20)) or something |
| 19:50:33 | FromGitter | <yyyc514> to get a pointer |
| 19:52:11 | * | Jesin joined #nim |
| 19:52:36 | PMunch | Ah, not entirely sure |
| 19:52:47 | FromGitter | <yyyc514> i need a “capture” method inside a template that can capture things that aren’t intended for output yet… but the template only allows one output call… like “result.add" |
| 19:52:56 | * | cozachk joined #nim |
| 19:52:58 | FromGitter | <yyyc514> so if i make it buffer[].add then i can retarget it just by changing buffer |
| 19:53:02 | * | craigger_ joined #nim |
| 19:54:09 | FromGitter | <yyyc514> hence playing with string pointers |
| 19:54:15 | * | craigger quit (Ping timeout: 252 seconds) |
| 19:54:27 | PMunch | Strings are already pointers though.. |
| 19:54:33 | PMunch | So you can just assign a string to it |
| 19:54:42 | PMunch | Oh wait |
| 19:54:48 | FromGitter | <yyyc514> well no because i don’t need copy on write |
| 19:55:03 | PMunch | Yeah, so you need to update both places |
| 19:55:13 | PMunch | Well you have alloc |
| 19:55:15 | * | zachk quit (Ping timeout: 268 seconds) |
| 19:55:33 | FromGitter | <kayabaNerve> Can I pass varargs to C++? |
| 19:55:36 | * | craigger joined #nim |
| 19:55:42 | FromGitter | <yyyc514> how would one use alloc in Nim? |
| 19:55:44 | PMunch | kayabaNerve, yes |
| 19:55:50 | PMunch | yyyc514 alloc |
| 19:55:57 | FromGitter | <kayabaNerve> I'm trying to interface with vectors. This damn library made constructors private. |
| 19:56:09 | FromGitter | <yyyc514> i’m not that familiar with C allocation what would it look like in a full example line |
| 19:56:17 | FromGitter | <kayabaNerve> Basically, trying to pass an unknown amount of objects and use the `{}` vector constructor |
| 19:56:48 | PMunch | yyyc514, look up create in system |
| 19:56:53 | * | craigger- joined #nim |
| 19:57:00 | FromGitter | <kayabaNerve> But that only works if I can convert the seq to a varargs |
| 19:57:06 | FromGitter | <kayabaNerve> :/ |
| 19:57:59 | * | craigger_ quit (Ping timeout: 276 seconds) |
| 19:58:05 | FromGitter | <yyyc514> actually maybe it would work without pointers but it’d do a lot of extra copying if you changed context a lot |
| 19:58:13 | PMunch | kayabaNerve, I remember seeing a module using C varargs recently |
| 19:58:29 | PMunch | Can't quite remember what it was though, maybe see if you can find a printf wrapper? |
| 19:58:38 | FromGitter | <kayabaNerve> Thanks for the idea. I'll go figure it out |
| 19:59:02 | FromGitter | <kaushalmodi> Araq: I was playing with the nightlies fork.. looks like it builds fine (testing linux only for now), but deployment fails: https://travis-ci.org/kaushalmodi/nightlies/builds/443358860#L805 |
| 19:59:28 | FromGitter | <kaushalmodi> It's not clear what the nightlies repo has to do with "fatal: repository 'https://github.com/nim-lang/Nim/' not found" during deployment |
| 19:59:37 | FromGitter | <kaushalmodi> I am not setting that URL anywhere in travis config |
| 19:59:41 | PMunch | yyyc514, or create a ref object with your string in it |
| 19:59:48 | * | craigger_ joined #nim |
| 19:59:49 | PMunch | That way it will be shared |
| 20:00:48 | * | craigger quit (Ping timeout: 268 seconds) |
| 20:01:44 | * | craigger_ quit (Read error: Connection reset by peer) |
| 20:02:02 | FromGitter | <yyyc514> i’m not sure that’d work i bet that’d still do copy of write semantica |
| 20:02:19 | FromGitter | <yyyc514> var a = obj.some_string; a.add(blah) … a is now different |
| 20:02:33 | * | craigger- quit (Ping timeout: 245 seconds) |
| 20:02:40 | FromGitter | <yyyc514> i think you have to use pointers or the shallowcopy stuff (which i couldnt’ get working) |
| 20:03:12 | PMunch | type RefString = ref object; str: string |
| 20:03:23 | PMunch | Then pass RefString objects around |
| 20:03:32 | FromGitter | <yyyc514> why not just ref string then? |
| 20:03:41 | FromGitter | <yyyc514> i don’t think what you’re proposing will work |
| 20:03:42 | PMunch | I couldn't get that to work.. |
| 20:03:58 | FromGitter | <yyyc514> mine works if i assign all the “real” strings to variables first |
| 20:04:03 | * | craigger joined #nim |
| 20:04:04 | FromGitter | <yyyc514> oh i’m using ptr sorry |
| 20:04:06 | FromGitter | <yyyc514> not ref |
| 20:04:25 | FromGitter | <yyyc514> `ptr string` :-) |
| 20:04:28 | PMunch | If I understand correctly what you are trying to do I'm 90% sure my solution would work |
| 20:05:44 | FromGitter | <yyyc514> hmmm how do you manually “deref” something |
| 20:05:52 | PMunch | [] |
| 20:06:04 | FromGitter | <yyyc514> no the opposite |
| 20:06:08 | FromGitter | <yyyc514> the equiv of “Addr" |
| 20:06:19 | FromGitter | <yyyc514> i guess that’d be reffing it lol |
| 20:06:26 | PMunch | Wait.. |
| 20:06:29 | FromGitter | <yyyc514> a : ref string |
| 20:06:32 | FromGitter | <yyyc514> how do i assign a string to a |
| 20:06:38 | PMunch | Oh |
| 20:07:52 | FromGitter | <yyyc514> oh you might be right about the object ref :) |
| 20:07:55 | FromGitter | <yyyc514> from what i just read |
| 20:08:21 | PMunch | This seems to work: http://ix.io/1puZ |
| 20:10:29 | FromGitter | <alehander42> @yyyc514 why do you need `ptr`? this is very rarely useful in normal nim code |
| 20:10:42 | FromGitter | <yyyc514> well refs might work |
| 20:10:48 | FromGitter | <yyyc514> but the first string op is still a copy |
| 20:11:02 | FromGitter | <yyyc514> then copying the refs around gets you references |
| 20:11:06 | FromGitter | <yyyc514> a[] = c # gets you a copy |
| 20:11:52 | FromGitter | <alehander42> i see you intend to use it in templates, but I still don't understand why do you need this buffer |
| 20:12:08 | FromGitter | <yyyc514> i need to swap the target of result.add |
| 20:12:29 | FromGitter | <yyyc514> so result needs to be a ptr or ref |
| 20:12:32 | FromGitter | <yyyc514> i think ref works |
| 20:12:40 | FromGitter | <yyyc514> just allocate it and then i can pass the ref around |
| 20:12:55 | FromGitter | <yyyc514> want to support something like this |
| 20:12:58 | FromGitter | <alehander42> ok, without any more details of the macro you're trying to write I can't tell |
| 20:12:59 | FromGitter | <yyyc514> (https://files.gitter.im/nim-lang/Nim/6C2E/Screen-Shot-2018-10-18-at-4.12.44-PM.png) |
| 20:13:13 | FromGitter | <yyyc514> where special_script doesn’t render but is saved for say the layout template |
| 20:13:20 | FromGitter | <alehander42> but it seems very overengineered to me, I can't believe there is no simpler solution |
| 20:13:50 | FromGitter | <yyyc514> seems pretty simple to me but maybe i’m missing something :) |
| 20:13:56 | PMunch | Well if it's a macro you can pas a symbol |
| 20:15:03 | FromGitter | <yyyc514> not sure i follow |
| 20:15:35 | FromGitter | <yyyc514> @alehander42 I have a template that’s writing to a buffer… but for something like “content_for” i need to change the target of the writes for a bit and then keep track of the sections like that for later use |
| 20:15:58 | FromGitter | <yyyc514> and with the build in code filters all you’re allowed is a single “emit" |
| 20:16:20 | FromGitter | <yyyc514> so it’d seemed easiest to emit to a ref string and then you could just change which string you’re emitting to easily |
| 20:16:55 | FromGitter | <alehander42> I see, I haven't played with code filters, it seems more reasonable now, sorry |
| 20:17:22 | FromGitter | <yyyc514> trying to make them stretch as far as possibl until i swap them with my own thing |
| 20:17:28 | FromGitter | <yyyc514> still struggling with how dynamic to make it |
| 20:18:13 | FromGitter | <alehander42> but still it seems you can change what emit does |
| 20:18:22 | FromGitter | <alehander42> it is `result.add` by default |
| 20:18:30 | FromGitter | <yyyc514> yep |
| 20:18:34 | FromGitter | <yyyc514> so i’d make it buffer.add :-) |
| 20:18:53 | FromGitter | <yyyc514> and then have content_for be a template that changes the buffer pointer at the beginning and end |
| 20:19:09 | FromGitter | <yyyc514> writes into a table of strings, etc |
| 20:19:31 | FromGitter | <yyyc514> which then the layout could easily grab |
| 20:19:47 | FromGitter | <alehander42> isn't emit called for each line? |
| 20:19:55 | FromGitter | <yyyc514> each line of HTML |
| 20:20:01 | FromGitter | <yyyc514> yes |
| 20:20:13 | FromGitter | <alehander42> ok, can't you have a custom proc templateAdd |
| 20:20:27 | FromGitter | <alehander42> which detects if the stripped line starts with content_for or something like this |
| 20:20:42 | FromGitter | <alehander42> and only then adds it to a different buffer |
| 20:20:46 | FromGitter | <yyyc514> well at that point yuo’re writing your own template parser from scratch |
| 20:20:58 | FromGitter | <alehander42> in this case you can just maintain a table of strings |
| 20:21:00 | FromGitter | <yyyc514> when the built in stuff it’s reading it and rewriting it in place |
| 20:21:13 | FromGitter | <yyyc514> you can’t “detect” a line |
| 20:21:23 | FromGitter | <yyyc514> only have the line do something by being a proc/template/macro etc |
| 20:21:24 | FromGitter | <alehander42> well if it's simpler to write a template engine |
| 20:21:27 | FromGitter | <alehander42> I'd do that :D |
| 20:21:46 | FromGitter | <yyyc514> i’ll never get around to actually making the HTTP part work if i do down every rabbit hole :) |
| 20:21:55 | FromGitter | <alehander42> that's why I like dsl-s like the karax one, much simpler to add this kind of functionality |
| 20:22:05 | FromGitter | <yyyc514> i want an ERB like one |
| 20:22:08 | FromGitter | <yyyc514> <%= syntax |
| 20:22:18 | FromGitter | <yyyc514> but otherwise the same idea as the built in |
| 20:22:55 | FromGitter | <yyyc514> conver to source code and pass that to a parseStatement or wahtever |
| 20:23:16 | FromGitter | <alehander42> well honestly I think writing a simple ERB-like template syntax and generating a nim file from it would take less time |
| 20:23:32 | FromGitter | <alehander42> but I can be wrong |
| 20:23:55 | FromGitter | <yyyc514> well that’s one approach… generate the nim template file vs generate the source code directly |
| 20:24:22 | FromGitter | <yyyc514> i guess technically you could still do both inside the compiler single pass |
| 20:25:00 | FromGitter | <alehander42> yeah, you can generate the code in memory and parseExpr it |
| 20:25:24 | FromGitter | <alehander42> I just parseExpr my karax templates |
| 20:25:32 | FromGitter | <yyyc514> what is karax? |
| 20:25:50 | FromGitter | <alehander42> a SPA framework |
| 20:25:56 | FromGitter | <alehander42> but it includes a html-building dsl |
| 20:26:21 | Araq | kaushalmodi: thanks for working on this, sadly the error is mysterious to me too |
| 20:26:24 | FromGitter | <yyyc514> oh yes i think you showed me that |
| 20:26:30 | FromGitter | <yyyc514> like HAML from ruby :) |
| 20:26:39 | FromGitter | <yyyc514> never cared for HAML though i know it’s like a religious war type topic :) |
| 20:27:00 | FromGitter | <yyyc514> i think progammers like that more and front-end people like HTML better, but i could be mistaken |
| 20:27:11 | FromGitter | <yyyc514> i’ve never minded HTML that much |
| 20:27:25 | FromGitter | <yyyc514> but then i also don’t like writing “end” in Ruby :) |
| 20:27:32 | FromGitter | <yyyc514> so i don’t have an issue with needing to “pair” elements :) |
| 20:27:49 | FromGitter | <yyyc514> i find i get lost sometimes in Nim/Python though editors can help with that |
| 20:28:37 | FromGitter | <alehander42> exactly, I don't like the need for closing tags in html/xml |
| 20:29:03 | FromGitter | <alehander42> well, one should strive to have relatively short functions/blocks of code |
| 20:29:05 | FromGitter | <yyyc514> yeah i think that’s the common gripe, mismatched tags - i’ve just never had a huge problem with that in practice |
| 20:29:19 | FromGitter | <yyyc514> > well, one should strive to have relatively short functions/blocks of code |
| 20:29:23 | FromGitter | <yyyc514> dat true too |
| 20:29:25 | FromGitter | <alehander42> it's not even that, it's just babysitting the syntax |
| 20:29:42 | FromGitter | <alehander42> with indentation the parser can figure out what is the tree of the code anyway |
| 20:29:44 | FromGitter | <yyyc514> i get that :) that’s why frameworks should support both options :) |
| 20:29:50 | FromGitter | <alehander42> yep |
| 20:29:54 | FromGitter | <yyyc514> so you just drop in yoru fav template engine and away you go :) |
| 20:30:00 | FromGitter | <yyyc514> or write your own :) |
| 20:30:05 | FromGitter | <alehander42> that's the spirit :) |
| 20:30:21 | FromGitter | <yyyc514> but don’t yell if the default is HTML ish :) |
| 20:30:38 | FromGitter | <yyyc514> for whatever default means, hopefully it’d be loosely coupled |
| 20:31:03 | FromGitter | <yyyc514> i’m thinking render might take a static string and then layout might be done at runtime |
| 20:31:22 | FromGitter | <yyyc514> render :index; layout = variable_that_is_a_layout_render_proc |
| 20:31:35 | FromGitter | <yyyc514> so you can dynamically change the layout at runtime |
| 20:31:45 | FromGitter | <yyyc514> but you have to kind of decide upfront which template you’re rendering |
| 20:32:02 | FromGitter | <yyyc514> or hard code it (if x; render :index; else render :oh_no; ) |
| 20:32:21 | FromGitter | <yyyc514> then all layout has to do is add a bunch of strings together |
| 20:32:28 | FromGitter | <alehander42> I haven't really thought how layouts would fit in this |
| 20:32:45 | FromGitter | <yyyc514> yeah Rails is big on layouts… like your overall site design so you don’t repeat it |
| 20:33:01 | FromGitter | <alehander42> in my solution this `if x: render :index else: render other` is autogenerated by a macro |
| 20:33:05 | FromGitter | <yyyc514> so at the simplest you have layouts and templates and your “main view” fits inside the layout |
| 20:33:14 | FromGitter | <alehander42> but I have to think how to combine it with layouts |
| 20:33:36 | FromGitter | <yyyc514> autogenerated how? |
| 20:33:44 | FromGitter | <yyyc514> like what code is the user writing in the controller |
| 20:34:53 | * | hoijui joined #nim |
| 20:35:02 | FromGitter | <alehander42> nope, I am talking about passing a variable path to render, as my templates are simpler |
| 20:35:38 | FromGitter | <yyyc514> you mean if there is a constrained list? i’m just not imaginign what you’re saying in practice |
| 20:35:43 | FromGitter | <yyyc514> or only seeing it very abstractly |
| 20:35:50 | FromGitter | <alehander42> a sec |
| 20:36:18 | FromGitter | <yyyc514> the only thing is if a template becomes a proc then it can’t access variables setup in the controller |
| 20:36:28 | FromGitter | <yyyc514> which views can because they are tempaltes running inline in the controller |
| 20:36:58 | FromGitter | <yyyc514> which would lead to hacks like passing things to the layout via the view |
| 20:37:06 | FromGitter | <yyyc514> not sure how often that would come up |
| 20:37:10 | FromGitter | <alehander42> well that's not fatal, because I can just use a template instead of proc |
| 20:37:34 | FromGitter | <yyyc514> well but then you can’t have a pointer to it at runtime |
| 20:37:50 | FromGitter | <yyyc514> i’m really feeling like layout should be a runtime choice (for me) |
| 20:37:52 | FromGitter | <alehander42> but you can collect all possible layouts |
| 20:37:56 | FromGitter | <yyyc514> hmmm |
| 20:38:01 | FromGitter | <alehander42> and generate something |
| 20:38:05 | FromGitter | <alehander42> that invokes them based on a string |
| 20:38:13 | FromGitter | <alehander42> similar to my runtime render solution |
| 20:38:39 | FromGitter | <yyyc514> i kind of get that but the invokation happens separately from the setting |
| 20:38:54 | FromGitter | <yyyc514> you’re saying i just generate all the possibilites in a case |
| 20:39:02 | FromGitter | <yyyc514> and then the run-time render calls that |
| 20:39:17 | FromGitter | <yyyc514> passing it the run-time value of layout which is resolved there |
| 20:40:18 | FromGitter | <yyyc514> and then they could be templates again and have the access back |
| 20:40:25 | FromGitter | <alehander42> well I am not sure how exactly the layouts worked in rails |
| 20:40:29 | FromGitter | <yyyc514> man the C code this stuff generates much be really ugly :) |
| 20:40:32 | FromGitter | <alehander42> i have to take a look at an example |
| 20:40:37 | FromGitter | <yyyc514> layouts just wrap views |
| 20:40:42 | FromGitter | <yyyc514> and yield when they want content |
| 20:40:50 | FromGitter | <yyyc514> <%= yield %> gives you the main content |
| 20:40:54 | FromGitter | <alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc8efd5c08b8b306736970d] |
| 20:41:06 | FromGitter | <yyyc514> <% yield js_snippter %> would give you some JS snippet you setup earlier in the view but didn’t want to render in the view |
| 20:41:21 | FromGitter | <alehander42> this was the code that collects render and generates a ⏎ ⏎ `````` [https://gitter.im/nim-lang/Nim?at=5bc8eff11e23486b93df05fb] |
| 20:41:26 | FromGitter | <yyyc514> and the walkDir stuff all works for you? |
| 20:41:30 | FromGitter | <yyyc514> you on nim 0.19 whatever? |
| 20:41:48 | FromGitter | <alehander42> yes, I always use one of the last versions |
| 20:41:55 | FromGitter | <yyyc514> i’ll try it again |
| 20:42:07 | FromGitter | <alehander42> everything before that is basically EOL |
| 20:42:12 | FromGitter | <yyyc514> sounds like that’d be the way to have the cake and eat it too |
| 20:42:16 | FromGitter | <alehander42> (until we get to 1.0) |
| 20:42:47 | FromGitter | <yyyc514> so which nim file do all your tempaltes get compiled to? |
| 20:43:16 | FromGitter | <yyyc514> i guess it could just be one big one you import everywhere but i don’t know if that would slow things down more than breaking it out |
| 20:43:23 | FromGitter | <yyyc514> for me views are OFTEN going to be scoped to controller |
| 20:43:37 | FromGitter | <alehander42> they don't compile to any file |
| 20:43:46 | FromGitter | <alehander42> whenever I need one |
| 20:43:59 | FromGitter | <alehander42> I just call `render("name of template without .nim")` |
| 20:44:11 | FromGitter | <alehander42> or `render(variable)` |
| 20:44:30 | FromGitter | <yyyc514> ok and that drops ALL your templates there |
| 20:44:36 | FromGitter | <yyyc514> in each spot |
| 20:44:45 | FromGitter | <alehander42> well it drops the template under this path |
| 20:44:54 | FromGitter | <alehander42> I have a single template directory |
| 20:44:57 | FromGitter | <alehander42> so no name clashes |
| 20:45:15 | FromGitter | <yyyc514> i mean everytime you call render you’re adding all the code to render ANY template yes? because of your case |
| 20:45:21 | FromGitter | <alehander42> e.g. main_view.nim chat_view.nim etc |
| 20:45:31 | FromGitter | <alehander42> and if i call render("main_view") I get only this template |
| 20:45:45 | FromGitter | <alehander42> well I add the code to render this particular template |
| 20:45:46 | FromGitter | <alehander42> only |
| 20:45:46 | * | nsf quit (Quit: WeeChat 2.2) |
| 20:45:51 | FromGitter | <yyyc514> you only get it RENDERED, but it’s still compiled (subject to dead code elimitation), yes? |
| 20:46:11 | FromGitter | <alehander42> well the case happens only if I call it with a variable |
| 20:46:15 | * | narimiran quit (Quit: Konversation terminated!) |
| 20:46:17 | FromGitter | <alehander42> which should be rare |
| 20:46:25 | FromGitter | <yyyc514> ah ok… you have both cases? |
| 20:46:31 | FromGitter | <alehander42> yes, that's the point |
| 20:46:39 | FromGitter | <yyyc514> right if you call it static it’d just inline the appropriate thing :) |
| 20:46:41 | FromGitter | <alehander42> the case version just invokes the constant string `render |
| 20:47:06 | FromGitter | <alehander42> this way I can reuse this for any template system that defines |
| 20:47:14 | FromGitter | <alehander42> a `render(path: static[string])` |
| 20:47:35 | FromGitter | <alehander42> which might be a bit limited for now, but it's fine for my usecases |
| 20:49:53 | FromGitter | <yyyc514> yeah i might be overengineering a bit but i’m learning a lot about Nim :) |
| 20:51:59 | FromGitter | <alehander42> great to hear :) |
| 20:53:27 | * | dddddd joined #nim |
| 20:55:05 | FromGitter | <yyyc514> HA |
| 20:55:06 | FromGitter | <yyyc514> Error: cannot 'importc' variable at compile time |
| 20:55:20 | FromGitter | <yyyc514> @alehander42 walkDir vs walkDirs |
| 20:55:26 | FromGitter | <yyyc514> walkDirs hits POSIX and is unsafe |
| 20:55:33 | FromGitter | <yyyc514> walkDir evidentally does something else entirely :) |
| 20:55:38 | FromGitter | <yyyc514> which works inside the VM |
| 20:55:43 | FromGitter | <yyyc514> i’m going to update my post on Github |
| 20:56:43 | * | sagax joined #nim |
| 20:57:12 | FromGitter | <alehander42> ah I see, well the pattern version is probably trickier |
| 20:57:24 | FromGitter | <alehander42> walkDir just takes a path, not a pattern |
| 20:57:54 | * | craigger_ joined #nim |
| 20:58:05 | FromGitter | <yyyc514> https://github.com/nim-lang/Nim/issues/9253 |
| 20:58:07 | * | craigger quit (Ping timeout: 244 seconds) |
| 20:58:16 | FromGitter | <yyyc514> yeah i see the diff :) |
| 20:58:19 | FromGitter | <yyyc514> THANKS! |
| 20:58:24 | FromGitter | <yyyc514> this helps a lot |
| 20:58:33 | FromGitter | <yyyc514> since now i can just collet the templates without having to import them specially |
| 20:58:47 | FromGitter | <yyyc514> more stupid boilerplate gone :) |
| 21:00:12 | FromGitter | <yyyc514> are you doing include vs import for templates? |
| 21:01:01 | FromGitter | <yyyc514> oh i get the “top level” error when i try to import/include the template at the call-site |
| 21:02:08 | FromGitter | <yyyc514> so how are you loading the templates themselves? i don’t see that in the code you pasted |
| 21:02:55 | FromGitter | <yyyc514> nim scares me cause i’m often importing the saem thing to every file and though i know it’s not getting confused i feel like it’s probably doing a lot of hard work to filter that all out and make sense of it |
| 21:02:58 | FromGitter | <yyyc514> something like strutils, etc |
| 21:03:25 | FromGitter | <yyyc514> i guess if the whole compiler builds in 12 seconds i’m probably good for a while, but still :) |
| 21:05:59 | FromGitter | <alehander42> well I am staticRead-ing my templates |
| 21:06:04 | FromGitter | <alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bc8f5bc38449236615efbd3] |
| 21:06:18 | FromGitter | <alehander42> but that's because they contain code |
| 21:07:03 | FromGitter | <alehander42> I mean I parseExpr them because they contain code, this way I have more control on what to do with that code |
| 21:07:16 | FromGitter | <alehander42> e.g. I can define a special proc instead or something else |
| 21:07:37 | FromGitter | <alehander42> with include I have to make sure the file itself contains something exported |
| 21:07:46 | FromGitter | <yyyc514> ah yeah i don’t think i can staticRead AND pass it to the code filter engine, i’d need to write my own engine |
| 21:07:57 | FromGitter | <yyyc514> i was trying to do an import_all macro but can’t because of the top-level things |
| 21:08:03 | FromGitter | <yyyc514> *trying to import each one in a loop) |
| 21:08:35 | FromGitter | <yyyc514> although i can swear i got this to work before |
| 21:08:39 | FromGitter | <yyyc514> if the macro was at the top level |
| 21:09:10 | FromGitter | <alehander42> I have no idea if parseExpr applies a code filter |
| 21:09:16 | FromGitter | <yyyc514> OH the import can’t be run from the macro lol :) |
| 21:09:20 | FromGitter | <yyyc514> it has to be generated AST :) |
| 21:09:21 | FromGitter | <yyyc514> one sec |
| 21:20:05 | FromGitter | <yyyc514> and now i can’t get it to work at all, grrrr |
| 21:21:00 | FromGitter | <yyyc514> it’s like my macro isn’t running at all |
| 21:21:31 | FromGitter | <yyyc514> ok it is just walkDir is doing nothing |
| 21:24:27 | * | hoijui quit (Ping timeout: 252 seconds) |
| 21:27:06 | FromGitter | <yyyc514> lkike walkDir is returning nothing |
| 21:33:04 | FromGitter | <yyyc514> it’s relative from current dir, weird |
| 21:33:15 | FromGitter | <yyyc514> not sure why it doesn’t throw an error if the directoy can’t be found - thought it did that earlier |
| 21:33:19 | FromGitter | <yyyc514> now it just does nothing silently |
| 21:33:53 | * | PMunch quit (Remote host closed the connection) |
| 21:40:17 | * | arecaceae quit (Remote host closed the connection) |
| 21:40:31 | FromGitter | <yyyc514> annoying that walkDir and import and running in different dirs :) |
| 21:41:54 | FromGitter | <citycide> can you constrain the parameter on a `proc {.importcpp.}` (js backend) to eg. one of 4 strings/cstrings? thought an `enum V = "V"` might do it but nah |
| 21:43:09 | FromGitter | <kaushalmodi> Araq: Success! with nightly deployment |
| 21:43:30 | FromGitter | <kaushalmodi> anyone up to test the nightly built release? I tested it fine on RHEL (Linux) |
| 21:43:36 | FromGitter | <kaushalmodi> any macOS/Win testers? |
| 21:44:16 | FromGitter | <kaushalmodi> Test nightly release: https://github.com/kaushalmodi/nightlies/releases/tag/untagged-c475d82c22c32f112cf9 |
| 21:45:05 | FromGitter | <kaushalmodi> download, extract https://github.com/kaushalmodi/nightlies/releases/download/untagged-c475d82c22c32f112cf9/nim-0.19.1.tar.xz |
| 21:45:29 | FromGitter | <kaushalmodi> the archive is a surprise zipbomb.. 3.xMB extracts to ~500MB |
| 21:51:09 | Araq | zipbomb ... we probably need to do something about that |
| 21:51:37 | Araq | the problem is we precompute the C sources for all CPU/OS combinations somebody ever was interested in |
| 21:52:06 | Araq | and they zip well, but don't "deduplicate" well anymore, even though deduplication has been implemented too |
| 21:53:09 | Araq | kaushalmodi: there is also a test for unzip + build.sh + tests via 'koch testinstall' |
| 21:54:00 | FromGitter | <yyyc514> any consistent naming for internalish variables? |
| 21:54:04 | Araq | which we can run if we don't already. anyway it's late here, will test the binaries tomorrow but create a PR already please |
| 21:54:05 | FromGitter | <yyyc514> i can’t use an underscore evidentaly :) |
| 21:54:29 | Araq | yyyc514: 'Impl' suffix is common for procs |
| 21:55:35 | FromGitter | <yyyc514> no this is a variable |
| 21:55:59 | FromGitter | <kayabaNerve> If anyone here is interested in BLS signatures, I just finished my lib for it. https://github.com/EmberCrypto/BLS |
| 21:56:16 | FromGitter | <yyyc514> maybe i’ll remove the inject and go back to leeting the compiler name it |
| 21:57:02 | FromGitter | <kaushalmodi> Araq: I wonder how koch testinstall works.. we will be deploying the archive only from one OS |
| 21:57:18 | FromGitter | <kaushalmodi> so we can testinstall only on the OS we are deploying from |
| 21:57:23 | Araq | well read koch.nim to find out |
| 21:57:42 | Araq | not everything written by me is hard to read... |
| 21:57:45 | FromGitter | <kaushalmodi> I didn't mean "how testinstall works" |
| 21:57:45 | FromGitter | <kayabaNerve> Read the source. It's like RTFM but for when you don't have proper docs. |
| 21:58:00 | FromGitter | <kayabaNerve> :thinking: |
| 21:58:14 | FromGitter | <kaushalmodi> My question is a practical one |
| 21:58:31 | FromGitter | <kaushalmodi> we are deploying one archive that can extract on all OSes |
| 21:58:42 | FromGitter | <kaushalmodi> so we can pick one of linux/osx to build that |
| 21:59:21 | FromGitter | <kaushalmodi> then running testinstall makes sense only on the OS whose built version we are deploying |
| 22:00:00 | FromGitter | <kaushalmodi> as I am deploying linux built archive (which technically should extract fine on osx too), running testinstall on linux won't prove that |
| 22:00:18 | FromGitter | <kaushalmodi> and running testinstall on osx on Travis makes no sense as we are not deploying the version built there |
| 22:01:14 | FromGitter | <kaushalmodi> so.. we need a "tester" repo that picks up the nightly built archive and tests on all OSes :P |
| 22:03:00 | Araq | yup |
| 22:03:11 | Araq | but 'koch testinstall' is better than nothing |
| 22:03:40 | FromGitter | <yyyc514> anyone else using Sublime? |
| 22:03:48 | Araq | also ... we can consider to avoid the zip bomb by producing nim_linux.tar.xz, nim_osx.tar.xz, nim_other.tar.xz |
| 22:04:06 | FromGitter | <yyyc514> my plugin is acting funny now because the path it uses to resolve the code when editing isn’t the same as when it’s compiling so my macro that walksDir is failing to compile inside the editor |
| 22:04:41 | FromGitter | <rayman22201> @Araq I think that is the solution I would favor. different zip packages for each main OS. |
| 22:04:58 | FromGitter | <rayman22201> @kaushalmodi trying the zip on windows now |
| 22:09:03 | FromDiscord | <hotdog> I found what was causing nim check to hang |
| 22:09:25 | FromDiscord | <hotdog> Here's the minimal example: https://ghostbin.com/paste/44wmm |
| 22:09:30 | FromGitter | <kaushalmodi> Araq: os-specific archives would be nice |
| 22:09:52 | FromDiscord | <hotdog> It compiles fine, but "nim check" will hang indefinitely, using 100% CPU |
| 22:11:45 | * | Tortice quit (Ping timeout: 252 seconds) |
| 22:27:13 | * | cozachk quit (Quit: Leaving) |
| 22:28:40 | FromGitter | <alehander42> hotdog, you are making me hungry |
| 22:28:51 | FromGitter | <alehander42> @yyyc514 I use sublime |
| 22:29:25 | FromGitter | <alehander42> are you talking about jump to? |
| 22:29:57 | FromGitter | <kaushalmodi> @rayman22201 ⏎ ⏎ > trying the zip on windows now ⏎ ⏎ what happened then? :P [https://gitter.im/nim-lang/Nim?at=5bc90964f659e67772da3790] |
| 22:30:40 | * | elrood quit (Quit: Leaving) |
| 22:30:49 | FromGitter | <rayman22201> sorry. Got distracted |
| 22:31:32 | FromGitter | <kaushalmodi> no worries. I am signing off for the day |
| 22:32:44 | FromGitter | <rayman22201> building now. I'll let you know. take it easy! @kaushalmodi |
| 22:40:05 | FromGitter | <yyyc514> https://github.com/Varriount/NimLime/issues/110 if anyone can help |
| 22:40:34 | FromGitter | <yyyc514> @alehander42 my magic import_all is failing the built in check because it’s not running the compiler from the root of the project evidentaly |
| 22:40:38 | FromGitter | <yyyc514> the issue explains it i think |
| 22:41:52 | FromGitter | <yyyc514> battery almost dead goign to have to head home |
| 22:42:30 | FromGitter | <yyyc514> maybe a nim.cfg seting can fix this? |
| 22:44:47 | FromGitter | <yyyc514> maybe not https://forum.nim-lang.org/t/3853 |
| 22:45:44 | FromGitter | <yyyc514> bbiab |
| 22:46:38 | FromGitter | <rayman22201> @kaushalmodi and @araq the nightly zip worked for me on windows with no problems 😀 👍 |
| 22:47:24 | * | endragor joined #nim |
| 22:47:27 | FromGitter | <kaushalmodi> 👍 |
| 22:50:44 | FromGitter | <alehander42> well for relative imports you need to use "path" |
| 22:50:48 | FromGitter | <alehander42> @yyyc514 ^ |
| 22:52:14 | * | endragor quit (Ping timeout: 268 seconds) |
| 23:12:55 | * | zachk joined #nim |
| 23:14:40 | * | zachk quit (Read error: Connection reset by peer) |
| 23:14:58 | * | zachk joined #nim |
| 23:21:59 | * | zachk quit (Changing host) |
| 23:21:59 | * | zachk joined #nim |
| 23:23:09 | FromGitter | <yyyc514> @alehander42 the problem is with walkDir not the imports |
| 23:23:12 | * | rockcavera quit (Read error: Connection reset by peer) |
| 23:23:15 | FromGitter | <yyyc514> the imports work if done manually |
| 23:23:28 | * | rockcavera joined #nim |
| 23:23:28 | * | rockcavera quit (Changing host) |
| 23:23:28 | * | rockcavera joined #nim |
| 23:23:52 | FromGitter | <yyyc514> i have ./app in my path |
| 23:24:20 | FromGitter | <yyyc514> walk dir seems to be local to the directory the compiler has as it’s PWD |
| 23:24:32 | FromGitter | <yyyc514> not the directory of the file in question |
| 23:25:30 | * | craigger joined #nim |
| 23:27:39 | * | craigger_ quit (Ping timeout: 252 seconds) |
| 23:28:32 | FromGitter | <alehander42> then just use |
| 23:28:40 | FromGitter | <alehander42> currentSourcePath & relativePath |
| 23:28:49 | * | darithorn quit () |
| 23:35:26 | FromGitter | <yyyc514> hmmm |
| 23:37:44 | * | zachk quit (Quit: Leaving) |
| 23:42:03 | FromGitter | <yyyc514> where is relativepath documented |
| 23:47:05 | FromDiscord | <treeform> hmm how can this fail: https://gist.github.com/treeform/4ed4e280143a7f751f67dc93a0321176 |
| 23:47:28 | FromDiscord | <treeform> it fails on this line: ` let alen = a.len` |
| 23:47:38 | FromGitter | <yyyc514> @alehander42 walkDir doesn’t like absolutely path |
| 23:47:52 | FromDiscord | <treeform> if a is null it returns 0, a is not null it returns len, how can it sigfault? |
| 23:56:28 | FromGitter | <yyyc514> @alehander42 thanks i got currentSourcePath working |
){kind=link}
){kind=link}
){kind=link}
){kind=link}