| 00:00:09 | Araq | I'm patient, no worries |
| 00:00:37 | Araq | misbehaviour: it produces "variable not initialized" even though it is |
| 00:00:50 | Araq | for some program of minor complexity |
| 00:01:01 | FromGitter | <ervinbosenbacher> :D |
| 00:01:04 | FromGitter | <ervinbosenbacher> ok |
| 00:01:20 | Araq | it's not critical at all, I wondered if you had experience with these algorithms |
| 00:01:42 | FromGitter | <ervinbosenbacher> i was looking into them sometime ago yes |
| 00:02:17 | FromGitter | <ervinbosenbacher> i will look into it |
| 00:02:45 | FromGitter | <ervinbosenbacher> building nim for jetson should be ok |
| 00:02:47 | Araq | and be gentle, it's my first DFA :P |
| 00:03:21 | Araq | only did control flow agnostic analyses before ;-) |
| 00:03:38 | FromGitter | <ervinbosenbacher> funny |
| 00:03:55 | FromGitter | <ervinbosenbacher> I was planning to build a FSM next week in Nim |
| 00:04:00 | FromGitter | <ervinbosenbacher> as a first exercise |
| 00:04:19 | FromGitter | <ervinbosenbacher> Then I will look into your DFA implementation |
| 00:05:24 | Araq | actually that's not true, but my other control flow dependent analysis are more adhoc |
| 00:05:50 | * | ehmry quit (Read error: Connection reset by peer) |
| 00:05:51 | * | unclechu quit (Read error: Connection reset by peer) |
| 00:05:52 | * | macsek1911[m] quit (Read error: Connection reset by peer) |
| 00:05:53 | * | Jipok[m] quit (Read error: Connection reset by peer) |
| 00:05:54 | * | dyce[m] quit (Remote host closed the connection) |
| 00:05:55 | * | hohlerde quit (Read error: Connection reset by peer) |
| 00:05:56 | * | byteflame quit (Remote host closed the connection) |
| 00:05:56 | * | hiway quit (Remote host closed the connection) |
| 00:05:56 | * | notdekka[m] quit (Remote host closed the connection) |
| 00:05:58 | * | TheManiac quit (Remote host closed the connection) |
| 00:05:58 | * | shashlick quit (Remote host closed the connection) |
| 00:05:58 | * | Demos[m] quit (Read error: Connection reset by peer) |
| 00:05:59 | * | Miguelngel[m] quit (Read error: Connection reset by peer) |
| 00:05:59 | * | planetis[m] quit (Read error: Connection reset by peer) |
| 00:06:20 | Araq | since Nim doesn't have 'goto' things are usally easy enough. this time I tried to follow the literature more closely though. |
| 00:07:00 | FromGitter | <ervinbosenbacher> ok |
| 00:07:00 | * | shashlick joined #nim |
| 00:07:01 | shashlick | https://github.com/genotrance/ff |
| 00:07:02 | shashlick | new CLI tool for Windows built using Nim, feedback appreciated! |
| 00:09:20 | * | kunev quit (Ping timeout: 260 seconds) |
| 00:10:19 | * | dandevelo joined #nim |
| 00:11:06 | * | kunev joined #nim |
| 00:14:05 | * | hiway joined #nim |
| 00:14:59 | * | dandevelo quit (Ping timeout: 255 seconds) |
| 00:15:57 | * | user0 quit (Quit: user0) |
| 00:16:15 | * | user0 joined #nim |
| 00:17:14 | Araq | shashlick: nice, did you use the improvements we merged into terminal.nim? |
| 00:18:11 | FromGitter | <tim-st> @Araq pls add an "add" proc to set/table |
| 00:18:21 | FromGitter | <tim-st> additionally to incl |
| 00:19:36 | FromGitter | <ervinbosenbacher> okay i successfully built nim on my ARM based ubuntu :) |
| 00:20:00 | FromGitter | <ervinbosenbacher> its a good start |
| 00:20:31 | * | user0 quit (Client Quit) |
| 00:20:46 | * | user0 joined #nim |
| 00:21:01 | * | user0 quit (Client Quit) |
| 00:21:50 | FromGitter | <tim-st> I remember I got sort working but now I cannot for seq[string] :( |
| 00:22:06 | FromGitter | <tim-st> in python the default call is easier |
| 00:23:18 | FromGitter | <tim-st> oh, it was in algorithm modul, now it works |
| 00:24:06 | * | byteflame joined #nim |
| 00:24:06 | * | dyce[m] joined #nim |
| 00:24:07 | * | Miguelngel[m] joined #nim |
| 00:24:07 | * | Demos[m] joined #nim |
| 00:24:07 | * | ehmry joined #nim |
| 00:24:07 | * | unclechu joined #nim |
| 00:24:09 | * | sz0 quit (Quit: Connection closed for inactivity) |
| 00:24:13 | * | TheManiac joined #nim |
| 00:24:13 | * | Jipok[m] joined #nim |
| 00:24:13 | * | macsek1911[m] joined #nim |
| 00:24:14 | * | hohlerde joined #nim |
| 00:24:14 | * | notdekka[m] joined #nim |
| 00:24:15 | * | planetis[m] joined #nim |
| 00:27:45 | FromGitter | <data-man> @shashlick: For Windows I recommend to use Everything http://www.voidtools.com ⏎ Very fast. There is SDK http://www.voidtools.com/support/everything/sdk/ |
| 00:30:27 | * | dandevelo joined #nim |
| 00:30:59 | FromGitter | <data-man> @tim-st: Your keys is strings? |
| 00:31:42 | FromGitter | <tim-st> yes, I got it working. I didnt remember that I needed to import algorithm for using sort |
| 00:32:06 | FromGitter | <tim-st> thought it was in system, but only cmp was there |
| 00:33:38 | FromGitter | <data-man> Try critbits |
| 00:34:11 | FromGitter | <tim-st> can it be constant? |
| 00:36:04 | FromGitter | <tim-st> I need to retrieve a sequence of sorted keys, for example everything between "ab"..."abe" |
| 00:36:14 | FromGitter | <tim-st> or an iterator |
| 00:37:26 | FromGitter | <data-man> CritBitTree support search with prefix |
| 00:38:33 | FromGitter | <tim-st> hm, looks very interesting indeed, thanks for letting me know! |
| 00:38:49 | FromGitter | <tim-st> not sure if its constant though |
| 00:38:57 | FromGitter | <tim-st> *can be |
| 00:40:22 | FromGitter | <data-man> What do you mean "constant"? |
| 00:41:28 | FromGitter | <tim-st> can I make `const cb = CritBitTree()`? |
| 00:42:01 | FromGitter | <tim-st> or `const cb = procThatReturnsSuchTree` |
| 00:43:36 | FromGitter | <data-man> For that? How many keys you need to store? |
| 00:44:18 | FromGitter | <tim-st> wikipedia index |
| 00:44:33 | FromGitter | <tim-st> and wiktionary |
| 00:45:41 | FromGitter | <tim-st> less than 4.5 million |
| 00:46:30 | FromGitter | <data-man> Oh! Then you better to use a key-value DBs or SQLite |
| 00:46:48 | FromGitter | <tim-st> why? it's read only |
| 00:47:38 | FromGitter | <tim-st> I'm pretty sure nims datastructures are much better for this by far |
| 00:51:27 | * | yglukhov quit (Remote host closed the connection) |
| 00:52:37 | FromGitter | <ervinbosenbacher> hey @Tim what are you building out of interest |
| 00:54:19 | * | endragor joined #nim |
| 00:55:05 | FromGitter | <tim-st> a system for natural processing, but with the aim to interpret very good or even correct, and after this I use it as a base for solving many interesting problems :) |
| 00:55:14 | FromGitter | <tim-st> *natural language |
| 00:55:30 | FromGitter | <ervinbosenbacher> yeah we are working on similar things |
| 00:55:38 | FromGitter | <ervinbosenbacher> :) go figure |
| 00:57:04 | FromGitter | <ervinbosenbacher> i am wokring on an finite state machine for fuzzy search. |
| 00:58:42 | * | endragor quit (Ping timeout: 246 seconds) |
| 00:58:46 | FromGitter | <tim-st> interesting^^, I want to support this too and also other very interesting things regarding linguistics |
| 01:08:40 | * | marenz__ quit (Ping timeout: 265 seconds) |
| 01:19:35 | * | Demos[m] quit (Ping timeout: 276 seconds) |
| 01:22:00 | * | yglukhov joined #nim |
| 01:26:44 | * | yglukhov quit (Ping timeout: 276 seconds) |
| 01:27:38 | * | Demos[m] joined #nim |
| 01:53:31 | * | xet7 quit (Quit: Leaving) |
| 01:57:56 | * | dandevelo quit (Ping timeout: 276 seconds) |
| 02:08:37 | * | S1t1Schu joined #nim |
| 02:12:28 | * | S1tiSchu quit (Ping timeout: 265 seconds) |
| 02:25:24 | * | yglukhov joined #nim |
| 02:29:42 | * | yglukhov quit (Ping timeout: 246 seconds) |
| 02:33:39 | FromGitter | <RedBeard0531> Tim, a critbit tree of that size will be pretty slow. Maybe try porting or wrapping a btree or qp tree impl |
| 02:50:36 | * | S1t1Schu quit (Remote host closed the connection) |
| 02:50:49 | shashlick | can someone close this issue? https://github.com/nim-lang/Nim/issues/1976 |
| 02:57:36 | * | yglukhov joined #nim |
| 02:58:27 | * | vlad1777d quit (Ping timeout: 240 seconds) |
| 02:59:51 | * | chemist69 quit (Ping timeout: 240 seconds) |
| 03:01:56 | * | yglukhov quit (Ping timeout: 255 seconds) |
| 03:02:41 | * | hosearch joined #nim |
| 03:02:43 | * | hosearch quit (Remote host closed the connection) |
| 03:14:00 | * | chemist69 joined #nim |
| 03:16:09 | * | arnetheduck joined #nim |
| 03:24:15 | * | dandevelo joined #nim |
| 03:29:48 | * | dandevelo quit (Ping timeout: 265 seconds) |
| 03:31:20 | * | MJCaley quit (Quit: MJCaley) |
| 04:01:08 | * | yglukhov joined #nim |
| 04:05:21 | * | yglukhov quit (Ping timeout: 248 seconds) |
| 04:21:48 | * | arnetheduck_ joined #nim |
| 05:12:56 | * | yglukhov joined #nim |
| 05:17:28 | * | yglukhov quit (Ping timeout: 268 seconds) |
| 05:26:44 | * | dandevelo joined #nim |
| 05:32:05 | * | dandevelo quit (Ping timeout: 265 seconds) |
| 06:13:28 | * | devdri joined #nim |
| 06:33:54 | * | nsf joined #nim |
| 06:38:55 | * | devdri quit () |
| 07:07:17 | * | yglukhov joined #nim |
| 07:11:40 | * | yglukhov quit (Ping timeout: 260 seconds) |
| 07:13:16 | * | TheManiac quit (*.net *.split) |
| 07:13:16 | * | Jipok[m] quit (*.net *.split) |
| 07:13:16 | * | Syneh_ quit (*.net *.split) |
| 07:13:16 | * | enigmeta quit (*.net *.split) |
| 07:13:17 | * | EastByte quit (*.net *.split) |
| 07:13:21 | * | Syneh__ joined #nim |
| 07:13:49 | * | Jipok[m] joined #nim |
| 07:13:49 | * | TheManiac joined #nim |
| 07:13:50 | * | EastByte joined #nim |
| 07:13:52 | * | enigmeta joined #nim |
| 07:16:20 | * | heinrich5991 quit (Ping timeout: 260 seconds) |
| 07:21:39 | * | wurui joined #nim |
| 07:26:07 | * | heinrich5991 joined #nim |
| 07:29:11 | * | dandevelo joined #nim |
| 07:47:45 | * | dandevelo quit (Ping timeout: 264 seconds) |
| 08:02:37 | * | wurui_ joined #nim |
| 08:03:37 | * | wurui_ quit (Max SendQ exceeded) |
| 08:03:51 | * | wurui quit (Ping timeout: 240 seconds) |
| 08:04:07 | * | wurui_ joined #nim |
| 08:05:15 | * | wurui_ quit (Remote host closed the connection) |
| 08:05:52 | * | wurui_ joined #nim |
| 08:06:45 | * | wurui_ quit (Remote host closed the connection) |
| 08:12:39 | * | floppydh joined #nim |
| 08:17:04 | * | dddddd joined #nim |
| 08:30:07 | * | PMunch joined #nim |
| 08:34:53 | FromGitter | <ervinbosenbacher> can we use pointers in Nim aka ptr? |
| 08:35:01 | * | claudiuinberlin joined #nim |
| 08:42:42 | * | sendell joined #nim |
| 08:43:37 | * | wurui joined #nim |
| 08:47:33 | PMunch | ervinbosenbacher, yes? |
| 08:48:04 | FromGitter | <ervinbosenbacher> trying to interface withsome level c libs and very new to nim |
| 08:48:12 | PMunch | Ah right |
| 08:48:22 | PMunch | Yeah you're able to create a ptr type |
| 08:48:25 | FromGitter | <ervinbosenbacher> some very low level |
| 08:48:44 | FromGitter | <ervinbosenbacher> simd intrinsics |
| 08:48:47 | PMunch | Have you checked that it hasn't been wrapped already by the way? |
| 08:49:17 | FromGitter | <ervinbosenbacher> :) |
| 08:49:20 | FromGitter | <ervinbosenbacher> good point |
| 08:49:56 | livcd | i remember something about simd intrisincs being mentioned on forums / reddit |
| 08:49:58 | PMunch | This is a pretty interesting project for SIMD: https://github.com/jackmott/nim_simd |
| 08:50:31 | FromGitter | <ervinbosenbacher> yes, found it, that saves me some time, working on a bitpacking / compression lib |
| 08:50:41 | FromGitter | <ervinbosenbacher> and helps to learn the language |
| 08:50:55 | FromGitter | <ervinbosenbacher> OpenMP |
| 08:51:12 | FromGitter | <ervinbosenbacher> mentioned here |
| 08:51:13 | FromGitter | <ervinbosenbacher> https://forum.nim-lang.org/t/212 |
| 08:52:05 | FromGitter | <ervinbosenbacher> thx for the help that heps to start to do what I do |
| 08:53:04 | PMunch | Happy to help, hope you find Nim to your liking :) |
| 08:53:11 | FromGitter | <ervinbosenbacher> @PMunch thank you |
| 08:53:19 | FromGitter | <ervinbosenbacher> yes I am in love |
| 08:53:21 | FromGitter | <ervinbosenbacher> :D |
| 08:53:38 | FromGitter | <ervinbosenbacher> I will help with the compiler and tools once I know better |
| 08:58:48 | * | tajmone joined #nim |
| 08:59:12 | * | tajmone quit (Client Quit) |
| 08:59:45 | * | yglukhov joined #nim |
| 09:00:08 | * | xet7 joined #nim |
| 09:00:26 | FromGitter | <ervinbosenbacher> previously I was working with D and rust |
| 09:01:18 | PMunch | And how do you feel Nim compares? |
| 09:01:36 | FromGitter | <ervinbosenbacher> Freedom |
| 09:01:54 | FromGitter | <ervinbosenbacher> well defined small language |
| 09:02:19 | FromGitter | <ervinbosenbacher> but what is critical in my case is that I can target various languages |
| 09:02:40 | PMunch | Mhm, that is pretty neat |
| 09:03:08 | FromGitter | <ervinbosenbacher> we have spent gazillion time to make sure that libs works across frameworks languages that we bridge things here and there |
| 09:03:48 | FromGitter | <ervinbosenbacher> some of the stuff implemented in rust to be able to close to metal some in c++11/14/17 some in D to interface easily with C++ some in JS for the front and so on |
| 09:04:03 | FromGitter | <ervinbosenbacher> With nim well 1 language |
| 09:04:16 | * | yglukhov quit (Ping timeout: 265 seconds) |
| 09:04:16 | FromGitter | <ervinbosenbacher> so I am busy porting our core libs to Nim right now |
| 09:04:24 | * | gmpreussner quit (Ping timeout: 268 seconds) |
| 09:04:43 | * | gmpreussner joined #nim |
| 09:05:30 | PMunch | Oh nice! Big project? |
| 09:05:39 | FromGitter | <ervinbosenbacher> Nim is pure awesomeness especially that I am into macros, templates metaprogramming |
| 09:06:07 | PMunch | Yeah, the metaprogramming stuff is great. So many cool projects that wouldn't be possible without it |
| 09:06:15 | FromGitter | <ervinbosenbacher> quite big, working on a search engine that you can deploy anywhere aka embedded devices |
| 09:06:36 | FromGitter | <ervinbosenbacher> I will add stuff to here as open source |
| 09:06:54 | FromGitter | <ervinbosenbacher> working on an DSL for queries |
| 09:07:06 | FromGitter | <ervinbosenbacher> imagine elasticsearch and lucene in Nim |
| 09:07:41 | FromGitter | <ervinbosenbacher> i added our startup to the list of companies |
| 09:07:43 | PMunch | Ooooh, very interesting |
| 09:07:45 | FromGitter | <ervinbosenbacher> finddit.com |
| 09:08:20 | FromGitter | <ervinbosenbacher> Nim is just perfect for our use case |
| 09:08:34 | PMunch | Yeah I'd imagine |
| 09:08:41 | FromGitter | <ervinbosenbacher> before I created a metaprogramming lib in c++14/17 |
| 09:09:20 | FromGitter | <ervinbosenbacher> trageting C with Pythonish syntax. almost allof our AI research is in pythob |
| 09:09:21 | * | yglukhov joined #nim |
| 09:09:30 | FromGitter | <ervinbosenbacher> so perfect |
| 09:11:01 | FromGitter | <ervinbosenbacher> anyway I will contribute as much as I can but its been only 3 days so patience :D |
| 09:11:08 | * | yglukhov quit (Read error: Connection reset by peer) |
| 09:11:41 | * | yglukhov joined #nim |
| 09:11:55 | FromGitter | <ervinbosenbacher> my interest is that Nim becomes powerful, its already is, by using Nim we are savin a lot of time as I had a similar project in my mind. |
| 09:12:01 | FromGitter | <ervinbosenbacher> so we just continue with Nim |
| 09:12:37 | PMunch | That's great to hear, looking forward to seeing you use Nim in the future :) |
| 09:13:10 | FromGitter | <ervinbosenbacher> I have extensive compiler experience so I will help with Nim v2 |
| 09:13:33 | FromGitter | <ervinbosenbacher> last nighti already packed up Nim on an ARM device |
| 09:13:50 | FromGitter | <ervinbosenbacher> compiled without effort |
| 09:13:57 | FromGitter | <ervinbosenbacher> high quality stuff |
| 09:14:52 | PMunch | I have yet to try it on ARM or even Arduinos, but it's something I've been meaning to do for the longest time.. |
| 09:15:48 | FromGitter | <ervinbosenbacher> We are targeting jetson tx2 with cuda, after checkng out the code on it took a few minutes to build it |
| 09:15:58 | FromGitter | <ervinbosenbacher> flawless |
| 09:16:24 | FromGitter | <ervinbosenbacher> (its also faster than rust with a smaller footprint for some reason) |
| 09:16:36 | FromGitter | <ervinbosenbacher> (in my case) |
| 09:20:00 | * | Vladar joined #nim |
| 09:23:05 | * | xet7 quit (Ping timeout: 240 seconds) |
| 09:35:57 | * | xet7 joined #nim |
| 09:44:32 | * | dandevelo joined #nim |
| 09:45:33 | * | yglukhov quit (Read error: Connection reset by peer) |
| 09:45:34 | * | wurui left #nim (#nim) |
| 09:46:09 | * | yglukhov joined #nim |
| 09:47:08 | * | yglukhov quit (Remote host closed the connection) |
| 09:49:10 | * | dandevelo quit (Ping timeout: 260 seconds) |
| 09:59:24 | * | yglukhov joined #nim |
| 10:00:32 | FromGitter | <ervinbosenbacher> what is the ! operator do? its been depreated it seems |
| 10:00:40 | * | yglukhov quit (Read error: Connection reset by peer) |
| 10:01:12 | * | yglukhov joined #nim |
| 10:04:48 | FromGitter | <alehander42> @ervinbosenbacher I am using karax in a middle-sized electron app, it works fine for me |
| 10:05:00 | FromGitter | <ervinbosenbacher> perfect |
| 10:05:21 | FromGitter | <alehander42> @araq is there a way to getCurrentDir() at compile time ? |
| 10:05:21 | FromGitter | <ervinbosenbacher> Is it open source can you share? Its ok if not |
| 10:05:26 | FromGitter | <alehander42> no, sorry |
| 10:05:37 | FromGitter | <ervinbosenbacher> its ok I was just wondering |
| 10:05:39 | FromGitter | <alehander42> but if you have any questions, feel free to ask |
| 10:05:44 | FromGitter | <ervinbosenbacher> thx |
| 10:06:03 | FromGitter | <ervinbosenbacher> ill try to figure it out |
| 10:06:27 | FromGitter | <ervinbosenbacher> can you see preformance improvement? |
| 10:06:36 | FromGitter | <ervinbosenbacher> are you more productive? |
| 10:06:46 | FromGitter | <ervinbosenbacher> What about the size of the running app? |
| 10:07:04 | FromGitter | <alehander42> well, we use Nim for the electron app since we created it, so I can't really compare haha |
| 10:08:27 | Araq | I am more productive with karax than I was with vue.js but |
| 10:08:34 | Araq | a) I loathe vue.js |
| 10:08:40 | Araq | b) I wrote Karax. |
| 10:08:47 | Araq | so I am as biased as possible. |
| 10:08:55 | FromGitter | <ervinbosenbacher> i know that I am just wndering about the process |
| 10:09:17 | FromGitter | <ervinbosenbacher> it would be nice to have at least a boilerplate to see how that works |
| 10:09:44 | FromGitter | <ervinbosenbacher> or avoid reinventing the wheel |
| 10:09:46 | Araq | https://nim-lang.org/araq/karax.html ? |
| 10:10:02 | FromGitter | <ervinbosenbacher> i mean integration with electron |
| 10:10:19 | FromGitter | <ervinbosenbacher> @araq i have a quikc question |
| 10:10:27 | Araq | oh ok, I don't use electron. shoot. |
| 10:10:28 | FromGitter | <alehander42> @Araq is staticRead of another nim binary that echo-es getCurrentDir the only way to get it at compile time :? |
| 10:10:56 | FromGitter | <ervinbosenbacher> is the ! operator used for negating things? |
| 10:10:58 | FromGitter | <ervinbosenbacher> if node.ident == !"simd": |
| 10:10:59 | Araq | alehander42: I don't know, getConfigDir works at compile-time and it did bite us |
| 10:11:17 | Araq | ervinbosenbacher, nah it's just identifier construction |
| 10:11:25 | FromGitter | <ervinbosenbacher> this line is deprecated in nim_simd I am wondering why |
| 10:11:38 | Araq | read it as if node.ident == "simd" |
| 10:11:53 | FromGitter | <ervinbosenbacher> ah |
| 10:12:23 | FromGitter | <ervinbosenbacher> so if node.ident == !"simd”: is the same as if node.ident == "simd”:? |
| 10:13:56 | Araq | well no, it's the same as if node.ident == ident"simd" but the 2nd ident only exists to make the type checker happy |
| 10:15:27 | FromGitter | <ervinbosenbacher> thx |
| 10:15:42 | FromGitter | <ervinbosenbacher> how about thsi one |
| 10:15:43 | FromGitter | <ervinbosenbacher> in countup(0,<a.len,simd.width div 4): |
| 10:15:48 | * | vlad1777d joined #nim |
| 10:15:50 | FromGitter | <ervinbosenbacher> The < is deprecated |
| 10:16:07 | FromGitter | <ervinbosenbacher> what would be the current equivalent |
| 10:16:27 | FromGitter | <ervinbosenbacher> I am guessing until ? |
| 10:16:54 | FromGitter | <alehander42> how did getConfigDir bite u |
| 10:19:38 | Araq | a.len-1 |
| 10:19:52 | FromGitter | <ervinbosenbacher> yes i already did it :) |
| 10:20:16 | Araq | alehander42: use it at compile-time, it resolves to /home/alehander42/.config, tada |
| 10:20:23 | Araq | a program that only works on your machine |
| 10:22:16 | FromGitter | <alehander42> aha, well I want to use it for a test macro (it reads filenames of examples on compile time), as usually I build it as a part of the test |
| 10:22:20 | FromGitter | <alehander42> but I can see your point |
| 10:23:38 | * | chemist69 quit (Ping timeout: 276 seconds) |
| 10:25:38 | * | xet7 quit (Ping timeout: 255 seconds) |
| 10:28:40 | FromGitter | <ervinbosenbacher> this is what worked |
| 10:28:42 | FromGitter | <ervinbosenbacher> if node.ident == toNimIdent("simd"): |
| 10:29:52 | Araq | that's the long version, yes. |
| 10:30:14 | FromGitter | <ervinbosenbacher> teh short version is ident”simd”? |
| 10:30:18 | Araq | but my version should also work |
| 10:30:28 | Araq | yes. |
| 10:31:12 | FromGitter | <ervinbosenbacher> whats the correct syntax? |
| 10:31:16 | FromGitter | <ervinbosenbacher> ``` node.ident = ident simdType`````` |
| 10:31:58 | FromGitter | <ervinbosenbacher> any shorter version? |
| 10:33:44 | FromGitter | <ervinbosenbacher> I am using 0.17.3 |
| 10:35:21 | * | xet7 joined #nim |
| 10:36:33 | * | xet7 quit (Remote host closed the connection) |
| 10:36:50 | * | xet7 joined #nim |
| 10:37:02 | * | chemist69 joined #nim |
| 10:42:32 | Araq | what are you trying to do? assignment or comparison? |
| 10:43:10 | FromGitter | <ervinbosenbacher> well thsi works ``` if node.ident == toNimIdent("simd"): ⏎ ⏎ ``` node.ident = toNimIdent(simdType)`````` [https://gitter.im/nim-lang/Nim?at=5a534b3e19147ac3231b536f] |
| 10:43:24 | FromGitter | <ervinbosenbacher> how does the short version looks like? |
| 10:49:05 | Araq | node = newIdentNode simdType iirc |
| 10:53:39 | * | floppydh quit (Quit: WeeChat 2.0.1) |
| 10:55:47 | * | floppydh joined #nim |
| 10:56:50 | FromGitter | <ervinbosenbacher> @Araq would this be correct code? If I want to introduce a new type |
| 10:56:55 | FromGitter | <ervinbosenbacher> `````` |
| 10:57:15 | * | xet7 quit (Ping timeout: 268 seconds) |
| 10:57:34 | FromGitter | <ervinbosenbacher> the original ```typedef float __m128 *attribute*((__vector_size__(16)));``` |
| 10:57:55 | FromGitter | <Araq> yeah, looks good |
| 10:58:11 | FromGitter | <ervinbosenbacher> thx |
| 11:07:23 | Araq | bah now I am curious. **why** is fmt causing a conflict? what's the error message? |
| 11:08:35 | * | yglukhov quit (Remote host closed the connection) |
| 11:09:30 | FromGitter | <tim-st> How do I call a proc that takes varargs[(string, string)] -> strutils.multiReplace ? |
| 11:10:11 | * | xet7 joined #nim |
| 11:11:23 | FromGitter | <tim-st> I dont understand it when reading: https://nim-lang.org/docs/tut1.html#advanced-types-varargs |
| 11:12:40 | PMunch | tim-st, what is your problem? |
| 11:13:29 | PMunch | multiReplace takes a string and then any number of tuples of two strings |
| 11:13:32 | FromGitter | <tim-st> how do I call the multiReplace proc? "hello".multireplace(("e", "x"), "h", "J") ? |
| 11:13:41 | FromGitter | <alehander42> does nim have some slice syntax similar to python's :: ? |
| 11:14:12 | Araq | there is no step parameter, alehander42 |
| 11:14:21 | PMunch | So you can call it like "hello".multireplace(("e", "x"), ("h","J")) and it should give you "Jxllo" |
| 11:14:35 | FromGitter | <tim-st> this dont work: `return w.multiReplace(("ä":"a"), ("ö":"o"))` |
| 11:14:40 | * | vlad1777d quit (Ping timeout: 248 seconds) |
| 11:14:47 | PMunch | Not : |
| 11:14:48 | PMunch | , |
| 11:14:56 | FromGitter | <tim-st> ah, sry my mistake |
| 11:15:26 | Araq | w.multiReplace({"ä": "a", "ö": "o"}) is also possible, I think |
| 11:16:18 | FromGitter | <tim-st> it works good, my problem was again that I didnt import it because I manually imported from strutils and forget this... |
| 11:16:50 | * | yglukhov joined #nim |
| 11:17:32 | livcd | can I use nim on Solaris/sparc or this is not advised ? |
| 11:17:51 | FromGitter | <alehander42> should we add some kind of `step` iterator to sequtils? |
| 11:18:15 | * | dandevelo joined #nim |
| 11:18:17 | * | wurui joined #nim |
| 11:18:19 | FromGitter | <ervinbosenbacher> @alehander42 Maybe you could try? I think it would be nice! |
| 11:21:14 | FromGitter | <ervinbosenbacher> is there a serialization framework in Nim? |
| 11:21:51 | * | wurui quit (Remote host closed the connection) |
| 11:23:16 | Araq | livcd: should work on Solaris |
| 11:24:44 | Araq | ervinbosenbacher: there is marshal.nim |
| 11:32:58 | FromGitter | <ervinbosenbacher> thx @Araq |
| 11:33:50 | FromGitter | <tim-st> It would be really nice if two runes() iterators could run parallel using zip() it would end when the shorter ends and if one could find out which iterator was the one that ended. That could improve my performance very much |
| 11:37:16 | * | MJCaley joined #nim |
| 11:39:55 | Araq | yeah well, write a zip that knows about runes, not ideal but also not much work |
| 11:40:08 | FromGitter | <ervinbosenbacher> especially if they could run on different threads |
| 11:40:45 | * | wurui joined #nim |
| 11:41:14 | FromGitter | <ervinbosenbacher> i have a similar problem the indexer is indexing different parts of the same docuemnt or different documents at the same time then I have to merge the results in a table |
| 11:41:20 | FromGitter | <alehander42> why is `()` deprecated |
| 11:41:51 | Araq | alehander42: Unknown interactions with the rest of the language |
| 11:42:52 | Araq | unclear benefits, you can overload `{}` instead. |
| 11:43:12 | FromGitter | <alehander42> ok, so when will it be removed completely ? |
| 11:43:56 | Araq | when we know. |
| 11:44:24 | PMunch | ervinbosenbacher, I also wrote this: https://github.com/PMunch/binaryparse |
| 11:44:30 | PMunch | To read and write binary formats |
| 11:44:35 | Araq | it's only deprecated to force an .experimental pragma |
| 11:44:46 | FromGitter | <ervinbosenbacher> ah thank you |
| 11:46:22 | FromGitter | <ervinbosenbacher> this is 2 years old https://github.com/dfdeshom/nimlz4 Can i pick this up and update it? |
| 11:47:01 | PMunch | Don't see why not :) |
| 11:47:18 | FromGitter | <ervinbosenbacher> ok I will continue to maintain it. |
| 11:49:28 | FromGitter | <ervinbosenbacher> @Tim I will experiment with multiple strems of compression using lz4, say you want to use multiple iterators in parallel through channels |
| 11:49:54 | FromGitter | <ervinbosenbacher> so will compress/decompress multiple streams at the same time |
| 11:51:55 | FromGitter | <tim-st> For compression it's good to use them, for me the reason was primarily to compare two string rune by rune and return as early as possible instead of operating two times on both complete string and than iterate over them |
| 11:52:15 | * | MJCaley quit (Quit: MJCaley) |
| 11:52:46 | FromGitter | <ervinbosenbacher> yes I get it |
| 11:55:10 | Araq | oh it's one of those days ... why does nimble dir use a dot? do not hide source code... |
| 11:55:51 | FromGitter | <tim-st> @Araq can I load data from filestream and make the returned result const? |
| 11:56:21 | FromGitter | <tim-st> I get error undeclared identifier |
| 11:57:21 | Araq | I forgot to tell you. embedding full wikipedia into your executable is not a good idea. |
| 11:57:34 | FromGitter | <tim-st> its only 45mb |
| 11:57:43 | FromGitter | <tim-st> or less |
| 11:58:36 | Araq | still meh, use getAppDir() / "resource.xml", your compiletimes will thank you |
| 11:59:06 | Araq | and to answer your question: I don't think you can. |
| 11:59:16 | FromGitter | <tim-st> ok, thanks |
| 12:03:26 | * | Snircle joined #nim |
| 12:03:43 | FromGitter | <ervinbosenbacher> that or a key value store |
| 12:03:47 | FromGitter | <ervinbosenbacher> or json |
| 12:04:22 | FromGitter | <tim-st> does a key value store communicate via tcp? |
| 12:04:35 | FromGitter | <tim-st> *every |
| 12:05:13 | FromGitter | <ervinbosenbacher> What is your preferece @tim-st |
| 12:05:16 | * | wurui_ joined #nim |
| 12:06:03 | FromGitter | <tim-st> I would a communication via ram not tcp |
| 12:06:09 | FromGitter | <tim-st> *prefer |
| 12:06:28 | FromGitter | <ervinbosenbacher> yeah I am not ready but eventually rocksdb will happen |
| 12:06:45 | * | wurui quit (Remote host closed the connection) |
| 12:06:55 | FromGitter | <ervinbosenbacher> onmy todo list which is embedded and very powerful and key value docuemnt store |
| 12:07:32 | FromGitter | <ervinbosenbacher> https://github.com/zielmicha/leveldb.nim |
| 12:07:35 | FromGitter | <ervinbosenbacher> try this |
| 12:07:41 | FromGitter | <ervinbosenbacher> rocksdb is based on leveldb |
| 12:08:15 | FromGitter | <ervinbosenbacher> someone had the same idea :D https://github.com/nimscale/rocksdb |
| 12:08:31 | FromGitter | <tim-st> Thanks, I will read into this database |
| 12:09:51 | FromGitter | <ervinbosenbacher> well here we go https://github.com/nim-lang/needed-libraries/issues/75 |
| 12:10:22 | * | wurui_ quit (Remote host closed the connection) |
| 12:10:33 | FromGitter | <ervinbosenbacher> I will work on this definitely as I am dare need as well and I have been toying with this for more than a year I have a POC in python was plannning to develop it in rust |
| 12:10:38 | FromGitter | <ervinbosenbacher> it will be nim |
| 12:13:07 | * | wurui joined #nim |
| 12:14:57 | * | zahary joined #nim |
| 12:23:03 | * | wurui quit (Remote host closed the connection) |
| 12:23:38 | FromGitter | <ervinbosenbacher> ill write a wrapper first and will try to figure how to make it idiomatic with Nim |
| 12:23:55 | FromGitter | <ervinbosenbacher> ro integrate it with the Nim orm lib |
| 12:25:06 | Araq | what is the Nim orm lib? |
| 12:25:22 | enthus1ast | ervinbosenbacher the last time i used lz4 module it leaked memory (so say a comment in old code) |
| 12:25:22 | FromGitter | <ervinbosenbacher> Whats the name you wrote it :D |
| 12:25:45 | FromGitter | <ervinbosenbacher> thx for the info |
| 12:27:21 | enthus1ast | ormin |
| 12:27:35 | FromGitter | <ervinbosenbacher> Thats the one thx :D |
| 12:28:07 | enthus1ast | Does ormin has a network api? Looks like it |
| 12:28:29 | FromGitter | <ervinbosenbacher> dont know, right now I have information overload |
| 12:28:39 | * | dandevelo quit (Ping timeout: 256 seconds) |
| 12:28:49 | FromGitter | <ervinbosenbacher> :D new language and trying to find my way through a bunch of stuff |
| 12:29:19 | enthus1ast | ormin + karax could be a nimmish "meteor" |
| 12:29:33 | FromGitter | <ervinbosenbacher> yes |
| 12:29:54 | FromGitter | <ervinbosenbacher> i have been nimbed |
| 12:30:10 | FromGitter | <ervinbosenbacher> my brain is nimb |
| 12:30:14 | FromGitter | <ervinbosenbacher> ok enuff |
| 12:32:18 | Araq | ormin is SQL only and hardly works with today's compiler :-) |
| 12:33:13 | FromGitter | <ervinbosenbacher> ok we will have something for key value stuff |
| 12:33:37 | FromGitter | <ervinbosenbacher> one of my core components is built around compression key value stores and encryption |
| 12:34:03 | FromGitter | <ervinbosenbacher> I need those in and possibly I will start with compression and key value store |
| 12:34:26 | FromGitter | <ervinbosenbacher> make them open source libs for the benfit of everyone |
| 12:34:55 | enthus1ast | ervinbosenbacher tried considering redis? |
| 12:35:13 | FromGitter | <ervinbosenbacher> yes but there are issues with redis |
| 12:36:02 | FromGitter | <ervinbosenbacher> One of them it is not really free when you want to embed it and that it occupies huge memory space |
| 12:36:54 | FromGitter | <ervinbosenbacher> this is what I have been using https://github.com/yinqiwen/ardb |
| 12:37:16 | livcd | Araq: thanks |
| 12:37:31 | dom96 | Redis isn't really free? how so? |
| 12:40:11 | FromGitter | <ervinbosenbacher> Well we have run into issues I dont know. But for our purpose MongoDB is better suited |
| 12:40:20 | FromGitter | <ervinbosenbacher> Sorry mixed up with mongo |
| 12:41:07 | FromGitter | <ervinbosenbacher> anyway mongodb woud be better for a search engine. Also redis is a huge memory footprint and rocksdb doesnt |
| 12:42:01 | FromGitter | <ervinbosenbacher> This was a long debate. Redis could be a could choice. |
| 12:42:08 | FromGitter | <ervinbosenbacher> It dpeends on the scenario |
| 12:42:41 | enthus1ast | Idk, but for a search engine i would look for these big table like dbs |
| 12:42:49 | FromGitter | <ervinbosenbacher> @Tim there is a wrapper Redis |
| 12:43:30 | FromGitter | <ervinbosenbacher> https://github.com/nim-lang/redis |
| 12:45:07 | FromGitter | <ervinbosenbacher> yes scylladb |
| 12:45:27 | FromGitter | <ervinbosenbacher> thats in c++ so we could have a Nim wrapper |
| 12:48:52 | enthus1ast | Araq do you know meteor js? Because ormin+karax could potentially be a nimish meteor. (withouth the annoying parts) |
| 12:49:00 | * | Yardanico joined #nim |
| 12:49:11 | Araq | never heard of it |
| 12:49:29 | FromGitter | <ervinbosenbacher> https://www.meteor.com/ |
| 12:50:29 | enthus1ast | meteor = write server/client js code side by side, client can register a view on an imaginary db table, client gets autoupdated, client has a simple `mongodb` like in the browser |
| 12:50:35 | Araq | oh wow, this needs some exposure, https://forum.nim-lang.org/t/3459 |
| 12:51:04 | FromGitter | <ervinbosenbacher> you mean like a universal app? |
| 12:51:43 | enthus1ast | database in the browser gets updated via websocket when data in the database change |
| 12:52:07 | Araq | wow I tried to build such a thing once and failed :-) |
| 12:52:58 | Araq | meteor seems the right idea for the wrong language |
| 12:53:17 | enthus1ast | the only thing that seems missing (karax+ormin) is a way to update the clients |
| 12:53:34 | Araq | no, that is covered |
| 12:53:47 | FromGitter | <ervinbosenbacher> then its done |
| 12:53:54 | Araq | the only thing is that that freaking websockets module doesn't work for me |
| 12:53:55 | FromGitter | <ervinbosenbacher> we just need that 2nd blog :) |
| 12:54:13 | Araq | that 2nd blog is done since forever |
| 12:54:15 | livcd | I want to run c2nim on some code pieces from reactos. Do I just the .h and .c files for those ? |
| 12:54:27 | FromGitter | <ervinbosenbacher> :) |
| 12:54:27 | Araq | livcd: https://forum.nim-lang.org/t/3459 |
| 12:54:41 | livcd | ah |
| 12:55:59 | FromGitter | <Yardanico> livcd: wow, reactos? :) |
| 12:56:10 | FromGitter | <ervinbosenbacher> you mean how karax + ormin works ? |
| 12:56:24 | Yardanico | livcd, I have some connects with guys from reactos :D |
| 12:56:28 | Yardanico | *connections |
| 12:56:33 | FromGitter | <ervinbosenbacher> @araq you said its not ready yet. |
| 12:56:49 | livcd | Yardanico: well reactos already has those wrappers defined so I hope i can use them |
| 12:57:19 | Araq | ervinbosenbacher: no I said, it hardly works with today's compiler. |
| 12:57:37 | Araq | it does work and will work better with a couple of compiler fixes. |
| 12:57:38 | FromGitter | <ervinbosenbacher> ok ok |
| 12:57:46 | Yardanico | livcd, I mean there's telegram groups (russian is the biggest one) and IRC channel #reactos #reactos-dev #reactos-testers |
| 12:59:25 | livcd | Yardanico: well i am only interested in simple stuff :D |
| 12:59:45 | livcd | like enumerating services,manipulating services etc |
| 13:00:26 | Yardanico | livcd, reactos should be generally compatible with winapi |
| 13:01:17 | Yardanico | ReactOS currently targets NT5.2 |
| 13:01:17 | livcd | so eg. I want to take this out and try nimgen/c2nim on it https://doxygen.reactos.org/d2/d57/cmdStart_8c_source.html |
| 13:01:47 | Yardanico | well it would be not that easy because winapi is always a lot of macros :) |
| 13:01:52 | Yardanico | but IDK about nimgen |
| 13:03:50 | livcd | uhm bummer i cant install nimgen |
| 13:04:12 | FromGitter | <ervinbosenbacher> @Araq https://nim-lang.org/araq/karax.html is there a follow up this post? |
| 13:04:30 | FromGitter | <ervinbosenbacher> i know it doesnt work with the current compiler |
| 13:06:04 | Yardanico | first part does AFAIK |
| 13:06:29 | FromGitter | <ervinbosenbacher> yes that i know, i have tested it |
| 13:09:37 | FromGitter | <ervinbosenbacher> I am after Ormin or whatever way I connect to any database backend |
| 13:15:49 | dom96 | you might prefer the db_* modules in the stdlib |
| 13:18:03 | Yardanico | but they don't have actual ORM :) |
| 13:20:02 | livcd | ah this wont do. I really need nimgen i guess but i cant install it |
| 13:21:30 | * | dandevelo joined #nim |
| 13:24:10 | dom96 | livcd: We need more info to help you |
| 13:25:58 | enthus1ast | livcd: such as the error message you get :) |
| 13:26:30 | livcd | sure but i think this is on whoever created nimgen. https://ghostbin.com/paste/9u7hc |
| 13:27:25 | dom96 | nope |
| 13:27:30 | dom96 | this is c2nim failing to build |
| 13:27:47 | dom96 | which sucks |
| 13:28:18 | livcd | aahh |
| 13:28:22 | dom96 | try `nimble install compiler@#devel` (make sure your shell doesn't comment that out, but you're on Windows so it shouldn't) |
| 13:28:32 | livcd | yeah ok |
| 13:37:55 | livcd | same error |
| 13:39:30 | * | Zevv joined #nim |
| 13:40:00 | Zevv | Hi #nim; is there some documentation available explaining how buffering is done in the net module, and especially, how to disable this? |
| 13:40:07 | FromGitter | <ervinbosenbacher> hey you need head of nim and compiler |
| 13:40:09 | FromGitter | <ervinbosenbacher> remove ~/.nimble |
| 13:40:34 | FromGitter | <ervinbosenbacher> i have spent a bit of time to getit right myself as well |
| 13:41:29 | Zevv | or should I switch to rawsockets instead? |
| 13:41:59 | Araq | why disable the buffering? |
| 13:42:01 | * | martinium_laptop joined #nim |
| 13:44:07 | Zevv | I want to receive UDP packets as they come |
| 13:44:32 | Zevv | variable size |
| 13:47:15 | enthus1ast | are you not receiving the whole packet with UDP? |
| 13:48:10 | Zevv | They are buffered until buffersize is reached |
| 13:48:22 | Zevv | thus, I loos packet boundaries and get unwanted latency |
| 13:49:57 | Araq | use "nativesockets" for UDP |
| 13:50:01 | livcd | ervin: i tried that. Still the same error i pasted earlier |
| 13:50:30 | Zevv | Araq: ok, thanks |
| 13:50:58 | FromGitter | <ervinbosenbacher> @lbvcd 1 sec |
| 13:51:39 | FromGitter | <ervinbosenbacher> can you repaste the error please |
| 13:51:45 | Yardanico | guys, wow |
| 13:51:46 | Yardanico | https://github.com/metacraft-labs/py2nim |
| 13:54:02 | FromGitter | <ervinbosenbacher> thats very cool |
| 13:55:31 | PMunch | Wow, that is pretty cool |
| 13:56:31 | Yardanico | it's developed by alehander42 and zah at the same time, so I think it will be great :) |
| 13:57:43 | Yardanico | Time to update my PoC https://github.com/Yardanico/nimpylib to add link to this py2nim :) |
| 13:57:47 | livcd | ervin:https://ghostbin.com/paste/bks4a |
| 13:59:16 | Araq | ervin, data-man, https://github.com/nim-lang/Nim/issues/7041 |
| 13:59:29 | FromGitter | <ervinbosenbacher> testing py2nim |
| 13:59:43 | Yardanico | it's in very early state as stated in the readme |
| 13:59:45 | FromGitter | <ervinbosenbacher> thx @Araq |
| 13:59:53 | * | Serenitor joined #nim |
| 13:59:55 | FromGitter | <ervinbosenbacher> still I want to try it |
| 14:00:02 | FromGitter | <ervinbosenbacher> eventually contribute |
| 14:00:31 | Yardanico | https://github.com/Yardanico/nimpylib/commit/77f8086d737ffe0618064ff508ccc8fe2e9d9c00 :) |
| 14:00:48 | FromGitter | <ervinbosenbacher> well ```python3 python-deduckt/deduckt/main.py example/fib.py ⏎ ⏎ Traceback (most recent call last) ⏎ py2nim.nim(64) py2nim ⏎ py2nim.nim(60) translate ... [https://gitter.im/nim-lang/Nim?at=5a53799097cedeb0482db7c3] |
| 14:01:15 | FromGitter | <ervinbosenbacher> using python 3.6 |
| 14:04:06 | FromGitter | <ervinbosenbacher> https://github.com/metacraft-labs/py2nim/issues/1 |
| 14:04:31 | Yardanico | ervin: are you sure you did it right? |
| 14:04:39 | Yardanico | readme says you should use './py2nim example/fib.py' |
| 14:04:48 | FromGitter | <ervinbosenbacher> yes |
| 14:05:01 | FromGitter | <ervinbosenbacher> ./py2nim example/fib.py |
| 14:05:26 | FromGitter | <ervinbosenbacher> 1 sec |
| 14:06:10 | FromGitter | <ervinbosenbacher> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5a537ad151915d962afdbefe] |
| 14:07:03 | dom96 | Zevv: Araq: Don't use nativesockets if you can help it. Just disable buffering, newSocket(buffered=false) |
| 14:07:12 | Araq | ok |
| 14:09:18 | * | martinium_laptop quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 14:11:40 | * | dandevelo quit (Ping timeout: 260 seconds) |
| 14:14:48 | Zevv | dom96: I tried that, but that didn't seem to work |
| 14:14:50 | Zevv | let me check |
| 14:14:58 | FromGitter | <ervinbosenbacher> @libvcd I had the same problem. ⏎ ⏎ ```rm -rf ~/.nimble ⏎ nimble install compiler@#head ⏎ nimble install c2nim@#head``` [https://gitter.im/nim-lang/Nim?at=5a537ce281cd9a5d7ea27181] |
| 14:15:08 | FromGitter | <ervinbosenbacher> @libvcd https://github.com/nim-lang/c2nim/issues/115 |
| 14:15:14 | FromGitter | <ervinbosenbacher> Please do that |
| 14:16:28 | dom96 | Zevv: That's odd, might be a bug |
| 14:18:16 | FromGitter | <tim-st> When I have `let s = returnsSomeSeq()` and I want to pass `s` by reference to a proc how should I write it? |
| 14:18:33 | FromGitter | <tim-st> s is of seq[string] |
| 14:19:26 | FromGitter | <tim-st> I tried `ref` and `var` nothing really works |
| 14:21:39 | FromGitter | <tim-st> `return word.exists(source=var s, ignoreCase=ignoreCase)` was my guess |
| 14:22:00 | Yardanico | your proc should accept var seq[string] probably |
| 14:22:09 | Yardanico | but you need to use "var s = returnsSomeSeq()" |
| 14:23:00 | PMunch | Why isn't there a reverse of this: https://github.com/nim-lang/Nim/blob/master/lib/pure/xmltree.nim#L199 |
| 14:23:40 | FromGitter | <tim-st> @Yardanico thanks, and can I use let for this? I only want to read from the var |
| 14:23:52 | Yardanico | I don't think so, but I might be wrong |
| 14:23:59 | FromGitter | <tim-st> thanks |
| 14:24:42 | Yardanico | PMunch, lol, there are 2 other places in nim stdlib where a proc like this is used |
| 14:24:48 | Yardanico | (copy-pasted) :P |
| 14:25:01 | Yardanico | ah, sorry, 4 other places |
| 14:27:51 | PMunch | Really? |
| 14:28:01 | PMunch | And there is no place that does the opposite conversion? |
| 14:28:22 | enthus1ast | ^ this, and we also need a mime.nim because parsing mime is done in smtp; http and what not else |
| 14:28:29 | Yardanico | PMunch, I didn't find any in nim repo |
| 14:30:21 | Yardanico | PMunch, duplicates are in cgi, compiler/ccgutils, unittest, xmldom, docutils, |
| 14:31:20 | Yardanico | well they're a bit different |
| 14:40:53 | Yardanico | So as I understand it seems like with the new runtime Nim will be moving towards making C++ the default compilation target? |
| 14:41:21 | * | zahary quit (Quit: Leaving.) |
| 14:42:10 | Araq | likely, yes. |
| 14:45:30 | FromGitter | <RedBeard0531> I was reading setLengthStr and I can't tell if this is a bug or intentional. It is passing the requested size to resizeStr in the argument for how much to *grow* the string: https://github.com/nim-lang/Nim/blob/316c693b9127ecc764e3fe26528b255b08afa239/lib/system/sysstr.nim#L231 https://github.com/nim-lang/Nim/blob/316c693b9127ecc764e3fe26528b255b08afa239/lib/system/sysstr.nim#L204 |
| 14:46:16 | FromGitter | <RedBeard0531> This is certainly weird, but similar enough to doubling growth that I thought it *could* be by design |
| 14:49:10 | FromGitter | <RedBeard0531> I was thinking setLen would be easier to use setLen if it treated nil input the same as "" and left the string with the desired size, which is why I was looking in the first place. |
| 14:52:12 | FromGitter | <RedBeard0531> C++ as default CG is great news. It seems to have much lower cost exception handling (in the happy path case) than sjlj-based exception handling. |
| 14:53:23 | Yardanico | yeah, that's the reason |
| 14:53:35 | Yardanico | because newruntime will use a lot more `try` blocks |
| 14:53:48 | Yardanico | (that's one of the main reasons) |
| 14:55:03 | FromGitter | <RedBeard0531> I've noticed a tendency of core code to use try/finally rather than defer blocks, even when defer blocks would have felt more natural to me at least. Is there a reason for that, or just different people's styles? |
| 14:56:17 | FromGitter | <RedBeard0531> ex: https://github.com/nim-lang/Nim/blob/devel/lib/pure/collections/sharedtables.nim#L66-L75 |
| 14:57:02 | livcd | ervin: maybe i am doing something wrong |
| 14:57:05 | livcd | let me try that |
| 14:57:12 | euantor | `defer` converts to `try/finally` if I remember correctly, so there's no actual difference between the two except from readability |
| 14:57:30 | Araq | 'defer' came later and I don't like it |
| 14:57:39 | euantor | I find `try/finally` easier to reason about, especially if there are several lines - you can clearly see when a resource is released that way |
| 14:57:50 | Araq | try finally is easier to follow IMO |
| 14:58:27 | livcd | well i also do not like defer: when it gets converted to try/finally |
| 14:58:31 | euantor | Same way as C#'s `using (IDisposable) {}` becomes `IDisposable.create() try { } finally { IDisposable.Dispose() }` |
| 14:58:55 | livcd | sucks to have to keep in mind what it is and how it works :D |
| 14:58:55 | euantor | Though with `using()` you get the easy to follow benefit due to a clear end of scope |
| 14:59:22 | * | Yardanico[Phone] quit (Quit: Quit) |
| 14:59:47 | FromGitter | <RedBeard0531> I guess I like to see the release next to the acquire so I can stop worrying about how/whether it gets released. Also, it avoids an unnecessary layer of indentation. But I'm coming from a C++ RAII mindset, so that could be coloring how I read code. |
| 15:01:31 | livcd | hmm i did exactly that and it is still broken. I'll try to update nim then |
| 15:01:34 | FromGitter | <RedBeard0531> @Araq, any thoughts on the setStrLength possible bug I mentioned earlier? |
| 15:04:25 | Araq | seems like an old bug that got never caught because it only allocates too much? |
| 15:06:49 | FromGitter | <RedBeard0531> ok, I'll PR when I get some time. Just wanted to make sure my understanding of the code was correct. |
| 15:08:55 | * | zahary joined #nim |
| 15:10:15 | FromGitter | <alehander42> @ervinbosenbacher about py2nim: actually you have to pass the full path to the file currently, that should be fixed soon ⏎ overally, it's still in an early stage indeed, but these days we might upload some more complete example translating a lib/project |
| 15:10:51 | Araq | alehander42: why not base it on PyPy? |
| 15:13:01 | FromGitter | <alehander42> why base it on PyPy? |
| 15:13:15 | * | skelett joined #nim |
| 15:13:54 | Araq | dunno, it can deduce some type information at runtime too |
| 15:14:26 | Araq | well I have no clue how python-deduckt works |
| 15:16:25 | FromGitter | <ervinbosenbacher> @alehander42 ok it worked! |
| 15:17:19 | FromGitter | <ervinbosenbacher> @alehander42 how do i get the output? |
| 15:18:00 | FromGitter | <ervinbosenbacher> ok got it |
| 15:18:08 | Araq | this is too clever for me |
| 15:18:39 | Araq | no idea how you can make Python output type information with this little amount of Python code and no core interpreter fixes |
| 15:21:17 | * | endragor joined #nim |
| 15:25:27 | * | endragor quit (Ping timeout: 240 seconds) |
| 15:26:49 | FromGitter | <ervinbosenbacher> @alehander42 python-deduckt.json |
| 15:27:06 | FromGitter | <ervinbosenbacher> how is it gets generated? its been cashed afer I compiled sum |
| 15:27:14 | FromGitter | <ervinbosenbacher> then I have deleted it |
| 15:28:55 | FromGitter | <alehander42> @Araq well Python has trace hooks, so you can execute a function on each call/line/return etc. We collect info on each call/return, then process it a bit and serialize it (we also merge some of it with the serialized python ast) |
| 15:29:18 | FromGitter | <alehander42> there are several other tools generating mypy definitions with a similar approach: pyannotate and monkeytype |
| 15:29:50 | FromGitter | <alehander42> of course, you can probably do even more powerful stuff with a patched vm, but it would be harder to install/maintain |
| 15:30:26 | * | Yardanico quit (Remote host closed the connection) |
| 15:30:40 | FromGitter | <RedBeard0531> Do you ignore inputs that result in TypeError? |
| 15:30:53 | FromGitter | <ervinbosenbacher> apart from the examples i wasnt able to transpile anythng else |
| 15:31:16 | FromGitter | <ervinbosenbacher> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5a538ec319147ac3231b5639] |
| 15:31:23 | * | Yardanico joined #nim |
| 15:31:59 | FromGitter | <ervinbosenbacher> i took a random python file of mine with full path |
| 15:32:04 | FromGitter | <ervinbosenbacher> and got this |
| 15:32:31 | FromGitter | <RedBeard0531> I'm always afraid of using tests as feedstock for anything that analyzes runtime behavior (such as PGO) because by their nature, many tests are testing edge cases like invalid inputs. |
| 15:33:08 | FromGitter | <ervinbosenbacher> i go back to my hand based transpilation of our python code to nim |
| 15:33:50 | FromGitter | <ervinbosenbacher> python-deduckt.json didnt get generated |
| 15:33:56 | FromGitter | <alehander42> @ervinbosenbacher the intended usage is to pass a file in the context of a package that results in good "function" coverage: usually a test file. |
| 15:34:12 | * | endragor joined #nim |
| 15:34:27 | FromGitter | <alehander42> anyway, it's probably too early to apply it to random real world folders |
| 15:34:34 | FromGitter | <alehander42> as I already said |
| 15:34:36 | FromGitter | <ervinbosenbacher> ok I wait |
| 15:34:44 | FromGitter | <ervinbosenbacher> just got excied |
| 15:34:58 | FromGitter | <ervinbosenbacher> as I am starting to do this by hand |
| 15:36:54 | FromGitter | <ervinbosenbacher> i have been working on python 3.6 to c++14 transpiler maybe I just will use that generate hooks using nim |
| 15:37:27 | FromGitter | <alehander42> @RedBeard0531 that's a valid concern, other possible problem with unit tests can be mocks etc |
| 15:38:40 | FromGitter | <alehander42> running it with integration tests might produce more reliable results sometimes |
| 15:39:03 | * | endragor quit (Ping timeout: 256 seconds) |
| 15:40:49 | FromGitter | <dom96> @alehander42 looks like an awesome tool :) |
| 15:42:32 | FromGitter | <ervinbosenbacher> yes it is, @alehander42 can I help instead of me trying to convert over 100k python code? :D |
| 15:45:22 | * | craigger quit (Quit: bye) |
| 15:45:33 | * | craigger joined #nim |
| 15:48:17 | * | Yardanico quit (Ping timeout: 248 seconds) |
| 15:50:24 | * | Yardanico joined #nim |
| 15:50:46 | * | craigger quit (Quit: bye) |
| 15:50:54 | * | craigger joined #nim |
| 15:59:11 | Araq | alehander42: one guiding principle could be "no matter what, we'll manage to output Nim syntax" |
| 15:59:25 | * | xet7 quit (Ping timeout: 268 seconds) |
| 15:59:35 | Araq | and then the more your runtime profile uncovers, the better Nim code is produced |
| 15:59:54 | * | floppydh quit (Quit: WeeChat 2.0.1) |
| 16:00:04 | Araq | so that people are not left with "ok, the tool crashes, back to hand translation" |
| 16:00:11 | Araq | :-) |
| 16:01:54 | Araq | but hey, you can also use the py2nim translator I wrote as a forum post (nah, just kidding) |
| 16:05:17 | * | gangstacat quit (Quit: Ĝis!) |
| 16:07:53 | FromGitter | <ervinbosenbacher> :D @Araq I was kidding not giving up |
| 16:08:32 | FromGitter | <ervinbosenbacher> Currently what I do is that using Cython I convert python code to C and then build a nim file y hand to call those functions |
| 16:08:38 | FromGitter | <ervinbosenbacher> without kidding |
| 16:09:15 | FromGitter | <ervinbosenbacher> downside is the huge C code and that I have to link with python lib |
| 16:09:54 | FromGitter | <ervinbosenbacher> the reason for my gazillion questions how do I interface with C |
| 16:10:19 | Araq | I wasn't kidding though. |
| 16:10:31 | FromGitter | <ervinbosenbacher> i know |
| 16:10:54 | FromGitter | <data-man> @Araq: What else should I do for my truecolored PR? |
| 16:11:08 | * | kivyion joined #nim |
| 16:11:10 | FromGitter | <ervinbosenbacher> I understand that there is a lot of work i am here to help, my main focus is on the compiler, destructor etc and python tooling |
| 16:13:45 | * | gangstacat joined #nim |
| 16:14:45 | * | xet7 joined #nim |
| 16:15:54 | Araq | data-man: your cache is still wrong. |
| 16:16:05 | Araq | fortunately it's also not required |
| 16:16:33 | Araq | if you want to avoid the string concats, use an array[size, char] |
| 16:16:44 | Araq | and pass that instead to the write |
| 16:17:03 | FromGitter | <alehander42> @dom96 thanks. (you should tag @zah too :D ) |
| 16:17:32 | Araq | you misunderstand my "This version works for a disabled taint mode too." remark, but it's not important, I can clean up your PR before merging it |
| 16:17:51 | * | dandevelo joined #nim |
| 16:18:00 | FromGitter | <alehander42> @Araq absolutely, the tool should be fault tolerant and it shouldn't crash in principle |
| 16:18:43 | Araq | alehander42: glad you agree. :-) |
| 16:18:46 | FromGitter | <alehander42> in the worst case it should produce warn messages and commented out original code/suggestions (still not the case, but that's the design goal) |
| 16:19:15 | FromGitter | <data-man> @Araq: Why not required? Cache significantly speeds up the output. |
| 16:19:38 | Araq | "Maybe to add (enable|disable|isSupported)TrueColor procs?" yes. |
| 16:19:48 | FromGitter | <data-man> Ok |
| 16:20:00 | FromGitter | <data-man> Thanks! |
| 16:20:10 | FromGitter | <alehander42> @ervinbosenbacher ofc, contributions would be great |
| 16:21:03 | * | Jesin quit (Quit: Leaving) |
| 16:22:17 | Araq | data-man: "\x1b[38;2;" & $rgb.r & ";" & $rgb.g & ";" & $rgb.b & 'm' can be written in a way that is faster than hashing it and looking it up |
| 16:22:34 | * | nsf quit (Quit: WeeChat 2.0.1) |
| 16:23:02 | Araq | and it will make the code shorter too |
| 16:23:02 | FromGitter | <ervinbosenbacher> yes I will |
| 16:23:11 | FromGitter | <ervinbosenbacher> currently trying to figure that one out |
| 16:23:25 | FromGitter | <ervinbosenbacher> Ihave no desire to rewrite our entre python codebase by hand |
| 16:23:39 | FromGitter | <ervinbosenbacher> would be silly, better to help you |
| 16:23:57 | Araq | ervinbosenbacher: the approach I used for Nim was a bit different than py2nim's approach though |
| 16:24:14 | FromGitter | <data-man> @Araq: strformat? :-) |
| 16:25:04 | Araq | I had a tool pas2nim that could translate Pascal to Nim. I introduced custom annotations for the Pascal code to help the pas2nim tool |
| 16:25:37 | Araq | and for a long time I kept working on the Pascal version until bootstrapping worked reliably with the translated compiler. |
| 16:25:50 | FromGitter | <ervinbosenbacher> yes @Araq just explained I am using cpython to compile python to C then by hand the nim interface |
| 16:26:08 | Araq | yeah but that might not your best option |
| 16:26:17 | FromGitter | <ervinbosenbacher> i know |
| 16:26:25 | FromGitter | <ervinbosenbacher> what would be the best option? |
| 16:26:35 | Araq | if py2nim allows for custom annotations in the Python code you can try what I did |
| 16:26:52 | FromGitter | <ervinbosenbacher> yes just whati was thinking |
| 16:27:16 | Araq | the trick is to not edit the translated code too quickly |
| 16:27:21 | FromGitter | <ervinbosenbacher> aka nuitka |
| 16:27:44 | Yardanico | well nuitka basically translated python code into python C API calls |
| 16:27:52 | Yardanico | and optimizes some constructs |
| 16:27:56 | Yardanico | *translates |
| 16:28:28 | FromGitter | <ervinbosenbacher> yes I was refering to annotations. but then you have to pile thrugh tons of code and add annotations. |
| 16:28:52 | FromGitter | <ervinbosenbacher> Cython does the same thing you still have to link to python |
| 16:29:13 | Araq | yes, it also depends on your goals. |
| 16:29:21 | * | Trustable joined #nim |
| 16:29:32 | FromGitter | <ervinbosenbacher> right quick and dirty solution until the tool is not ready |
| 16:29:41 | FromGitter | <ervinbosenbacher> iteration 1 if you will |
| 16:30:00 | FromGitter | <ervinbosenbacher> python to c (somehow) call it from Nim |
| 16:30:03 | Araq | with my approach I never had a wasteland of code I couldn't run. |
| 16:30:43 | FromGitter | <ervinbosenbacher> so what did you do again? You were talking about pascal |
| 16:31:13 | Araq | yeah ofc the problem was easier since I only used the subset of Pascal I know would translate easily |
| 16:32:14 | Araq | and I had this lovely {@ignore} pascal code; {@emit Nim version} construct |
| 16:32:34 | * | Jesin joined #nim |
| 16:32:38 | Araq | I think c2nim now has something similar. |
| 16:34:58 | Araq | my point is "quick and dirty solutions" are good if you don't end up in a messy situation like "Python version too slow and hasn't kept up with the recent features, Nim version has more features but crashes" :-) |
| 16:35:57 | * | dandevelo quit (Remote host closed the connection) |
| 16:36:22 | FromGitter | <alehander42> something like magic comment directives for py2nim might be useful indeed |
| 16:36:31 | * | dandevelo joined #nim |
| 16:36:57 | FromGitter | <ervinbosenbacher> what I do in detail is that I convert the python code to cpython code, basically. And adding annotations to the python code. |
| 16:37:15 | Araq | that can never hurt :-) |
| 16:38:07 | FromGitter | <alehander42> some other hooks should be also eventually possible: mapping some specific for your project dynamic components to handwritten nim stuff (eg decorators to provided by the user macros), providing our idioms dsl for the user etc |
| 16:38:45 | * | PMunch quit (Quit: Leaving) |
| 16:38:54 | Araq | alehander42: you also often need abstraction breaking transformations |
| 16:39:03 | Yardanico | ervin: don't confuse cpython with cython :) |
| 16:39:16 | FromGitter | <alehander42> loading info from mypy annotations is also possibly useful, the reason it didn't have much priority is that mypy definitions still don't seem to be popular enough (it's not like definitely-typed and ts) |
| 16:39:59 | Araq | like regex based rewrite rules that are specific to the project and don't generalize well |
| 16:40:01 | FromGitter | <alehander42> @Araq that's true, as a whole often you need architectural changes etc which just can't be expressed with a general translation engine |
| 16:40:14 | dom96 | alehander42: already told him, we've got a Slack channel (possibly should open that to everyone) |
| 16:40:28 | Araq | dom96: don't you dare. |
| 16:40:33 | dom96 | another suggestion I made: type hints in Python could help |
| 16:40:55 | dom96 | Araq: You can still have a private channel for your secret zahary/Araq discussions :P |
| 16:40:55 | FromGitter | <alehander42> yeah, two messages up I gave my view on that |
| 16:41:15 | Araq | dom96: they are not private, you should read them :P |
| 16:41:26 | dom96 | fine, me/you and zahary then :P |
| 16:41:33 | FromGitter | <alehander42> but if somebody wants to contribute, mypy support might be a very cool start |
| 16:42:48 | Araq | alehander42: better would be something like Nim's using statement. tell the tool externally how to infer the type based on the parameter name |
| 16:43:31 | FromGitter | <ervinbosenbacher> Then I will use cython for now, second step I will try to help with py2nim |
| 16:43:31 | Araq | how many def signatures are there to annotate? many thousands. how many different parameter names are used? much fewer, I bet. |
| 16:43:51 | dom96 | btw, important-ish announcement. I've decided to change the way Nimble works: if `bin` is specified, Nimble will no longer install .nim files, only compile the binary and install it (this is a more common case so it makes sense by default IMO). |
| 16:44:01 | FromGitter | <ervinbosenbacher> dont know I need to get this done now, we will work on improvements later |
| 16:44:44 | FromGitter | <ervinbosenbacher> if I export functins from cython I can have a quick nim interface job done |
| 16:44:58 | FromGitter | <ervinbosenbacher> its not as glamorous and easy as the tool but better |
| 16:45:08 | FromGitter | <ervinbosenbacher> than rewriting the while thing |
| 16:46:27 | FromGitter | <alehander42> yes, for a big project as yours, probably some process well tweaked for it might be best (at least on the current stage of py2nim) |
| 16:47:09 | FromGitter | <ervinbosenbacher> yes I also get my hand dirty with Nim and get insights how to link the stuff together then I can help with the tool |
| 16:47:44 | Araq | you can build your Nim code as a DLL and call it from Python. |
| 16:49:00 | FromGitter | <ervinbosenbacher> yes but the main point is not to have python right now. |
| 16:49:11 | * | skelett quit (Ping timeout: 240 seconds) |
| 16:50:06 | FromGitter | <ervinbosenbacher> that is a good option too tw |
| 16:50:08 | FromGitter | <ervinbosenbacher> tw |
| 16:50:12 | FromGitter | <ervinbosenbacher> testedt yesterday |
| 16:50:31 | FromGitter | <ervinbosenbacher> sorry for the typos, on the phone |
| 16:51:11 | FromGitter | <RedBeard0531> @dom96 will nimble still support hybrid lib/bin packages? |
| 16:51:56 | dom96 | RedBeard0531: yep, but you will need to explicitly tell Nimble that you want the .nim files installed for that |
| 16:54:33 | * | chemist69 quit (Ping timeout: 265 seconds) |
| 16:54:42 | * | skelett joined #nim |
| 17:03:06 | FromGitter | <ervinbosenbacher> i still think Nim is the best thing currently we just need to figure out how to make it even more awesome and these are just teething issues that we are going to fix |
| 17:05:06 | * | azur_kind joined #nim |
| 17:07:59 | * | chemist69 joined #nim |
| 17:08:28 | * | claudiuinberlin quit (Quit: Textual IRC Client: www.textualapp.com) |
| 17:13:25 | * | darithorn joined #nim |
| 17:15:41 | skelett | Hello, I'm trying to implement an application which should listen on SIGUSR1 for toggeling an inplace configuration reload. In C, only certain operations/functions are safe to be called inside a signal handler. Does anybody know whether the nim compiler pays attention so that signal handlers are always safe or is this up to me (as in C)? Currently I'm using the `onSignal` template from the `posix` module. |
| 17:17:39 | dom96 | Unfortunately Nim doesn't do any checking for this AFAIK |
| 17:17:57 | dom96 | But I think Nim's effect system would be perfect for this, somebody just has to define what is allowed |
| 17:22:07 | skelett | dom96: Ok thank you. Indeed that could solve the problem in the future. Afaik `man 7 signal-safety` should give an overview over all allowed functions. |
| 17:22:32 | dom96 | skelett: feel free to create an issue for this :) |
| 17:28:23 | Araq | skelett, dom96 .gcsafe is a good first approximation of what is valid in a signal handler, I think |
| 17:29:45 | * | miran joined #nim |
| 17:30:07 | dom96 | Here are the relevant man pages: http://man7.org/linux/man-pages/man7/signal-safety.7.html |
| 17:31:31 | dom96 | huh, this implies that printf uses a global buffer, am I misreading? I find that surprising. |
| 17:32:28 | * | yglukhov quit (Remote host closed the connection) |
| 17:33:01 | FromGitter | <RedBeard0531> I think all of the FILE* functions in c use locking and locking inside a signal hander is a big no-no (possibly unless it is reentrent) |
| 17:33:19 | skelett | dom96: printf is not a safe function |
| 17:33:22 | dom96 | ahh |
| 17:34:48 | skelett | But the low level write function should be IIRC |
| 17:34:49 | FromGitter | <RedBeard0531> write() is safe because it is unbuffered and non-locking (at least in user-space) |
| 17:36:42 | FromGitter | <RedBeard0531> IIRC snprintf is banned because some implementations malloc internally which is another no-no |
| 17:36:55 | dom96 | Fun :) |
| 17:37:53 | * | sendell quit (Remote host closed the connection) |
| 17:41:19 | * | claudiuinberlin joined #nim |
| 17:42:29 | dom96 | oh, just realised: https://developers.google.com/open-source/gsoc/timeline |
| 17:43:20 | dom96 | If we want to apply we've got until January 23rd |
| 17:44:53 | dom96 | Last time we made an attempt was 2016 it seems: https://github.com/nim-lang/Nim/wiki/GSoC-2016-Ideas |
| 17:46:52 | * | sz0 joined #nim |
| 18:01:08 | * | yglukhov joined #nim |
| 18:03:09 | * | themagician joined #nim |
| 18:04:37 | shashlick | araq: from earlier, no I didn't have to use terminal, fzf does most of that |
| 18:05:19 | * | yglukhov quit (Ping timeout: 248 seconds) |
| 18:07:13 | miran | dom96: we want to apply, right? :) |
| 18:07:37 | dom96 | we do, but I need people who are willing to mentor/write up tasks for this |
| 18:15:02 | miran | what of those 2016 tasks has been completed and what is still open? |
| 18:17:04 | dom96 | I think most of them are no longer applicable |
| 18:17:55 | dom96 | One task I can think of after looking at this is a webassembly target for Nim |
| 18:19:33 | miran | and is there a roadmap with some clear goals for nim in next year of two? |
| 18:20:38 | dom96 | sort of, me and Araq have been meaning to make it public for a while now. |
| 18:21:15 | Yardanico | it's private? :) |
| 18:21:17 | miran | please do it :) |
| 18:22:29 | miran | one thing at top of my mind when it comes to nim improvements is nim's vscode plugin |
| 18:22:53 | miran | as this is something that most of us use every day |
| 18:23:00 | dom96 | That's true |
| 18:23:11 | dom96 | I would like to see a language server implementation |
| 18:23:17 | dom96 | That builds on top of nimsuggest |
| 18:23:35 | dom96 | and does its own parsing for faster responses in some cases |
| 18:23:42 | FromGitter | <data-man> https://gist.github.com/data-man/62b186bfd39694cfbc9976321f11359b ⏎ ⏎ |benchGetColor.nim | relative | time/iter | iters/s ⏎ --- | ---| --- | --- ⏎ getColorSimpleConcat | | 80.15ms | 12.48 ... [https://gitter.im/nim-lang/Nim?at=5a53b72e0505ad8b45cce4cf] |
| 18:28:54 | Araq | still not as optimized as I imagine :P |
| 18:30:53 | FromGitter | <data-man> You can always improve after merging :-D |
| 18:31:06 | FromGitter | <Araq> I can't merge, it's not correct |
| 18:31:38 | FromGitter | <Araq> you cannot use SharedTable with string keys unless you protect() the string |
| 18:31:58 | FromGitter | <Araq> it's really messy |
| 18:34:24 | FromGitter | <data-man> Then I will remove caching for threads mode. ⏎ It's ok? |
| 18:34:26 | * | nsf joined #nim |
| 18:34:56 | * | azur_kind quit (Remote host closed the connection) |
| 18:35:36 | * | endragor joined #nim |
| 18:36:14 | * | endragor_ joined #nim |
| 18:36:48 | Araq | yes! |
| 18:37:04 | Araq | oh, I misread, oh well |
| 18:40:05 | * | endragor quit (Ping timeout: 240 seconds) |
| 18:40:18 | FromGitter | <data-man> fmt is thread safe? :-) |
| 18:40:28 | * | endragor_ quit (Remote host closed the connection) |
| 18:41:00 | * | endragor joined #nim |
| 18:42:15 | * | endragor quit (Remote host closed the connection) |
| 18:42:49 | * | endragor joined #nim |
| 18:44:06 | * | endragor quit (Remote host closed the connection) |
| 18:45:06 | * | Trustable quit (Remote host closed the connection) |
| 18:45:09 | FromGitter | <xmonader> Hi |
| 18:45:32 | FromGitter | <xmonader> I see https://github.com/nim-lang/Nim/issues/6778 all of the methods to be deprecated are already done hmm |
| 18:50:43 | Araq | data-man: it is. |
| 18:59:40 | FromGitter | <RedBeard0531> Why deprecate repr? In languages without it (basically everything but python and nim) I end up missing it. |
| 19:00:30 | * | zahary quit (Quit: Leaving.) |
| 19:05:58 | * | yglukhov joined #nim |
| 19:06:38 | Araq | macros.repr is not system.repr |
| 19:06:48 | Araq | and it should have been macros.$ |
| 19:06:53 | * | Ven`` joined #nim |
| 19:07:19 | Araq | but I don't want to do this anymore either, pointless renamings everywhere when we've much better things to do |
| 19:07:39 | FromGitter | <RedBeard0531> ahh, got scared for a second :) |
| 19:10:25 | * | yglukhov quit (Ping timeout: 248 seconds) |
| 19:15:40 | * | miran quit (Quit: Konversation terminated!) |
| 19:17:48 | FromGitter | <RedBeard0531> Even if the renames aren't a good use of time, extending nimsuggest (or building a similar tool) to support renames like this seems pretty valuable. I know I use that functionality in rtags in our c++ codebase fairly frequently. It is great being able to hit a few keys and have all usages rewritten. |
| 19:17:48 | * | Ven`` quit (Read error: Connection reset by peer) |
| 19:18:05 | * | Ven`` joined #nim |
| 19:22:35 | * | zahary joined #nim |
| 19:32:12 | Araq | I can do that with nimgrep, more or less. |
| 19:32:38 | Araq | the problem is I don't have "all usages", I have plenty of Nim projects on multiple different machines |
| 19:33:09 | Araq | and "project X now only works with the latest Nim" is not much of a feature for everybody else |
| 19:33:21 | * | Serenitor quit (Ping timeout: 248 seconds) |
| 19:38:09 | * | Ven`` quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 19:59:38 | FromGitter | <ervinbosenbacher> we should keep the stuff simple and consistent without renaming things which can break things. Imho once the language is finalized its should be kept that way. Countless languages fall in the same trap. |
| 19:59:53 | FromGitter | <ervinbosenbacher> just my cents out of the blue. |
| 20:00:04 | * | mckendricks joined #nim |
| 20:00:12 | FromGitter | <RedBeard0531> I think macros.repr highlights the limitations of using nimgrep for this, but I get your larger point. Breaking stuff without good reasons sucks for everyone. I've gotten too used to working on a unified codebase where I can change everything that doesn't change rpc inputs and outputs in a single commit. |
| 20:03:02 | * | rauss quit (Quit: WeeChat 2.0.1) |
| 20:04:08 | * | rauss joined #nim |
| 20:06:18 | * | martinium joined #nim |
| 20:11:02 | * | Vladar quit (Quit: Leaving) |
| 20:12:43 | Yardanico | I'm currently trying to run py2nim on a ~1k loc project :) |
| 20:13:12 | FromGitter | <ervinbosenbacher> Ok let me know how you getting on |
| 20:15:38 | Yardanico | wow, python-deduckt.json is 123mb :) |
| 20:16:08 | FromGitter | <ervinbosenbacher> how big is the result ? |
| 20:17:22 | FromGitter | <ervinbosenbacher> i mean the nim file(s) |
| 20:19:48 | Yardanico | Well I'm redoing it (I merged tests and project file into one file) |
| 20:20:03 | FromGitter | <ervinbosenbacher> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5a53d27351915d962afdc472] |
| 20:20:22 | FromGitter | <ervinbosenbacher> i cant submdule init due to permissions |
| 20:21:10 | Yardanico | yeah I fixed that |
| 20:21:26 | * | yglukhov joined #nim |
| 20:21:31 | FromGitter | <ervinbosenbacher> thats 2 minutes ago |
| 20:21:32 | Yardanico | change file .git/config |
| 20:21:38 | FromGitter | <ervinbosenbacher> ok |
| 20:21:44 | Yardanico | change url in submodule to https://github.com/metacraft-labs/python-deduckt |
| 20:21:46 | Yardanico | it will work |
| 20:22:07 | FromGitter | <ervinbosenbacher> thx |
| 20:25:37 | * | yglukhov quit (Ping timeout: 248 seconds) |
| 20:28:52 | Yardanico | well py2nim sigsegvs :) |
| 20:29:10 | FromGitter | <martinium> Happy New Year guys! |
| 20:29:26 | Yardanico | you're a bit late :P |
| 20:30:33 | FromGitter | <martinium> yeah :) |
| 20:30:43 | * | zahary quit (Quit: Leaving.) |
| 20:30:57 | FromGitter | <martinium> haven’t been in here in a bit so wanted to say hi and catch up |
| 20:31:05 | * | jjido joined #nim |
| 20:31:14 | dom96 | alehander42: why not use nimble instead of submodules? |
| 20:31:49 | Yardanico | because his submodule is in python ? :) |
| 20:32:01 | Yardanico | it's a python project for tracing python types at runtime |
| 20:32:14 | Yardanico | (this is possible with some magic) |
| 20:33:41 | * | jjido quit (Read error: Connection reset by peer) |
| 20:33:53 | Yardanico | grumpy works the same way btw |
| 20:35:09 | Yardanico | https://github.com/google/grumpy/tree/master/compiler |
| 20:35:32 | Yardanico | but it seems it uses static analysis |
| 20:38:15 | FromGitter | <ervinbosenbacher> right now I will use nim modules from python following the advice |
| 20:38:35 | * | SitiSchu joined #nim |
| 20:38:50 | FromGitter | <ervinbosenbacher> mnaged to comile python for for our embdded platform so we go with that and gradually I will add new stuff using nim |
| 20:39:47 | * | SitiSchu quit (Client Quit) |
| 20:40:08 | * | SitiSchu joined #nim |
| 20:42:12 | * | jjido joined #nim |
| 20:43:51 | * | Yardanico quit (Remote host closed the connection) |
| 20:45:14 | * | jjido quit (Read error: No route to host) |
| 20:49:22 | darithorn | how do I use macros.quote if I want to change the op parameter? |
| 20:50:17 | * | jjido joined #nim |
| 20:50:46 | Araq | https://github.com/nim-lang/Nim/issues/6989 now somebody figures out the reason please |
| 20:50:50 | Araq | see you later |
| 21:00:30 | dom96 | Yardanico: oh lol, my bad. |
| 21:02:30 | jjido | What does ^rot do? |
| 21:07:20 | FromGitter | <zetashift> ^rot = rot.len |
| 21:07:49 | FromGitter | <zetashift> I think actually: https://nim-lang.org/docs/tut1.html#advanced-types-subranges under 'Slices' |
| 21:17:20 | * | jjido quit (Read error: Connection reset by peer) |
| 21:20:23 | * | jjido joined #nim |
| 21:25:05 | * | yglukhov joined #nim |
| 21:29:23 | * | yglukhov quit (Ping timeout: 255 seconds) |
| 21:35:10 | * | miran joined #nim |
| 21:36:04 | miran | Araq: i just seen your comment about strange behaviour in 0.17.3 |
| 21:36:23 | miran | it seems like ^ is the cause? |
| 21:37:03 | * | jjido quit (Read error: No route to host) |
| 21:37:38 | miran | jjido: ^x is like -x in python, indexing from the right |
| 21:45:04 | * | jjido joined #nim |
| 21:51:47 | * | martinium quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 21:51:59 | * | deech joined #nim |
| 21:53:08 | * | miran quit (Quit: Konversation terminated!) |
| 21:54:07 | * | martinium joined #nim |
| 22:27:17 | * | nsf quit (Quit: WeeChat 2.0.1) |
| 22:38:50 | * | dddddd quit (Remote host closed the connection) |
| 22:47:08 | FromGitter | <data-man> https://github.com/data-man/SciTENim ⏎ Just for fun :) |
| 22:56:38 | * | Serenitor joined #nim |
| 22:57:28 | * | claudiuinberlin quit (Quit: Textual IRC Client: www.textualapp.com) |
| 23:06:28 | * | vlad1777d joined #nim |
| 23:08:00 | * | Yardanico[Phone] joined #nim |
| 23:25:52 | * | jjido quit (Read error: Connection reset by peer) |
| 23:33:26 | * | mckendricks quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 23:52:23 | * | marenz__ joined #nim |