<< 26-09-2018 >>
00:00:56*kapil___ quit (Quit: Connection closed for inactivity)
00:01:16*Alforz20 quit (Remote host closed the connection)
00:08:48*adulteratedjedi5 joined #nim
00:08:57*adulteratedjedi5 quit (Remote host closed the connection)
00:10:26*banym1 joined #nim
00:12:08*banym1 quit (Remote host closed the connection)
00:13:57*bus7d joined #nim
00:17:05*skellock quit (Quit: WeeChat 1.9.1)
00:18:24*bus7d quit (Read error: Connection reset by peer)
00:20:51*csaunders21 joined #nim
00:21:41*lstdgtfp_ joined #nim
00:22:26*csaunders21 quit (Remote host closed the connection)
00:25:07*lstdgtfp_ quit (Remote host closed the connection)
00:28:11*hays_ joined #nim
00:29:29*hays_ quit (Read error: Connection reset by peer)
00:38:10*bews16 joined #nim
00:41:42*gtonello_ joined #nim
00:42:50*gtonello_ quit (Remote host closed the connection)
00:42:55*bews16 quit (Remote host closed the connection)
00:45:13*bviktor joined #nim
00:49:58*bviktor quit (Remote host closed the connection)
00:50:49*optiz0r28 joined #nim
00:51:05*optiz0r28 quit (Remote host closed the connection)
00:53:16AlexMaxshashlick_: *nod*
00:53:49AlexMaxI feel like my nimnuklear work isn't done, but I kinda wanna upstream what I have anyway
00:54:26AlexMaxwithout my changes, it doesn't even work
00:54:35AlexMaxwith my changes, it works, but there are some annoying type problems to work out
01:03:46*noonien quit (Quit: Connection closed for inactivity)
01:05:29shashlick_Ya sure please do
01:12:31*vivus quit (Remote host closed the connection)
01:15:20*zeus21 joined #nim
01:20:12*zeus21 quit (Remote host closed the connection)
01:20:49AlexMaxshashlick_: https://github.com/genotrance/nimnuklear/pull/2
01:26:23*jya2 joined #nim
01:27:27*ChrisWarrick24 joined #nim
01:29:15*jya2 quit (Remote host closed the connection)
01:29:36*ChrisWarrick24 quit (Remote host closed the connection)
01:30:01*dave492515 joined #nim
01:30:40*phry8 joined #nim
01:33:10*dave492515 quit (Remote host closed the connection)
01:33:22*drale2k13 joined #nim
01:34:18*phry8 quit (Remote host closed the connection)
01:34:37*drale2k13 quit (Remote host closed the connection)
01:35:24FromGitter<gogolxdong> why don't you just work on https://github.com/zacharycarter/nuklear-nim
01:37:33FromGitter<gogolxdong> Example works with some changes, also got some issues on editString of our own testing.
01:44:47*asilenkov13 joined #nim
01:45:36*asilenkov13 quit (Remote host closed the connection)
01:46:43*Taneb20 joined #nim
01:50:43*Taneb20 quit (Remote host closed the connection)
01:53:15*vquicksilver10 joined #nim
01:54:36*noctavian28 joined #nim
01:55:23*noctavian28 quit (Read error: Connection reset by peer)
01:57:16*vquicksilver10 quit (Remote host closed the connection)
01:59:18shashlick_@gogolxdong: this was with zachcarter's blessings
01:59:30shashlick_There were lots of changes upstream
01:59:58shashlick_Now using nimgen to automate the wrapper generation
02:10:40shashlick_@AlexMax: merged
02:11:00shashlick_You need to join the nimgen gitter channel
02:11:36shashlick_https://gitter.im/nimgen/Lobby
02:16:20*ironus19 joined #nim
02:16:24*ironus19 quit (Remote host closed the connection)
02:29:06AlexMax@gogolxdong We were actually in communication with zacharycarter already. He has had to turn his attention away from his bindings due to other obligations, and nimgen actually makes the hard first step of properly exporting the C library much easier
02:30:17c15ade4vs code golang module support
02:30:17FromGitter<gogolxdong> that's to say, you take over nuklear atm.
02:31:29AlexMaxI don't actually want to take over the nuklear nim bindings. But zach's current nim bindings seemed incomplete and didn't keep up with a years worth of changes to the upstream nuklear library itself
02:32:28AlexMaxI think zach's intention is to, when he gets time, use the nimgenned binding as an implementation detail to help further expand the scope of his library
02:32:53AlexMaxIt's a lot of binding code generated that he doesn't have to write by hand.
02:33:29FromGitter<gogolxdong> exactly.
02:34:30AlexMaxAnd all I was doing was helping the nimgenned binding stuff actually work out of the box
02:36:32AlexMaxthat helps me (because I wanted to use nuklear), that helps zach (since it gets rid of a bunch of work he has to do) and it helps shashlick (since it's yet another successfully nimgenn'ed library)
02:36:39AlexMaxhooray for open source!
02:36:41AlexMaxeverybody wins!
02:36:56FromGitter<gogolxdong> brilliant.
02:38:09AlexMax🎶 all we do is win win win 🎶
02:39:43FromGitter<gogolxdong> lol
02:56:54*ftsf joined #nim
03:12:35*mutin-s joined #nim
03:13:14*mutin-s quit (Remote host closed the connection)
03:20:58*lorenzh12 joined #nim
03:22:01*lorenzh12 quit (Remote host closed the connection)
03:22:56*Spydar00725 joined #nim
03:23:56*Spydar00725 quit (Remote host closed the connection)
03:34:06*rimd2r9 joined #nim
03:34:07*rrva13 joined #nim
03:35:21*rimd2r9 quit (Read error: Connection reset by peer)
03:36:36*rrva13 quit (Remote host closed the connection)
04:12:06*kapil___ joined #nim
04:50:08*Sembei joined #nim
04:52:58*Pisuke quit (Ping timeout: 245 seconds)
05:00:22*darithorn quit (Quit: Konversation terminated!)
05:02:03*jmarion joined #nim
05:06:45*jmarion quit (Remote host closed the connection)
05:11:02*aptanet13 joined #nim
05:12:49*mu5 joined #nim
05:14:28*aptanet13 quit (Remote host closed the connection)
05:15:47*mu5 quit (Remote host closed the connection)
05:23:55*nsf joined #nim
05:24:03*leorize quit (Ping timeout: 252 seconds)
05:24:54*DTZUZO joined #nim
05:25:23*DTZUZO quit (Read error: Connection reset by peer)
05:28:51FromGitter<retsyo> @Vindaar thanks. Currently I am using numpy/scipy which runs some slow, so some calculation is replaced by numba's cfunc in fact. It seems that numpy/scipy-like package is not in nim world currently and a long time is need to evolve.
05:32:17*AxForest12 joined #nim
05:34:31*KiNgMaR29 joined #nim
05:35:22*KiNgMaR29 quit (Remote host closed the connection)
05:37:12*AxForest12 quit (Read error: Connection reset by peer)
05:52:01*mongy joined #nim
05:52:04*ldiamond8 joined #nim
05:52:23*ldiamond8 quit (Remote host closed the connection)
05:52:41*mongy quit (Remote host closed the connection)
05:55:55*leorize joined #nim
06:04:31*Guest50053 joined #nim
06:06:09*Guest50053 quit (Remote host closed the connection)
06:19:08*Skainet joined #nim
06:21:28*Skainet quit (Remote host closed the connection)
06:22:48*twiddern18 joined #nim
06:23:08*ghormoon21 joined #nim
06:25:42*grahamg joined #nim
06:26:30*ghormoon21 quit (Remote host closed the connection)
06:27:07FromGitter<narimiran> @retsyo for some simple linear algebra you can use Neo. if you want to just to rewrite some slow parts in Nim and call that from Python (like you would do with Cython), take a look at Nimpy
06:27:08*twiddern18 quit (Remote host closed the connection)
06:27:18grahamgHi. I just started looking at nim today. I notice that the exception message produced with an IOError doesn't relay specific information that reflects C's errno. E.g. EACCESS or ENOENT. It just gives "cannot open".
06:27:41grahamg(ie. the open() proc)
06:27:43*Vermont18 joined #nim
06:28:18*Vermont18 quit (Remote host closed the connection)
06:28:37*wililupy28 joined #nim
06:29:02*wililupy28 quit (Remote host closed the connection)
06:29:10grahamgIs there a canonical way to open files that does provide useful error messages?
06:31:14FromGitter<gogolxdong> someone familiar with IPFS?
06:33:39*MatrixBot715 joined #nim
06:34:32*namho joined #nim
06:35:05FromGitter<Bennyelg> Morning
06:36:27*georgemp15 joined #nim
06:36:44*namho quit (Remote host closed the connection)
06:36:50*MatrixBot715 quit (Remote host closed the connection)
06:38:51FromGitter<gogolxdong> @mratsim does Status develops all in Nim?
06:39:56*georgemp15 quit (Remote host closed the connection)
06:45:42*z01d joined #nim
06:46:50*z01d quit (Remote host closed the connection)
06:50:57*da2ce74 joined #nim
06:55:17*da2ce74 quit (Remote host closed the connection)
06:58:45*Guest46016 joined #nim
06:59:41*zkms21 joined #nim
07:00:48*Guest46016 quit (Remote host closed the connection)
07:00:54*leru joined #nim
07:00:56*zkms21 quit (Read error: Connection reset by peer)
07:03:43*NegativeFlare10 joined #nim
07:04:11*grummund13 joined #nim
07:05:46*grummund13 quit (Remote host closed the connection)
07:05:59*NegativeFlare10 quit (Remote host closed the connection)
07:17:54FromGitter<mratsim> @gogolxdong Ethereum 2.0 will be all Nim for the backend
07:18:10FromGitter<mratsim> otherwise Ethereum 1.0 is backend in Go, FrontEnd in ClojureScript.
07:19:27FromGitter<gogolxdong> How many devs do you have ?
07:20:07FromGitter<mratsim> 6 Nim devs at the moment
07:20:50FromGitter<mratsim> Maybe half of the company are devs, 40~50 people otherwise? Not sure
07:23:45*mihait13 joined #nim
07:24:40*terrorjack17 joined #nim
07:24:41*mihait13 quit (Remote host closed the connection)
07:26:46*terrorjack17 quit (Remote host closed the connection)
07:29:40*Tanger quit (Quit: Leaving)
07:37:53FromGitter<gogolxdong> all work on Ethereum?
07:41:03*ToeSnacks_20 joined #nim
07:44:25*ToeSnacks_20 quit (Remote host closed the connection)
07:48:40*noonien joined #nim
07:50:11FromGitter<mratsim> All dev are on Ethereum yes
07:50:50FromGitter<gogolxdong> I cannot believe it. Ethereum 2.0 will be all Nim for the backend, how do you persuade Vitalik Buterin?
07:51:05FromGitter<mratsim> But Ethereum means a lot of things: P2P networking, Cryptography (and Ellptic curve math), Virtual machines
07:52:04FromGitter<mratsim> The goal of the foundation is to have many clients when Ethereum 2.0 is released, so that consensus issues are detected (today if there is an issue on Geth, since everyone is running Geth we might get a problem)
07:52:24FromGitter<mratsim> foundation == Ethereum Foundation
07:52:47FromGitter<gogolxdong> means Status got funding by Ethereum Foundation?
07:53:47FromGitter<mratsim> We applied but we started without funding (besides Status ICO)
07:53:59FromGitter<mratsim> Here is the funding detail: https://blog.ethereum.org/2018/01/02/ethereum-scalability-research-development-subsidy-programs/
07:55:01FromGitter<gogolxdong> got it , know a little how Foundation works.
07:57:45*clog21 joined #nim
08:00:03*clog21 quit (Read error: Connection reset by peer)
08:01:17FromGitter<gogolxdong> to be precise, didn't know how foundation works until now.
08:03:11*gmpreussner quit (Quit: kthxbye)
08:03:29*gmpreussner joined #nim
08:04:25FromGitter<gogolxdong> There was such a thought flashed through my head when I looked at the smart contract for the first time, don't believe it come true. Ethereum in Nim.
08:04:49FromGitter<mratsim> You need an underdog field to adopt an underdog language ;)
08:05:14FromGitter<mratsim> By the way you’re working at Tencent right? Congrats on the Starcraft 2 AI bot!
08:05:25FromGitter<mratsim> THat’s one thing I want to do with Arraymancer
08:05:32*NobodyCam15 joined #nim
08:06:26FromGitter<mratsim> @narimiran didn’t you say 0.19.0 was for yesterday?
08:06:45FromGitter<narimiran> @mratsim it seemed so
08:06:49FromGitter<gogolxdong> thanks, and I'm not. :) , we deal with Tencent Cloud only.
08:06:58FromGitter<mratsim> ah ok
08:07:00FromGitter<narimiran> https://github.com/nim-lang/Nim/commit/c240806756579c3375b1a79e1e65c40087a52ac5
08:07:34*shanth16 joined #nim
08:08:16*NobodyCam15 quit (Remote host closed the connection)
08:08:28*Liskni_si24 joined #nim
08:09:42FromGitter<gogolxdong> Arraymancer is a matrix operation library, not the algorithm , do you know what kind of neural network they used?
08:10:13FromGitter<mratsim> Paper is here: https://arxiv.org/abs/1809.07193
08:10:50*shanth16 quit (Remote host closed the connection)
08:10:55*xniega9 joined #nim
08:11:16FromGitter<gogolxdong> Arraymancer GPU backend is still WIP , is that hard?
08:11:37FromGitter<mratsim> it’s easier than CPU backend actually
08:11:43*Liskni_si24 quit (Remote host closed the connection)
08:11:47FromGitter<mratsim> at least for Cuda/CuDNN
08:11:58FromGitter<mratsim> OpenCL is no-man’s land
08:12:41FromGitter<mratsim> It seems like contrary to Alpha Go, for SC2 they encoded some human expertise in the bot according to the paper “165 macro actions"
08:12:45*Freebien24 joined #nim
08:13:05FromGitter<mratsim> and they used a variant of Deep Q-Learning
08:13:17*xniega9 quit (Read error: Connection reset by peer)
08:13:25FromGitter<gogolxdong> not pure AI.
08:13:33*Freebien24 quit (Remote host closed the connection)
08:13:55FromGitter<mratsim> not sure what “Distributed Rollout Infrastructure” is but I suppose it’s Monte-Carlo Tree Search
08:14:09*poljar joined #nim
08:14:10FromGitter<mratsim> to get a distribution of outcomes probability after each action
08:16:22FromGitter<gogolxdong> better AI lies in theory and research breakthrough of paper, neural network works as a blackbox atm.
08:16:25*poljar quit (Remote host closed the connection)
08:18:52FromGitter<mratsim> not true anymore
08:19:37FromGitter<mratsim> Copy-pasting myself from a data science slack: ⏎ ⏎ > SHAP (SHapley Additive exPlanations) is a unified approach to explain the output of any machine learning model. SHAP connects game theory with local explanations, uniting several previous methods [1-7] and representing the only possible consistent and locally accurate additive feature attribution method based on expectations ⏎ ⏎ 1)
08:19:37FromGitter... https://github.com/slundberg/shap ... [https://gitter.im/nim-lang/Nim?at=5bab4119fe37781110fc4f7a]
08:25:26*chemist69 joined #nim
08:31:48*Mahjong4 joined #nim
08:34:24*Mahjong4 quit (Remote host closed the connection)
08:36:45*ftsf quit (Remote host closed the connection)
08:37:06*ftsf joined #nim
08:37:57FromGitter<Vindaar> @retsyo see my response here: https://gist.github.com/Vindaar/aab2926425400fc57274b521e80398dd
08:38:07*inaki24 joined #nim
08:39:29federico3https://hackaday.com/2018/09/25/nim-writes-c-code-and-more-for-you/
08:40:18*skellock joined #nim
08:42:03*inaki24 quit (Remote host closed the connection)
08:42:11*skule9 joined #nim
08:42:31*PMunch joined #nim
08:43:31*dmorris joined #nim
08:43:58*skule9 quit (Remote host closed the connection)
08:46:09*joestewart9 joined #nim
08:47:39*joestewart9 quit (Remote host closed the connection)
08:48:16*f8l0 joined #nim
08:49:19*linduxed7 joined #nim
08:49:53*dmorris quit (Ping timeout: 252 seconds)
08:51:13*f8l0 quit (Remote host closed the connection)
08:51:29*chemist69 quit (Quit: WeeChat 1.9.1)
08:52:09*chemist69 joined #nim
08:53:18*dograt6 joined #nim
08:53:53*linduxed7 quit (Remote host closed the connection)
08:57:33*eofs2 joined #nim
08:57:34*ftsf quit (Read error: Connection reset by peer)
08:57:57*dograt6 quit (Remote host closed the connection)
09:00:15*eofs2 quit (Remote host closed the connection)
09:01:43*OS-7153_18 joined #nim
09:01:56*leorize quit (Ping timeout: 252 seconds)
09:02:04*OS-7153_18 quit (Remote host closed the connection)
09:02:40*outofbounds5 joined #nim
09:04:05*outofbounds5 quit (Remote host closed the connection)
09:04:47*hk6 joined #nim
09:04:50*hk6 quit (Remote host closed the connection)
09:07:56*ghavil joined #nim
09:08:01*ghavil is now known as Guest58840
09:08:05*altendky15 joined #nim
09:11:34*altendky15 quit (Read error: Connection reset by peer)
09:11:46*lerc7 joined #nim
09:12:23*Guest58840 quit (Remote host closed the connection)
09:14:12*krux02 joined #nim
09:15:05*lerc7 quit (Remote host closed the connection)
09:18:25*skule5 joined #nim
09:21:02*skule5 quit (Remote host closed the connection)
09:23:04*p_l12 joined #nim
09:26:12*couven92 quit (Quit: Client disconnecting)
09:27:52*p_l12 quit (Remote host closed the connection)
09:34:52*quaddle joined #nim
09:36:51*quaddle quit (Client Quit)
09:39:26*Mikail3 joined #nim
09:43:10*Mikail3 quit (Read error: Connection reset by peer)
09:43:15*onu3 joined #nim
09:47:26*onu3 quit (Remote host closed the connection)
09:51:08*arkblitz joined #nim
09:52:19*arkblitz quit (Remote host closed the connection)
09:52:37*tomoyat113 joined #nim
09:55:56*tomoyat113 quit (Remote host closed the connection)
10:03:15*ObsidianX joined #nim
10:03:41*ventures26 joined #nim
10:04:43*ObsidianX quit (Remote host closed the connection)
10:04:58*stefanos82 joined #nim
10:05:01*ventures26 quit (Read error: Connection reset by peer)
10:07:27*UTAN_dev10 joined #nim
10:08:34*thrillgore17 joined #nim
10:08:54FromGitter<narimiran> @Vindaar that might be a nice blog post!
10:09:00*UTAN_dev10 quit (Remote host closed the connection)
10:09:37*habibilus22 joined #nim
10:11:25*thrillgore17 quit (Remote host closed the connection)
10:11:50*habibilus22 quit (Remote host closed the connection)
10:21:00*Vladar joined #nim
10:21:56*tzui joined #nim
10:22:11*WillMoogle_ joined #nim
10:24:35*WillMoogle_ quit (Remote host closed the connection)
10:27:41*zzattack1 joined #nim
10:28:45*tzui quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
10:29:16*zzattack1 quit (Read error: Connection reset by peer)
10:32:05*shiranaihito joined #nim
10:35:27*tzui joined #nim
10:36:49*shiranaihito quit (Remote host closed the connection)
10:36:59*Gerrymandered10 joined #nim
10:38:49*Gerrymandered10 quit (Read error: Connection reset by peer)
10:38:53*DoctorDick22 joined #nim
10:39:20*DoctorDick22 quit (Read error: Connection reset by peer)
10:42:27*TheRealNoob4 joined #nim
10:43:01*TheRealNoob4 quit (Remote host closed the connection)
10:43:21*experienceRCOS joined #nim
10:51:51*experienceRCOS quit (Ping timeout: 252 seconds)
11:05:36*XXCoder3 joined #nim
11:05:44*npope7 joined #nim
11:05:59*XXCoder3 quit (Remote host closed the connection)
11:07:17*npope7 quit (Remote host closed the connection)
11:08:36*amunak joined #nim
11:10:06*amunak quit (Read error: Connection reset by peer)
11:10:14*halirc joined #nim
11:10:51*lakonig_ joined #nim
11:13:52*lakonig_ quit (Remote host closed the connection)
11:19:26*stnutt joined #nim
11:20:10*tobberydberg15 joined #nim
11:21:26*pugfantus10 joined #nim
11:21:54FromGitter<zacharycarter> that's true
11:22:31FromGitter<zacharycarter> it's a good conversation starter - not something to form a persuasive argument around
11:23:57*pugfantus10 quit (Remote host closed the connection)
11:24:10*stnutt quit (Remote host closed the connection)
11:28:14FromGitter<Vindaar> @narimiran yeah, that idea is what made me right this up in more than a few lines here, haha. Will try to fix that up a little in the next few days and turn it into one :)
11:28:15*tobberydberg15 quit (Ping timeout: 250 seconds)
11:30:22FromGitter<zacharycarter> that hackaday article is on my google news feed :O
11:32:17FromGitter<narimiran> @Vindaar thumbs up! :)
11:33:22PMunchSpeaking of articles and such, did anyone ever do something with the FOSDEM lightning talk idea?
11:36:01*exos27 joined #nim
11:37:30*exos27 quit (Remote host closed the connection)
11:38:25*kungtotte quit (Remote host closed the connection)
11:39:34*leorize joined #nim
11:41:30*dlech1 joined #nim
11:43:05*dlech1 quit (Remote host closed the connection)
11:58:46*ZeroFlame00719 joined #nim
12:01:06*EmmyNoether24 joined #nim
12:03:03*EmmyNoether24 quit (Read error: Connection reset by peer)
12:03:29*ZeroFlame00719 quit (Remote host closed the connection)
12:06:50*Krenair12 joined #nim
12:07:19*Krenair12 quit (Remote host closed the connection)
12:08:34*xet7 joined #nim
12:11:05federico3PMunch: I started a deck of slides that can be used for any talk https://github.com/FedericoCeratto/nim-slides
12:11:42*tzui quit (Ping timeout: 244 seconds)
12:15:21PMunchOh cool
12:16:25PMunchA bit long for a lightning talk though
12:17:13PMunchI was thinking of writing a set on Nim macros, and the power they bring to the language
12:17:24PMunchI think that could be a cool 5-minute talk
12:18:05*SenasOzys quit (Ping timeout: 250 seconds)
12:22:34*wmann joined #nim
12:24:41FromGitter<narimiran> federico3: that is a nice template for some general 'intro to nim' presentations!
12:25:00*wmann quit (Remote host closed the connection)
12:25:01FromGitter<narimiran> PMunch: is it possible to show (off) macros in 5 minutes?
12:25:54PMunchWell, I think you could talk a little bit about what they can be used for, then maybe show an example or two
12:26:00federico3thanks - contributions are welcome. There could be longer and shorter decks on the same repo
12:27:12federico3PMunch: lightning talks at FOSDEM are short. Getting into macros would need a regular talk
12:28:37PMunchWell they're not that short, I'm not talking about explaining how they work or anything. Just give people a teaser on their power
12:32:15*flacko18 joined #nim
12:33:47*riidom quit (Ping timeout: 240 seconds)
12:35:32*flacko18 quit (Remote host closed the connection)
12:35:50*swatarianess23 joined #nim
12:36:07*leru quit (Ping timeout: 240 seconds)
12:38:29*swatarianess23 quit (Remote host closed the connection)
12:39:08federico3PMunch: keep in mind that most people in the crowd know nothing about Nim
12:41:09FromGitter<deech> Does the asyncdispatch mechanism launch a separate OS thread for each process?
12:43:04PMunchfederico3, yeah I know. The idea wasn't to teach them much about Nim either, just to leave them with a desire to do so .)
12:43:37PMunchNo one is going to check out a language if you just pitch the basic facts about it
12:45:44*egon102418 joined #nim
12:46:00*egon102418 quit (Remote host closed the connection)
12:47:32*Loshki5 joined #nim
12:48:11FromDiscord<Shield> @deech no, it just automatically converts your procedures into iterator closures, everything runs on single thread
12:48:51FromGitter<narimiran> PMunch: i wouldn't check out a language if you just pitch "macros are cool" ;)
12:50:07*Loshki5 quit (Remote host closed the connection)
12:51:26*dddddd joined #nim
12:51:52federico3same here
12:52:28PMunchnarimiran, well it wouldn't be "macros are cool" but rather "this is cool, and made possible by macros"
12:53:16*klntsky26 joined #nim
12:55:17*Dmitrii-Sh4 joined #nim
12:57:40*Dmitrii-Sh4 quit (Remote host closed the connection)
12:58:03halircHello. I'm using the & operator from strformat like this &"{parseFloat(stringVarWithFloat):.4f}" and I get the result I'm looking for most of the time but sometimes I see resulting strings like this -5.722e-006. How can I prevent that? Am I doing something wrong?
13:01:42halircGenerally my question would be: If I have a number with decimal places, read in as a string, how can I round it to 4 decimal places consistently and have the result be a string again.
13:01:43*tay_ joined #nim
13:01:44*klntsky26 quit (Ping timeout: 252 seconds)
13:01:47*Astro746724 joined #nim
13:02:44halircIf I just use round I saw things like 8.000000000001
13:02:52PMunchfloatString[0..floatString.find('.')+4]
13:02:58PMunchShould work I think
13:03:16*mikky6 joined #nim
13:03:37PMunch!eval let floatString = "123.45612121212"; echo floatString[0..floatString.find('.')+4]
13:03:40NimBot123.4561
13:04:14PMunchIt requires the float string to be formatted as number.number though
13:04:19PMunchNo e notation
13:04:32FromGitter<Vindaar> @PMunch but this would just truncate the decimal places, no?
13:04:42PMunchVindaar, ah good point
13:05:03PMunchIt would indeed just take the first n characters of the decimal, no rounding
13:05:43*Astro746724 quit (Remote host closed the connection)
13:07:52*dexta15 joined #nim
13:07:56halircI will always be in the number.number format but yes, it has to be rounded, too.
13:08:00FromGitter<Vindaar> Hm, so according to https://nim-lang.org/docs/strformat.html#standard-format-specifier-for-strings-integers-and-floats I understand it such that it should always print it as a fixed point number. `g` is the notation which automatically switches to exponent notation
13:08:01FromGitter<narimiran> halirc: the problem with `-5.722e-006` when you try to make it a string with four decimal places is that that number is much smaller than 0.0001
13:08:06*mikky6 quit (Read error: Connection reset by peer)
13:08:36FromGitter<narimiran> this is 0.0000057. how do you print that with only 4 decimal places?
13:09:26FromGitter<Vindaar> well, it should just print 0 or round up the last digit to 1?
13:09:30halircIt's fine if it would get rounded to 0, I just don't want any e notation in my output string
13:09:40halircyes, 0 or 0.0001 all ok
13:09:47*dexta15 quit (Read error: Connection reset by peer)
13:11:02FromGitter<Vindaar> weird, @halirc it works fine for me
13:11:42FromGitter<Vindaar> !eval import strformat; let a = 0.00000000057342; echo &"{a:4f}"
13:11:46NimBot0.000000
13:12:01*JinnX27 joined #nim
13:12:33PMunch!eval import strformat; let a = "0.00000000057342"; echo &"{parseFloat(a):4f}"
13:12:35NimBotCompile failed: lib/core/macros.nim(377, 1) Error: undeclared identifier: 'parseFloat'
13:13:00PMunch!eval import strformat, strutils; let a = "0.00000000057342"; echo &"{parseFloat(a):4f}"
13:13:04NimBot0.000000
13:13:06FromGitter<narimiran> i really wouldn't like for something two orders of a magnitude smaller to be rounded/shown as 0.0001 ;)
13:13:15FromGitter<Vindaar> haha, agreed
13:13:33PMunch!eval import strformat, strutils; let a = "0.0000057"; echo &"{parseFloat(a):4f}"
13:13:37FromGitter<Vindaar> shouldn't it only echo 4 places though?
13:13:37NimBot0.000006
13:13:59PMunchYeah that's strange
13:14:07FromGitter<Vindaar> oh, need to do:
13:14:08PMunch!eval import strformat, strutils; let a = "12,34567"; echo &"{parseFloat(a):4f}"
13:14:08FromGitter<narimiran> Pmunch: use `:.4f`
13:14:11NimBot<no output>
13:14:13FromGitter<Vindaar> yep
13:14:15halirc-5.722e-006 is the putput I get from &"{float:4f}", I don't actually know at this time what the input is, it would be a bit annoying to dig that out, but I could
13:14:16PMunch!eval import strformat, strutils; let a = "12.34567"; echo &"{parseFloat(a):.4f}"
13:14:20NimBot12.3457
13:14:38PMunch!eval import strformat, strutils; let a = "0.0000057"; echo &"{parseFloat(a):.4f}"
13:14:41NimBot0.0000
13:14:46halircI just was surpised it made any e notation output
13:14:54FromGitter<Vindaar> rightly so
13:15:05*JinnX27 quit (Remote host closed the connection)
13:15:12*tay_ quit (Remote host closed the connection)
13:15:12PMunchhalirc, which Nim version are you using?
13:15:35halircNim Compiler Version 0.18.0 [Windows: amd64]
13:15:38*tay_ joined #nim
13:17:39*mpan19 joined #nim
13:17:43FromGitter<narimiran> i've tried and in both v0.18 and v0.18.1 i get 0.0000
13:17:50*mpan19 quit (Read error: Connection reset by peer)
13:18:06FromGitter<Bennyelg> @Vindaar https://gist.github.com/Vindaar/aab2926425400fc57274b521e80398dd Fantastic boost.
13:18:12halircI'll find out the input that generates this
13:18:17PMunchIt's probably only some floating point numbers that cause this
13:19:20FromGitter<narimiran> https://forum.nim-lang.org/t/4243 who left?
13:21:08*kapil___ quit (Quit: Connection closed for inactivity)
13:21:09AraqI don't know :-)
13:22:26FromGitter<Vindaar> @Bennyelg I agree, it's pretty nice. :) Although I guess for something like this other Python tools probably achieve similar speed ups (so it won't be able to convince a python user to try Nim by itself)
13:23:41FromGitter<Bennyelg> Yea
13:23:41pqflx3[m]This? https://irclogs.nim-lang.org/24-09-2018.html#02:35:49
13:23:45FromGitter<Vindaar> Back when I did this I actually wrote the integrand in C and called it with ctypes. That was quite a little more cumbersome though
13:24:31PMunchhttp://ix.io/1nBg/Nim gives no output
13:24:57PMunchSo I'm not able to repeat the output you get halirc
13:25:07FromGitter<narimiran> @Vindaar @Bennyelg well, you can use cython. which is sometimes quite simple with huge benefits. but sometimes it isn't. ;) btw, here is what i have written on the same topic: https://narimiran.github.io/2018/05/10/python-numpy-nim.html
13:25:44FromGitter<narimiran> numba is much simpler than cython. and *when* it works, it is very nice
13:25:59*YoY joined #nim
13:26:40*MrAxilus[m] joined #nim
13:28:03FromGitter<Vindaar> Yep, I remember reading that post. I tried cython in the past (related to this actually) but wasn't a fan. And we use numba currently in a readout software. It does work, but to me it feels weird.
13:29:09FromGitter<Vindaar> It's probably different if you actually spend time to learn numba well, but to me it feels like "let's @jit it and hope this is faster than before". I like certainty :)
13:29:12*YoY quit (Remote host closed the connection)
13:29:27FromGitter<Vindaar> but well, I guess everyone /here/ will agree with me anyways :D
13:30:22FromGitter<narimiran> yeah, that's how i use numba too :D
13:34:42halircI turns out my input is not always in the number.number format, as I really thought it was. So the regex to find these numbers did not catch that one with the e in it and it was never processed by &. So it is all good with the & .4f formatting.
13:34:56halircSorry about wasting your time with this and the false alarm.
13:35:28halircAt least it got me to find the error. That's what I get when I assume the input is all correctly or consistently formatted
13:37:28FromGitter<Vindaar> glad you found the issue :)
13:39:55*Gohla5 joined #nim
13:40:55FromGitter<kaushalmodi> @narimiran The docopt package dev has left as I learned from @zacharycarter and @dom96
13:41:50*zampano21 joined #nim
13:41:54Araqthat was years ago. He is kind enough to keep maintaining his packages though, afaict
13:43:37FromGitter<kaushalmodi> I think that's it then I believe. You can mention (will be more official) in that forum thread to prevent any panic :)
13:44:03*zampano21 quit (Remote host closed the connection)
13:44:13AraqI replied
13:44:34*Gohla5 quit (Remote host closed the connection)
13:47:47*junland quit (Ping timeout: 240 seconds)
13:48:58Araqhttps://docs.python.org/3/c-api/memory.html
13:49:17FromDiscord<Shield> whenever I ask people about Nim they always complain about the lack/outdated libraries and bindings, the last one complained about the GTK3 bindings and the lack of destructors
13:50:06Araqthere is always a good reason for not using something
13:50:33AraqGTK3 is written in C. C lacks destructors.
13:52:49*madnight joined #nim
13:52:55FromDiscord<Shield> I too have a problem with it, the SFML bindings are really outdated, but I just went with it because Nim is too good, at least openGl and SDL are covered for gamedev
13:53:42*madnight quit (Remote host closed the connection)
13:54:10Araqthe destructors are getting into shape, but most languages live fine without them
13:55:14FromDiscord<Shield> yeah I don't get the deal behind destructors, you can invoke the free/destroy function on your own
13:55:43Araqoften it's also pushed into a finalizer and you don't have to call free
13:55:50FromGitter<arnetheduck> about forum thread - indeed, outdated libraries is a symptom of people leaving nim, and that was what the conversation was about
13:56:45Araqoutdated libraries are everywhere though. there is a constant stream of newly developed libs too
13:57:26FromGitter<Bennyelg> people always seeking how and where to complain.
13:57:43FromGitter<Bennyelg> I don't take them seriously if they leaving because library is outdated.
13:57:48FromGitter<timotheecour> araq since you’re here, I’m really curious about how to integrate existing react components in karax? (and vice versa) - Nim forum (https://forum.nim-lang.org/t/4241)
13:58:10Araqin open source development, maintenance almost always leaves a lot to be desired
13:58:51Araqand of course, Nim has problems with stability. the macro system is Nim's key unique feature
13:59:16Araqso people are happy to explore the macro system, only to notice that the next version changed the AST slightly and their macros break
13:59:40FromGitter<arnetheduck> @Bennyelg the other way around - I'm not curious about people complaining that libraries are outdated - I'm curious about why people make the effort to write a library, publish it, and then stop using nim (ie stop maintaining the library - if the authers were using it / using nim, it would get maintenance)
13:59:54FromDiscord<Shield> This is also a confusing time, some libraries already require 0.18.1 (found one with 0.19 for some reason) when 0.18.0 is still the official version, so the libraries are split even more
14:00:08Araq(even though the AST is part of the spec...)
14:00:36*Perkol joined #nim
14:01:14Araqtimotheecour: I don't know aobut react components.
14:01:20FromGitter<Bennyelg> @arnetheduck Agree but I can give you a bunch of libraries in python which not updated anymore. ⏎ + ⏎ remember that nim is not yet version 1.* so everything is under work.
14:01:52FromDiscord<Shield> @arnetheduck, exactly, nim's bindings for SFML are outdated while the author kept the crystal bindings updated, it's kinda unfair
14:02:56FromGitter<timotheecour> does that mean we can’t reuse any of the components (eg youtube, google maps, tree nav etc) written for react in karax? or there could be a way but it’s not trivial?
14:03:23*KAT joined #nim
14:03:47*Qiming10 joined #nim
14:03:50Araqthere is experiments/echartstest.nim
14:04:42Araqyou use setForeignNodeId "echartSection" to make karax aware of a subsection of the DOM not under its control
14:04:43*Qiming10 quit (Remote host closed the connection)
14:05:43*KAT quit (Remote host closed the connection)
14:06:04AraqI also wanted to add 'nativeNode' virtual DOM node kind for even better interop but never got to do it. PRs welcome, it's not hard.
14:06:41Araqthe idea is that the virtual DOM -> DOM transformation keeps 'nativeNodes' as they are
14:07:12FromGitter<timotheecour> cool, i was really afraid of a completely separate ecosystem, ok will look at it
14:09:39FromGitter<timotheecour> for some reason `nim js echartstest.nim && open echartstest.html` doesn’t do much (clicking on Products doesn’t do anything) but maybe i’m doing smthg wrong (other simple karax stuff works)
14:10:29FromDiscord<hotdog> Does anyone have an example of "--hotCodeReloading:on" working with browserSync/Livereload?
14:14:56*digitalgrease3 joined #nim
14:15:26*tay_ quit (Ping timeout: 252 seconds)
14:19:21*tay_ joined #nim
14:19:33*digitalgrease3 quit (Remote host closed the connection)
14:22:56*mooasaurus7 joined #nim
14:22:56*mooasaurus7 quit (Remote host closed the connection)
14:25:00*PMunch quit (Quit: Leaving)
14:27:38*enthus1ast joined #nim
14:28:31enthus1astfolks have a look at this tui library: https://github.com/nais314/stui/
14:28:52enthus1asthe has just fixed id for devel
14:28:54enthus1astit
14:29:12Araq that's what I've been waiting for
14:29:13enthus1astbuild all the ssh guis
14:29:15enthus1ast:D
14:29:26enthus1astme to
14:32:27FromGitter<kaushalmodi> enthus1ast: woh! good find!
14:33:06*Samian9 joined #nim
14:35:26*Jbain_ joined #nim
14:36:12*Perkol quit (Quit: Leaving)
14:37:59*Samian9 quit (Remote host closed the connection)
14:38:37*lukexj21 joined #nim
14:38:55*Jbain_ quit (Remote host closed the connection)
14:38:58*ephemer0l23 joined #nim
14:40:15*lukexj21 quit (K-Lined)
14:40:34*Perkol joined #nim
14:41:21federico3shameless plug https://github.com/FedericoCeratto/nim-dashing
14:41:52*abm joined #nim
14:42:24*abm left #nim (#nim)
14:42:42*ephemer0l23 quit (Remote host closed the connection)
14:46:36*mr_noob1 joined #nim
14:47:41*Guest41066 joined #nim
14:48:06*Guest41066 quit (Remote host closed the connection)
14:54:15FromGitter<gogolxdong> @nais314
14:54:34FromGitter<gogolxdong> invisible
14:56:33*BadHorsie joined #nim
14:58:55*BadHorsie quit (Read error: Connection reset by peer)
15:00:33FromGitter<gogolxdong> Error: 'spawn' takes a GC safe call expression
15:01:32*dietricha_ joined #nim
15:02:42FromGitter<gogolxdong> on branch nim019 `Error: 'tb2test' is not GC-safe as it calls 'setCursorPos'`
15:03:06*dietricha_ quit (Remote host closed the connection)
15:05:02FromGitter<tim-st> why is `newSeqUninitialized` only available for number types?
15:05:31FromGitter<tim-st> I tried for enum and didnt work
15:05:58Araqgogolxdong: you can enforce gc-safety via
15:05:59FromDiscord<Shield> @federico3 windows did not like that demo
15:06:06Araq{.gcsafe.}:
15:06:12Araq echo "my dangerous code here"
15:06:53*amiconn13 joined #nim
15:07:47FromGitter<tim-st> ok, I have a better idea, I cast the string direct instead of using `newSeqUninitialized`
15:07:53enthus1astgogolxdong the stui_test1.nims build the test for me
15:09:14*television18 joined #nim
15:09:20*amiconn13 quit (Read error: Connection reset by peer)
15:10:49FromGitter<gogolxdong> It builds for me too on dev enviroment, but not on WLS. many frontend devs uses windows, sigh.
15:10:54*tay_ quit (Ping timeout: 252 seconds)
15:11:22FromGitter<gogolxdong> use
15:11:43*television18 quit (Remote host closed the connection)
15:11:59*user_5117 joined #nim
15:12:54*dau13 joined #nim
15:13:25federico3Shield: never tested on windows. I'll try to enable appveyor. PRs are welcome
15:14:04FromGitter<gogolxdong> a lame linux system.
15:14:11*user_5117 quit (Remote host closed the connection)
15:14:41*notbrent19 joined #nim
15:14:59*dau13 quit (Remote host closed the connection)
15:15:30FromDiscord<Shield> same problem with stui library, window's terminal just looses its mind, I guess it just sucks compared to linux's
15:15:53*notbrent19 quit (Remote host closed the connection)
15:17:26FromGitter<gogolxdong> I would like to force them to use ubuntu18.04 afterwards for enrolled devs.
15:17:52enthus1asti think terminals suck in general
15:19:27*Trustable joined #nim
15:22:32enthus1astFedericoCeratto dashing.nim(76, 12) Error: usage of 'isNil' is a user-defined error
15:23:12enthus1ast^ federico3 ?
15:23:22FromGitter<kaushalmodi> Araq: Would you plan to first publish an RC1 for 0.19.0?
15:23:32federico3enthus1ast: Nim version?
15:24:16FromGitter<kaushalmodi> That way all the package developers who don't use Nim devel at least fix backward incompatible changes like isNil before a larger user crowd is faced with those broken packages
15:24:31FromGitter<kaushalmodi> federico3: devel
15:25:07FromGitter<kaushalmodi> Araq: RC1 -> release candidate 1
15:25:50Araqno, this is getting ridiculous
15:26:07Araq"release candidates" for some odd version that should have been v1 5 years ago
15:26:29AraqI'm willing to publish 0.19.2 soon after though
15:26:45enthus1astfederico3, just pulled in the freshest
15:26:49Araqbut I've ported plenty of my own code to 0.19 without much hickups
15:26:59*gweedo7676 joined #nim
15:27:20FromGitter<kaushalmodi> yeah, the only concern is that many people still don't know of the breaking nil string change
15:28:02Araqthere is --nilseqs:on
15:28:15FromGitter<kaushalmodi> yeah
15:28:17Araqand it works
15:28:21enthus1astwhere can i read what 0.19 brings?
15:28:22*gweedo7676 quit (Read error: Connection reset by peer)
15:28:25*ritchie_ joined #nim
15:28:33Araqthere is also --laxStrings:on and it works
15:29:13FromGitter<kaushalmodi> The problem is that people don't know that they should use that switch
15:29:17FromGitter<kaushalmodi> related issue: https://github.com/nim-lang/Nim/issues/9074
15:29:37Araqthey do know, it's in the changelog
15:29:53*Jesin quit (Quit: Leaving)
15:29:55enthus1asthttps://github.com/nim-lang/Nim/blob/devel/changelog.md
15:29:59FromGitter<kaushalmodi> should the deprecated message be more advertised? and also the error message suggest using that switch temporarily till the package gets fixed?
15:30:39*mr_noob1 quit (Quit: Page closed)
15:31:29FromDiscord<Shield> i'm also curious about the changelog
15:32:12FromGitter<kaushalmodi> Shield: See the above link that enthus1ast posted
15:33:30AraqI don't know how that "deprecated message be more advertised"?
15:33:37Araqwhat do you mean?
15:33:56FromGitter<kaushalmodi> Araq: See that issue
15:34:35FromGitter<kaushalmodi> isNil is missing a deprecated message in bold. It should have looked something like this: https://nim-lang.github.io/Nim/system.html#xlen,string
15:34:35federico3perhaps more descriptive: Error: isNil is deprecated. Replace it with "== nil"
15:34:54FromGitter<kaushalmodi> federico3: that wouldn't work either
15:34:56FromGitter<kaushalmodi> strings cannot be nil
15:35:14FromGitter<kaushalmodi> what it should be replaced with depends on the context
15:35:26*ft27 joined #nim
15:35:30FromGitter<kaushalmodi> it could be replaced with `""`, or may be even `options[string]`
15:35:37federico3ah! One reason more to give context-dependent error messages
15:35:56FromGitter<kaushalmodi> the context can be known only by humans
15:36:06AraqisNil is not deprecated, it's not allowed anymore
15:36:21FromGitter<kaushalmodi> Araq: I meant to give that as an example
15:36:33FromGitter<kaushalmodi> point is that the docs don't highlight that error at all!
15:36:38FromGitter<kaushalmodi> see the screenshot in that issue
15:36:53FromGitter<kaushalmodi> if someone sees that error on the terminal, and visits the docs, they will be confused
15:37:11FromGitter<kaushalmodi> they wouldn't know to fish for the "error" pragma *after* clicking on the ellipsis
15:37:20enthus1astafaik isNil does not echo the line number where the nil check is
15:37:26*Torro joined #nim
15:37:32enthus1astafaik it points to system.nim
15:37:48enthus1astah
15:37:50enthus1astnevermind
15:38:03Araqkaushalmodi: it's a good point but I don't know what to do.
15:38:03FromGitter<kaushalmodi> the `isNil` on devel looks like this:
15:38:16FromGitter<kaushalmodi> https://user-images.githubusercontent.com/3578197/46038011-f3778c80-c0d7-11e8-94bf-2d87bc9edb48.png
15:38:44AraqI would change the docgen and re-upload the docs but then *not* the release was used to build the "Latest stable" docs
15:39:57*ft27 quit (Remote host closed the connection)
15:39:59FromGitter<kaushalmodi> Araq: I am not understanding the complication.. would it not be possible to have error pragma print an error in bold in the docs, as the deprecated pragma is printing this (example): "Deprecated since version 0.18.1. Use len() instead. "?
15:40:18planetis[m]hey, should the `name=` proc in macros override the export marker?
15:40:41Araqplanetis[m]: ha, good question. probably not
15:41:31Araqkaushalmodi, I've already tagged v0.19, that's the problem
15:42:24*ritchie_ quit (Quit: Leaving)
15:42:50FromGitter<kaushalmodi> ah ok, saw that you did that 8 mins ago :P
15:43:10FromGitter<kaushalmodi> may be the docs get fixed in a quick followup 0.19.1 release?
15:43:16Araqit's always the same, polish, polish, polish required for everything and in the meantime people wonder why the fuck 'destructors' are STILL not ready
15:44:41AraqI would use 0.19.1 to rebuild the docs for 0.19.0 but people hate when I do that
15:45:00FromGitter<kaushalmodi> why would people hate?
15:45:21Araqhttps://github.com/nim-lang/Nim/issues/9063
15:45:25FromGitter<kaushalmodi> as docgen is tied with Nim, it makes sense to tag new release if you fix the docgen
15:46:09Araqonce 'nightlies' work I can release much more often but travis messages like "nothing new, skipping deploy step" make it completely untrustworthy
15:46:43Araqnot to mention that it doesn't understand git tags and appveyor only understands git tags or something like that
15:47:11FromGitter<kaushalmodi> huh? My experiment with nightly docs worked fine
15:47:24Araqdocs are not binaries
15:47:32FromGitter<kaushalmodi> also tag based deployment works fine using Travis..
15:47:36FromGitter<kaushalmodi> that's with binaries
15:48:05FromGitter<kaushalmodi> See https://github.com/kaushalmodi/hello_musl/blob/master/.travis.yml, https://github.com/kaushalmodi/hello_musl/releases
15:48:09FromGitter<kaushalmodi> I cannot comment on Appveyor
15:48:25Araqwell I have a special repo for nightlies, it's not part of the Nim core repo
15:48:35Araqbecause that caused the builds to time-out
15:48:57FromGitter<kaushalmodi> wouldn't forced updates work?
15:49:01Araqand it's not required for every commit anyway, "nightly" would be fine
15:49:45Araqprobably something will eventually work but I'm really bad at these things and my time is much better spent elsewhere
15:49:57Araqso please be my guest, look at nightlies and fix it
15:50:05Araqyou can fork it and play with it
15:50:12FromGitter<kaushalmodi> makes sense, will have a look later
15:50:41*acute11 joined #nim
15:50:44AraqI think my scripts are complete to 90%
15:54:27*acute11 quit (Remote host closed the connection)
16:00:53*skellock quit (Ping timeout: 245 seconds)
16:05:22*darithorn joined #nim
16:08:19*Jesin joined #nim
16:09:29*leru joined #nim
16:09:49*nsf quit (Quit: WeeChat 2.2)
16:12:22*skellock joined #nim
16:15:56*borbyu joined #nim
16:15:59*Guest31642 joined #nim
16:16:13*DrMcCoy15 joined #nim
16:16:54*Vigdis2 joined #nim
16:18:11*borbyu quit (Read error: Connection reset by peer)
16:18:27*Jesin quit (Quit: Leaving)
16:18:54*Vigdis2 quit (Remote host closed the connection)
16:19:17*halirc quit (Quit: Leaving)
16:19:24*Jesin joined #nim
16:19:48*Guest31642 quit (Remote host closed the connection)
16:21:23*DrMcCoy15 quit (Ping timeout: 252 seconds)
16:24:41*Sembei quit (Ping timeout: 252 seconds)
16:24:44*MyMind joined #nim
16:26:49*jemarch10 joined #nim
16:28:06*strong_ joined #nim
16:28:13*chris|16 joined #nim
16:28:41*strong_ quit (Remote host closed the connection)
16:29:56*noonien quit (Quit: Connection closed for inactivity)
16:30:25*jemarch10 quit (Remote host closed the connection)
16:31:15*chris|16 quit (Remote host closed the connection)
16:32:34*Guest85318 joined #nim
16:34:03*Guest85318 quit (Remote host closed the connection)
16:35:42*hfp_work joined #nim
16:42:04*hfp_work quit (Remote host closed the connection)
16:47:38FromGitter<alehander42> Araq, is something like typeSectionMacro possible? e.g. to have a way to define a compile time macro which generates code to be included in a type section and after the end of the section
16:47:57FromGitter<alehander42> (possible to add eventually to the compiler from a design viewpoint)
16:48:11FromGitter<alehander42> e.g. you invoke your macro with
16:49:19FromGitter<alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babb88feba8e60bc64ba9dc]
16:49:46shashlick_need some c2nim help
16:49:47shashlick_https://github.com/facebookresearch/fastText/blob/a5d22aba45f38c12d195ecc6c3e448aa3690fbbd/src/fasttext.h#L46
16:49:59shashlick_what do I change that line to that c2nim will like?
16:50:15shashlick_Error: identifier expected, but got: {
16:53:39*avium1 joined #nim
16:53:39*lari joined #nim
16:55:49*lari quit (Remote host closed the connection)
16:56:02*zveda17 joined #nim
16:56:24*zveda17 quit (Remote host closed the connection)
16:57:09*avium1 quit (Remote host closed the connection)
16:57:31FromGitter<mratsim> isn’t that something that is changed by the preprocessor?
16:59:19shashlick_well, i'm trying to wrap it so trying to convert to equivalent nim
16:59:38shashlick_for now, i'm changing `tokenCount_{}` to `tokenCount_` but i'll bet it doesn't mean the same thing
17:07:02*fatalbit5 joined #nim
17:10:19*TotallyNotKim17 joined #nim
17:10:34*Trustable quit (Remote host closed the connection)
17:11:25*fatalbit5 quit (Remote host closed the connection)
17:11:36FromGitter<bung87> can’t find the document about the suffix `{}`
17:12:02*TotallyNotKim17 quit (K-Lined)
17:12:08*noonien joined #nim
17:13:54FromGitter<bung87> seems like a initial
17:15:46FromGitter<bung87> http://aristeia.com/EC++11-14/parens%20or%20braces%202014-03-18.pdf
17:18:16shashlick_bung87: making decent progress
17:19:06*Richlv20 joined #nim
17:19:12FromGitter<bung87> are you the guy at nimgen channel ?
17:19:27shashlick_yep
17:19:36*Ysblokje joined #nim
17:20:29FromGitter<bung87> ah , sounds good, I thought that might be difficult if it’s relys on much libs.
17:21:25*Richlv20 quit (Remote host closed the connection)
17:21:52*grepper3 joined #nim
17:23:59*Ysblokje quit (Read error: Connection reset by peer)
17:24:22*Serus joined #nim
17:24:50FromGitter<bung87> I’d like port some others libs depends on fastext, since I think it gets good quality for some task.
17:25:19*Serus quit (K-Lined)
17:25:55*grepper3 quit (Remote host closed the connection)
17:27:07*lumidify joined #nim
17:27:15*lumidify quit (Remote host closed the connection)
17:29:19*esainane4 joined #nim
17:29:38shashlick_so far so good
17:29:46shashlick_but i've had limited luck wrapping c++ libs
17:30:55*esainane4 quit (Remote host closed the connection)
17:32:44FromGitter<bung87> I’m waiting here if you need someone to talk with, I just got a little QT framework experience in c++.
17:32:49*tagomago2 joined #nim
17:35:31*enthus1ast quit (Ping timeout: 260 seconds)
17:37:31*tagomago2 quit (Remote host closed the connection)
17:38:36*Jesin quit (Quit: Leaving)
17:40:03*Guest46751 joined #nim
17:41:23*Guest46751 quit (Remote host closed the connection)
17:42:16shashlick_Mostly minor stuff getting imports right, etc
17:42:28shashlick_Main challenge is to get an example to compile
17:43:12*Jesin joined #nim
17:43:12shashlick_Considering you've used this lib before, you might have better luck handling that
17:44:16shashlick_Like I got nuklear wrapped and compile cleanly but AlexMax had multiple changes to actually get it working in an example
17:46:15*Newa10 joined #nim
17:46:17*Torro quit (Quit: -bye)
17:48:04FromGitter<bung87> yeah, I have a repo use its language detetion feature, https://github.com/bung87/whatlangid I can use the language model test it.
17:48:41planetis[m]boom its done: https://github.com/b3liever/protocoled opinions?
17:48:46*Newa10 quit (Read error: Connection reset by peer)
17:50:55*feigned4 joined #nim
17:53:22*Guest16583 joined #nim
17:53:58*feigned4 quit (Remote host closed the connection)
17:55:05FromGitter<bung87> it’s like AClass implements Bclass ?
17:56:10FromGitter<bung87> like in java or typescript
17:57:17*Guest16583 quit (Remote host closed the connection)
18:00:00*rockcavera quit (Remote host closed the connection)
18:01:25*xet7 quit (Quit: Leaving)
18:01:46*xet7 joined #nim
18:02:49*xet7 quit (Remote host closed the connection)
18:02:53*annalee_ joined #nim
18:03:16*annalee_ quit (Remote host closed the connection)
18:03:23*xet7 joined #nim
18:05:03*tay_ joined #nim
18:09:16planetis[m]well and that, read the readme on how it works
18:14:54*rockcavera joined #nim
18:16:04*skellock quit (Ping timeout: 240 seconds)
18:17:34FromGitter<bung87> good night,it’s 2:17 am now.
18:19:59planetis[m]night night!
18:25:04planetis[m]if i have time i will do another one that generates code like `interfaced`
18:28:59*tay_ quit (Quit: Leaving)
18:35:05*lvgx27 joined #nim
18:36:31*Jesin quit (Quit: Leaving)
18:37:39*lvgx27 quit (Read error: Connection reset by peer)
18:39:37*witti5 joined #nim
18:39:53*nsf joined #nim
18:40:27*Jesin joined #nim
18:41:37Araqshashlick_: just remove the {} ?
18:43:00shashlick_araq: I did that, converts fine now but no idea if it will work but will deal with that later
18:43:10shashlick_more immediate problem is https://en.cppreference.com/w/cpp/utility/pair
18:43:48shashlick_```
18:43:58shashlick_```
18:43:59shashlick_proc predict*(this: var Model; a3: vect[int32]; a4: int32; a5: real;
18:44:00shashlick_ a6: var vect[pair[real, int32]]) {.stdcall, importcpp: "predict",
18:44:02shashlick_ header: headermodel.}
18:44:03shashlick_```
18:44:15*witti5 quit (Remote host closed the connection)
18:44:23shashlick_that `pair[real, int32]` needs to be declared
18:44:35shashlick_i have ` pair {.importc: "std::pair", header: "utility".}[T, U] = object`
18:45:12shashlick_but it doesn't work, I get:
18:45:13shashlick_```Error: cannot instantiate pair
18:45:15shashlick_got: <void, type int32>
18:45:16shashlick_but expected: <T, U>
18:45:17shashlick_```
18:48:21*japanoise3 joined #nim
18:50:05Araq'real'? it's called 'float'
18:51:14shashlick_there's a typedef - https://github.com/facebookresearch/fastText/blob/master/src/real.h
18:52:35shashlick_it gets converted to `type real* = cfloat`
18:54:22shashlick_but nim doesn't like it, if I change real to float, it works
18:55:13Araqok...
18:56:39shashlick_interesting - if the type is imported from another file (real.nim), it doesn't work, but if I just add the type to the same file, it works
18:57:34*japanoise3 quit (Ping timeout: 272 seconds)
19:02:32*kiltzman13 joined #nim
19:03:24*Mangy_Dog joined #nim
19:03:25*endre27 joined #nim
19:03:52*xace_ quit (Ping timeout: 246 seconds)
19:04:16*endre27 quit (Remote host closed the connection)
19:05:00*xace joined #nim
19:07:04*Mangy_Dog quit (Remote host closed the connection)
19:07:17*kiltzman13 quit (Remote host closed the connection)
19:11:11*ritchie_ joined #nim
19:14:02*Comnenus1 joined #nim
19:14:47shashlick_@bung87: I think it is working now, will post on github for you to take it further
19:15:11*Comnenus1 quit (Remote host closed the connection)
19:17:38FromGitter<bung87> wow! thanks! I will check it at day time.
19:17:57*nullptr|24 joined #nim
19:19:07*nullptr|24 quit (Remote host closed the connection)
19:19:22*ktch_ joined #nim
19:20:31*leru quit (Ping timeout: 260 seconds)
19:20:34*SkillyOnFire5 joined #nim
19:21:27*SkillyOnFire5 quit (Remote host closed the connection)
19:21:38*the_real_plumps2 joined #nim
19:22:40*rattboi13 joined #nim
19:22:41*Cthulhon26 joined #nim
19:24:12*ktch_ quit (Remote host closed the connection)
19:25:22*Cthulhon26 quit (Read error: Connection reset by peer)
19:25:45FromGitter<bung87> it will post here right ? https://github.com/genotrance/nimfastText
19:26:13*rattboi13 quit (Remote host closed the connection)
19:26:20*skellock joined #nim
19:26:24*the_real_plumps2 quit (Remote host closed the connection)
19:26:25shashlick_yep - just pushed
19:27:53FromGitter<bung87> ok I think I’ll give a simple test for now, since I’m not sleep well.
19:28:17shashlick_only tested on Windows, will check on linux right now
19:32:49*chowmeined29 joined #nim
19:33:02*unixman18 joined #nim
19:35:51*zyk14 joined #nim
19:36:23*chowmeined29 quit (Remote host closed the connection)
19:38:34*zyk14 quit (Remote host closed the connection)
19:39:01shashlick_pushed some fixes for Linux
19:40:31*unixman18 quit (Ping timeout: 250 seconds)
19:41:10*francisl joined #nim
19:42:17*KBme7 joined #nim
19:43:39*KBme7 quit (Remote host closed the connection)
19:46:06*Pisuke joined #nim
19:47:27*MyMind quit (Ping timeout: 240 seconds)
19:49:06*ssbarnea3 joined #nim
19:50:20planetis[m]what do you think of my macro by the way?
19:50:37*Guest51823 joined #nim
19:51:14FromGitter<kaushalmodi> Araq: No one asked this :) Why isn't Nim 0.19.0 announced yet? (release notes, blog post, etc.) Are you planning to do that tomorrow?
19:52:02*ssbarnea3 quit (Read error: Connection reset by peer)
19:52:18FromGitter<bung87> `../../../.nimble/pkgs/nimfastText-0.1.0/nimfastText/dictionary.nim(61, 56) Error: invalid token: _ (\95)`
19:52:41FromGitter<bung87> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babe389eba8e60bc64cd684]
19:52:52*Guest51823 quit (Read error: Connection reset by peer)
19:53:41shashlick_which os
19:58:25shashlick_how is `nullptr` changing into `__get_nil_t()` for you
19:58:34*skellock quit (Ping timeout: 240 seconds)
19:59:44FromDiscord<Shield> I thought you'll release 0.18.1
20:00:09FromGitter<bung87> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babe549cdc501317259f280]
20:00:29planetis[m]comeon congratulate me already, it is pretty cool ;)
20:01:59FromGitter<kaushalmodi> @bung87 You need to update devel :)
20:03:28shashlick_@bung87: you're on Mac? don't have a way to test that
20:03:46FromGitter<bung87> is that true? just two days ago version.
20:03:48*CC6610 joined #nim
20:03:58FromGitter<bung87> I’m on Mac
20:04:19FromGitter<bung87> anyway I will try update nim
20:04:30*chaky|work8 joined #nim
20:04:37FromGitter<kaushalmodi> ```code paste, see link``` ⏎ ⏎ *Why would I lie? :)* [https://gitter.im/nim-lang/Nim?at=5babe6558196693171bc9343]
20:04:55*alphageek joined #nim
20:05:33FromGitter<kaushalmodi> @bung87 btw if you build just 2 days back, there's not a big difference between then and today, just few commits inbetween and version change.
20:05:35FromGitter<bung87> I just thought some thing about nil done days ago
20:05:39*alphageek quit (Remote host closed the connection)
20:06:07FromGitter<kaushalmodi> the nil string change happened a month or more ago, I think
20:06:17FromGitter<Vindaar> @bung87 the issue is that `__get_nil_t` is an invalid Nim identifier (must not start with an underscore). Or are you talking about something else?
20:07:05shashlick_Need to change that to nil
20:07:38*chaky|work8 quit (Remote host closed the connection)
20:07:39FromGitter<bung87> yeah , anyway update it then check others
20:08:33*CC6610 quit (Remote host closed the connection)
20:10:30*SitiSchu joined #nim
20:11:02FromGitter<bung87> ok ,that’s old version nimble installation problem.
20:12:35*nox_x22 joined #nim
20:16:29*nox_x22 quit (Remote host closed the connection)
20:21:02*SitiSchu quit (Quit: Leaving)
20:21:23*Perkol quit (Remote host closed the connection)
20:22:22*emsal joined #nim
20:22:58planetis[m]anyway, room is pretty empty
20:23:46*shashlick_ quit (Remote host closed the connection)
20:23:47*emsal quit (Remote host closed the connection)
20:23:52FromGitter<bung87> what’s the command to regen the nim files?
20:24:04*shashlick joined #nim
20:24:39shashlickhave you checked out the repo?
20:24:44shashlickthen just `nimble install -y`
20:25:51*francisl quit (Quit: francisl)
20:26:37FromGitter<kaushalmodi> Has anything changed with borrow pragma?
20:26:47FromGitter<kaushalmodi> this used to work: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babeb8798245f4ae08adde2]
20:27:16FromGitter<bung87> ok gen the same thing, manual change it check others.
20:27:32*SpartanPalace joined #nim
20:28:00shashlickare you still seeing the __get_nil_t()?
20:28:44FromGitter<bung87> yeah , I got same thing, just manually change it
20:29:22*Tuju26 joined #nim
20:29:53*SpartanPalace quit (Remote host closed the connection)
20:31:13*nsf quit (Quit: WeeChat 2.2)
20:32:05*Tuju26 quit (Remote host closed the connection)
20:36:48*skellock joined #nim
20:39:12shashlickjust pushed a fix for you, please check
20:46:25*shadowbane quit (Remote host closed the connection)
20:46:29FromGitter<bung87> `fasttext.h:14:10: fatal error: 'atomic' file not found ⏎ ⏎ #include <atomic>` ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babf0258ddf5867dc7b28dd]
20:48:20*SenasOzys joined #nim
20:48:43*shadowbane joined #nim
20:49:55FromGitter<vivekimsit> Hello guys, I just almost finished my first nim lib
20:49:56FromGitter<vivekimsit> https://github.com/vivekimsit/pretty-bytes
20:50:11FromGitter<vivekimsit> please feel free to give me some feedback
20:50:35FromGitter<vivekimsit> I also tried to put it in nimble but got errors
20:51:16shashlickbung87: not sure, never used clang before
20:52:54FromGitter<bung87> ok I use llvm_gcc but raise ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5babf1a641177e0bc7afb9b2]
20:55:04shashlickmight need -std=c++0x
20:55:08shashlickper the makefile
20:57:05*oats14 joined #nim
20:57:18*tribly quit (Quit: WeeChat 2.2)
20:58:17*tribly joined #nim
21:00:41FromGitter<bung87> how you compile and run the test file ?
21:01:01*oats14 quit (Read error: Connection reset by peer)
21:01:07FromGitter<bung87> I tried clang gcc llvm_gcc
21:03:12shashlicknimble test
21:05:18cavariuxIs there any way to install the dependencies of a lib/bin without running it?
21:05:24cavariuxlike in npm the `npm install` ?
21:05:39FromGitter<vivekimsit> Is there anything equivalent to `any` of typescript in nim?
21:06:24cavariuxYou could either use any or a generic type
21:06:29FromGitter<bung87> ok I got it `~/Nim/bin/nim cpp --cc:gcc -t:-std=c++0x -r tests/testlang.nim`
21:06:56*Whisket10 joined #nim
21:07:09FromGitter<vivekimsit> so `any` is a valid keyword in nim?
21:07:22*ritchie_ quit (Quit: Leaving)
21:07:58cavariuxvivekimsit yep, you can look more info at https://nim-lang.org/docs/typeinfo.html
21:08:20FromGitter<vivekimsit> nice
21:09:31*francisl joined #nim
21:09:38FromGitter<zacharycarter> but it's not a concrete type right?
21:09:54FromGitter<zacharycarter> I think the `any` equivalent from TypeScript would be `JsObject`
21:10:18shashlick@bung87: so is it working for you now?
21:10:23cavariuxOh, if you want to interact with a js backend I would say JsObject but if you want to port a lib it would be Any
21:10:46FromGitter<zacharycarter> ah gotcha
21:10:51FromGitter<zacharycarter> like JS to C backend?
21:11:04cavariuxLike Nim to Js backend
21:11:09FromGitter<zacharycarter> Oh I see
21:11:47*Whisket10 quit (Ping timeout: 252 seconds)
21:11:48FromGitter<zacharycarter> going to start resuming work on my JS libs this evening - been in a massive funk the past two days because of what went on at work
21:11:59FromGitter<zacharycarter> but I need to build myself a virtual resume - so I need to finish these libraries anyway
21:15:15FromGitter<bung87> @shashlick compiled pass but I can’t find the right arg pass to `loadModel`
21:16:34FromGitter<zacharycarter> @bung87 what are you working on again?
21:16:53shashlickwrapper for fastText
21:17:12FromGitter<zacharycarter> ah
21:17:24shashlickI just wrapped it using nimgen, he's trying to use it now
21:17:26shashlickhttps://github.com/genotrance/nimfastText
21:18:43*Vladar quit (Remote host closed the connection)
21:19:14FromGitter<zacharycarter> hrm - I haven't played with Nimgen in a while - I've been porting more libs rather than actually binding to them lately
21:19:23FromGitter<zacharycarter> but I plan to put it through it's paces when I port bullet3 later thi syear
21:20:12FromGitter<zacharycarter> I just found it easier when I last worked with it - to do the bindings myself because I knew c2nim better than nimgen and it was easier to just use c2nim rather than learn a whole new way of attacking a problem
21:20:12FromGitter<vivekimsit> any idea how to check is an object is not null?
21:20:19FromGitter<zacharycarter> objects cannot be nil
21:20:24FromGitter<vivekimsit> I am just trying to do TDD with nim
21:20:37FromGitter<vivekimsit> I am writing a lib to parse ini
21:20:47FromGitter<zacharycarter> there's already an ini parser in Nim pretty much
21:20:49FromGitter<zacharycarter> parsecfg module
21:20:52FromGitter<bung87> https://github.com/Nim-NLP/nimfastText/commit/f763c4305ad8933f1b000c5ad47edfcdac8c1e30
21:20:55FromGitter<vivekimsit> my first test is to check if there is something returned from the method
21:21:12FromGitter<vivekimsit> yes, but I learn most while doing
21:21:15FromGitter<zacharycarter> okay - there's been a lot of internal debate about this in the community
21:21:26FromGitter<zacharycarter> objects can't be null - they're allocated on the stack not the heap
21:21:43FromGitter<zacharycarter> traditionally in C - you'd attack this with returning a non-zero number or something - a negative one or something like that from the procedure
21:21:52FromGitter<zacharycarter> or an uninitialized struct or something like that
21:21:53FromGitter<vivekimsit> I read that they can be allocated on the heap as well
21:22:00FromGitter<zacharycarter> ref objects can
21:22:10FromGitter<zacharycarter> but not just pain objects - they're analogous to structs in C
21:22:23FromGitter<vivekimsit> well this distinction of ref object and object is completely new to me
21:22:26*tribly quit (Quit: WeeChat 2.2)
21:22:30FromGitter<zacharycarter> let me find you a good explanation then
21:22:39FromGitter<vivekimsit> and I don't know C that well
21:22:50FromGitter<zacharycarter> https://nim-by-example.github.io/types/objects/ - so here's kind of a high level overview
21:23:00shashlick@zacharycarter: well, nimgen is all about automation so the main reason i push it is so that libs can keep up with upstream changes
21:23:00FromGitter<vivekimsit> I am more of a JS, Python and Java guy
21:23:15shashlickdid you get a chance to try out nimnuklear? @AlexMax spent some time getting it functional
21:23:24FromGitter<vivekimsit> @zacharycarter are you into TDD?
21:23:27FromGitter<zacharycarter> https://peterme.net/nim-types-originally-a-reddit-reply.html
21:23:49FromGitter<vivekimsit> cool
21:23:55FromGitter<vivekimsit> into my todo :D
21:24:04FromGitter<zacharycarter> @vivekimsit - no not really - the company I work at is a fan (even though no one actually does it at work) - but I think it has value in certain places
21:24:43FromGitter<vivekimsit> so, I haven't implemented my method yet
21:24:51FromGitter<vivekimsit> its like a stub function right now
21:24:55FromGitter<zacharycarter> mmhmm
21:25:00FromGitter<vivekimsit> I need to pass my test
21:25:11FromGitter<vivekimsit> I have declared the return type as table
21:25:23FromGitter<zacharycarter> can you post a gist maybe?
21:25:29FromGitter<zacharycarter> or something similar?
21:25:29FromGitter<vivekimsit> but I am not concerned about its content right now
21:25:39FromGitter<zacharycarter> sure - are you using the unittest module?
21:25:40FromGitter<vivekimsit> wait
21:26:34*tribly joined #nim
21:31:56FromGitter<bung87> https://pastebin.com/qbyLsLQe
21:32:29FromGitter<vivekimsit> @zacharycarter here it is ⏎ ⏎ https://github.com/vivekimsit/ini
21:32:41FromGitter<zacharycarter> good looks - I will have a look now
21:33:19FromGitter<vivekimsit> its almost new :p
21:33:55*zachcarter joined #nim
21:34:37FromGitter<zacharycarter> :) it's always fun to have new community members to try and help!
21:35:11*defragger1 joined #nim
21:35:51shashlickbung87: what's in your test file
21:36:02shashlickmight be best to follow what's in the http://main.cc file example
21:36:22*zachcarter quit (Client Quit)
21:36:34*zachcarter joined #nim
21:37:03FromGitter<vivekimsit> thanks
21:37:31FromGitter<bung87> https://github.com/Nim-NLP/nimfastText/blob/lang_test/tests/testlang.nim just try to load model file
21:38:40*defragger1 quit (Remote host closed the connection)
21:40:42FromGitter<bung87> and it throw `note: candidate function not viable: no known conversion from 'NimStringDesc *' to 'const std::string'`
21:40:52shashlickbung87: try `var ft: FastText` instead
21:44:26FromGitter<bung87> hmm same thing
21:48:46*francisl quit (Quit: francisl)
21:49:14FromGitter<zacharycarter> @vivekimsit - I started a job complaint thread in off-topic asking Nim members for recommendations about work - so I've taken myself away from your problem - but i haven't forgotten. I will reply asap.
21:51:09FromGitter<bung87> @zacharycarter hello
21:51:19FromGitter<zacharycarter> @bung87 Hi
21:52:47*urluck joined #nim
21:53:01shashlick@bung87: can you post that ftz file?
21:53:55FromGitter<bung87> https://github.com/Nim-NLP/nimfastText/blob/lang_test/tests/testlang.nims it’s here if you r on linux you will get it automaticlly
21:54:00FromGitter<vivekimsit> @zacharycarter np, I am myself trying to job hunt too :)
21:54:19FromGitter<vivekimsit> btw I am making it OOPsy now
21:55:24shashlickno the langModel file you are openeing
21:55:27shashlickit's not in source control
21:55:58FromGitter<bung87> https://s3-us-west-1.amazonaws.com/fasttext-vectors/supervised_models/lid.176.ftz
21:56:38shashlickokay, got the link from the nims file, sorry
21:57:09*urluck left #nim (#nim)
21:57:28FromGitter<bung87> I start to thinking it compile with nim cpp , what happen when other project depens on it ?
22:00:25shashlickbung87: your problem is that loadModel() expects std::string but you are providing a nim string
22:00:39shashlicki don't know how to convert/cast from nim string to that
22:00:40FromGitter<zacharycarter> @vivekimsit OOP is a dicey road w/ Nim
22:01:07FromGitter<zacharycarter> not that you can't write OOP with Nim - but it really doesn't mesh well with Nim's design
22:01:17FromGitter<zacharycarter> Nim was designed with imperative coding in mind
22:01:31FromGitter<bung87> so what the `proc loadModel*(this: var FastText; a2: string)` does ?
22:01:35FromGitter<zacharycarter> and thus classes / objects are not treated the way they are in Java / C# etc
22:01:56FromGitter<bung87> the type `string ` does not make effect ?
22:02:01*Mitar18 joined #nim
22:02:48FromGitter<bung87> if as it that means all the interface needs a type converter...
22:03:29shashlickanyone knows how to convert a nim string into std::string
22:03:40shashlickya - c2nim converted std::string to nim string but that might not be right or functional
22:05:00*superbia joined #nim
22:05:14superbiahow to quickly remove choosenim from my linux system?
22:05:25FromGitter<bung87> hmm, I’m worried about the probability of this project gets works...
22:06:20*Mitar18 quit (Remote host closed the connection)
22:06:44FromGitter<zacharycarter> well that's always the problem with automation
22:07:04FromGitter<zacharycarter> but at least if you run the automated process frequently enough - you should catch and be able to remedy
22:08:59shashlickbung87: shouldn't be too hard
22:09:05shashlickjust takes time
22:09:43FromGitter<bung87> I just thought nimgen has examples of cpp wrap ,found the author is you...
22:09:45AlexMaxsuperbia: I think choosenim installs itself to ~/.choosnim
22:09:52AlexMaxand there's also some stuff in ~/.nimble too I think
22:10:12AlexMaxerm
22:10:14superbiaso delete those two folders, anything else?
22:10:15AlexMax~/.choosenim
22:10:32AlexMaxwas missing an e there
22:10:42AlexMaxI think that's it - if you care, you can remove it from your PATH
22:11:24FromGitter<bung87> atleast we can dig something from nim source
22:11:31superbiain .bashrc just ?
22:12:40*UFC17 joined #nim
22:13:41superbiaremoved, thank you AlexMax
22:14:05FromGitter<bung87> https://github.com/nim-lang/Nim/blob/27b081d1f77604ee47c886e69dbc52f53ea3741f/tests/exception/tcpp_imported_exc.nim it defined a new type `std_string* `
22:14:48*UFC17 quit (Remote host closed the connection)
22:15:40FromGitter<bung87> which can contruct from cstring
22:15:59shashlickbung87: great - let me update the generation to do that instead of hard coding to string
22:16:31FromGitter<bung87> ok
22:26:03FromGitter<arnetheduck> hey, @alehander42 , just looked at zero_functional, and it looks very nice! is there any way not to use that `it` name and let the caller specify it instead? it's quite limiting in the sense that if you want to nest stuff, you're screwed - likewise with any "outside" variables you might have - example: ``` ⏎ x.any(y.any(y-it == x-it))``` to see if there's overlap in two seqs.. it's like having `select` without
22:26:03FromGitter... `join` - nice for simple examples, but not terribly useful.
22:27:03*superbia left #nim ("WeeChat 2.2")
22:28:11FromGitter<zacharycarter> @vivekimsit is your project targeting the JS backend?
22:31:57shashlickbung87: works now
22:32:02shashlickgoing to push
22:35:48shashlickplease try now - updated the test sample as well
22:41:54FromGitter<bung87> Illegal storage access. (Attempt to read from nil?)
22:43:03FromGitter<bung87> https://github.com/Nim-NLP/nimfastText/commit/f8d0a34cc0412c663154efec88902c081c2315be
22:46:39*rockcavera quit (Remote host closed the connection)
22:48:08FromGitter<bung87> I dont understand the assign operator of `FastText` removed how it gets initialized?
22:50:23shashlickRemove tests from path
22:54:33FromGitter<bung87> which *path*?
22:57:13shashlickThe path to the ftz file
22:58:23shashlickYou are running in the tests directory
23:00:18FromGitter<bung87> I use .nims swich the path
23:00:19FromGitter<mratsim> @bung87 the most useful NLP libs I’ve used are Syuzhet in R, GloVe in python/C++ and spaCy in Python/Cython.
23:00:43*craigger quit (Quit: bye)
23:02:04FromGitter<bung87> I know Glove and spaCy, spaCy is cool.
23:02:43FromGitter<mratsim> Syuzhet is good for out-of-the-box sentiment analysis on tweets or TripAdvisor rating for example
23:04:56FromGitter<bung87> I do some works on word segment ,word pos tag , similarity, summary and sentiment these months. just thoughts port some libs to nim ,make it more faster
23:04:59FromGitter<mratsim> I sorry it wasn’t glowe, it was gensim. There is also Rasa-NLU and MITIE that impressed me
23:05:59FromGitter<bung87> Syuzhet I have not try, not familar with R and the R syntax looks ugly to me..
23:06:05FromGitter<mratsim> mitie can probably be wrapped in Nim as it has a C API: https://github.com/mit-nlp/MITIE
23:07:16FromGitter<mratsim> iirc, RASA is built on top of mitie but it’s been a while since I’ve checked
23:07:44shashlickbung87: I had the same illegal crash, cause it was pointing to a file that doesn't exist
23:08:43FromGitter<bung87> shashlick `echo existsFile($langModel)` I checked ,how that happend?
23:08:54*skellock quit (Ping timeout: 244 seconds)
23:09:09shashlickhmm
23:10:22FromGitter<bung87> @mratsim well ,seems spaCy has more features than MITIE
23:10:35shashlickbung87: how are you running the test
23:11:23FromGitter<bung87> `~/Nim/bin/nim cpp --cc:llvm_gcc -t:-std=c++0x -r tests/testlang.nim`
23:11:28*skellock joined #nim
23:13:26FromGitter<mratsim> yeah of course, MITIE is more out of the box preconfig built on top of dlib
23:13:48FromGitter<mratsim> dlib being a C++ general ML lib
23:14:47FromGitter<mratsim> well gotta sleep. Hopefully I can finish GRU in Arraymancer this weekend and start playing with pure Nim NLP ;)
23:14:51FromGitter<bung87> https://github.com/frcchang/zpar there’s a word segment lib have check yet
23:15:40FromGitter<bung87> @mratsim it’s target to Deep learnning ?
23:15:45FromGitter<mratsim> something not updated since 2015 is probably severely out of date
23:16:04FromGitter<bung87> yeah , its features looks good though.
23:16:12FromGitter<mratsim> yes Arraymancer will offer state of the art DL stuff
23:16:54FromGitter<mratsim> though I will get as much useful stuff from sklearn ported as well, like TfIdf + TruncatedSVD for text classification
23:17:17FromGitter<bung87> that’s the cool thing I’m waiting, I only worry about I have no good machine to train
23:17:54FromGitter<mratsim> I’m always developing CPU first on a 2015 dual-core Macbook, and I take good care of optimization
23:17:59FromGitter<mratsim> I wouldn’t worry for NLP
23:18:00shashlickbung87: not sure what to tell you - works fine here with a correct path
23:18:45FromGitter<bung87> @shashilick can you post you test code to GitHub ,I will check it
23:19:28FromGitter<mratsim> basically the bottleneck for NLP is using softmax if you have a choice of 50k words :P
23:20:06shashlicklatest on github is working copy
23:20:53FromGitter<bung87> @mratsim hmm , I only have experience with trained mode in DL, tried once but dont know how long it will takes to train..
23:21:12FromGitter<zacharycarter> don't forget about - https://github.com/apple/turicreate
23:21:18FromGitter<zacharycarter> built for mbp hardware
23:22:46AlexMaxhrm
23:22:47AlexMaxthis is odd
23:23:40AlexMaxhttps://paste.ee/p/BBMzW
23:23:45AlexMaxNot sure how to parse this error message
23:24:25AlexMaxI thought the way to convert an array to a pointer is to addr array[0]
23:24:33AlexMaxI'm sure it's worked for me before...
23:26:37FromGitter<bung87> shashlick I tried with your test , got same error, no clue how it happens..
23:26:50AlexMaxIt's like nim can't figure out what type it's dealing with
23:27:46*bear45418 joined #nim
23:29:46FromGitter<zacharycarter> probably need `.ord` or something AlexMax
23:30:03*rockcavera joined #nim
23:30:23FromGitter<zacharycarter> I think the `[]` proc is interpreting your enum values as integers
23:30:28FromGitter<zacharycarter> when you really want ordinals
23:30:30shashlickbung87: i'm on win/lin so don't know why Mac isn't working
23:30:36AlexMaxOH
23:30:40AlexMaxThat would explain it
23:30:41shashlickanyone has a Mac VM somewhere
23:31:17FromGitter<zacharycarter> I feel like there's a mac version of qvm or something somewhere
23:31:27FromGitter<bung87> may you can continue test https://github.com/bung87/whatlangid/blob/master/tests/test_predict_lang.py
23:31:32FromGitter<zacharycarter> https://github.com/kholia/OSX-KVM
23:31:33FromGitter<zacharycarter> sorry
23:31:34FromGitter<zacharycarter> kvm
23:31:54AlexMax@zacharycarter actually, accessing the table items using the enum works for me. It's using a raw integer like 0 that confuses things.
23:31:58FromGitter<bung87> the rest of things , I currently dont know what problem I am facing..
23:32:15AlexMaxtable[nuklear.COLOR_TEXT] works, table[0] does not
23:32:30*bear45418 quit (Remote host closed the connection)
23:32:46shashlick@zacharycarter: are you able to download ISOs without purchase?
23:36:50FromGitter<bung87> shashlick if it works fine you will get something like ` [ (“__label__ja”,0.33434),…]` when you put a sentence
23:38:26FromGitter<bung87> `loadModel` and `predict` sames to other models
23:40:08shashlickwhat do you mean when you say put a sentence
23:42:04FromGitter<bung87> `wtl.predict_pro(["English sentence", "അമ്മ"]) ⏎ [('en', 0.8848170638084412), ('ml', 0.9535570740699768)]` like this in python wrapper
23:43:02*Fusl joined #nim
23:43:48FromGitter<bung87> I checked the `predict` proc seems not like offical python api takes optional args
23:44:01FromGitter<zacharycarter> <shashlick> @zacharycarter: are you able to download ISOs without purchase? - yeah I know how to put a patch over one eye and a wooden peg on one of my legs
23:44:07*Fusl quit (Remote host closed the connection)
23:44:40FromGitter<zacharycarter> I need to go do a few things around the house - but I'll be back shortly
23:45:07FromGitter<bung87> https://github.com/facebookresearch/fastText/blob/master/python/fastText/FastText.py#L100
23:48:27FromGitter<bung87> basically I just interested in the use parts, not the train parts , since it provide all pre-trained models to public .
23:59:38AlexMaxoh dear
23:59:46AlexMaxI knocked the style demo out of the way rather quickly
23:59:57AlexMax....not so much the overview
23:59:57*noonien quit (Quit: Connection closed for inactivity)