<< 11-09-2019 >>
00:05:30*dddddd quit (Ping timeout: 246 seconds)
00:08:33*dddddd joined #nim
00:12:14*krux02_ quit (Remote host closed the connection)
00:14:04*wildtrees quit (Quit: Leaving)
00:31:01*fov60 joined #nim
00:31:41*m712 quit (Quit: bye-nyan!)
00:47:23*ryukoposting quit (Quit: WeeChat 1.6)
00:52:41fov60yo is there a way to import nil and as?
00:53:21disruptekimport foo as bar except nil
00:55:35disruptekfrom foo as bar import nil
00:56:05disruptekcan't say i'm sure what you're after, honestly. what are you trying to do?
00:56:52fov60yeah that's what i was trying to do, but turns out I didn't even need it *smh*
00:58:38disruptektrue; it's rarely needed.
00:59:21disrupteki wonder why my machine isn't slow due to this json fragment behavior.
01:17:43FromGitter<mratsim> json fragment?
01:18:43disruptekit's really due to a split on whitespace across a 360mb logfile, i guess. but each line is json that needs parsing. and for federico, it was super slow. but, fast for me.
01:19:20disruptekhttps://github.com/nim-lang/Nim/commit/2f10f5fac3c676daf34179e93ff5fdc26c98b72c
01:34:14*sealmove joined #nim
01:49:10FromGitter<gogolxdong> https://github.com/nim-lang-cn/db-analyzer/ the first project of Chinese nim-lang community
01:51:25FromGitter<gogolxdong> We aim to build a production level web server to beat Spring of Java and fasthttp of Go.
01:54:29FromGitter<gogolxdong> 3X than fasthttp ,4X than Netty ,20X than Spring
02:00:11*mal`` quit (Quit: Leaving)
02:05:11*mal`` joined #nim
02:13:54sealmovegood luck :-)
02:17:39*mal`` quit (Ping timeout: 240 seconds)
02:22:43FromGitter<zacharycarter> +1 - netty is pretty fast
02:23:12*fov60 left #nim ("WeeChat 2.5")
02:23:32FromGitter<zacharycarter> @Varriount thanks bud <3
02:24:10*mal`` joined #nim
02:25:19*dddddd quit (Remote host closed the connection)
02:30:17*btop joined #nim
02:41:52*shashlick joined #nim
02:52:19sealmoveprobably stupid question but does "owned" keyword make sense for seqs?
03:06:01leorizeno
03:11:38sealmoveby the way leorize, I have an issue with Nim... regarding namespaces. Say I have 3 types, then I have to make procs like getA, getB, getC. It would be nice to have something like A.get, B.get, C.get where A, B, C are not object but are typedesc. Is this possible?
03:12:03leorizetotally
03:12:14FromGitter<zacharycarter> lmao - was checking out the Chinese Nim website and used google translate (at least I hope this is a google translation issue)
03:12:28FromGitter<zacharycarter> on the community page I saw this - `"Nim Development Concentration Camp" QQ group`
03:12:47sealmovehelp? :D what would a proc signature look like?
03:13:10leorizeproc get(_: typedesc[A], <args here>)
03:13:23sealmoveah I see, this is great :)
03:17:06sealmoveI am making a Nim backend for the Kaitai Struct compiler
03:17:26sealmoveIt's in Scala for which I have mixed feelings
03:21:26sealmoveBut anyway this is my first compiler project :), and it involved Nim! (even if it's just generating Nim code with Scala)
03:22:49shashlickwhat's the cheapest osx setup i can get to do some nim testing
03:22:50leorizeg'luck :)
03:23:20FromGitter<mratsim> Travis CI? :P
03:23:35shashlickit never gives me complete logs
03:23:42leorizeazure pipelines
03:26:43leorizehere's a CI pipeline for Nim if you're interested: https://github.com/alaviss/Nim/blob/azure-pipelines/azure-pipelines.yml
03:26:59*chemist69_ joined #nim
03:28:44*laaron quit (Remote host closed the connection)
03:29:43*chemist69 quit (Ping timeout: 246 seconds)
03:30:36*laaron joined #nim
03:34:34*rockcavera quit (Read error: Connection reset by peer)
03:34:54*rockcavera joined #nim
03:34:54*rockcavera quit (Changing host)
03:34:54*rockcavera joined #nim
03:38:53*deansher quit (Read error: Connection reset by peer)
03:38:55*msmorgan quit (Ping timeout: 276 seconds)
03:40:09*deansher joined #nim
03:41:59*nuxdie quit (Read error: Connection reset by peer)
03:44:36*nuxdie joined #nim
03:44:50*deansher quit (Ping timeout: 276 seconds)
03:45:32*msmorgan joined #nim
03:46:19*deansher joined #nim
03:47:39sealmoveif I pass an owned ref as a parameter, then the proc that received it becomes the new owner?
03:47:52sealmove(is this the old "sink"?)
03:49:48leorizeit works like sink
03:49:54leorizestill distinct from sink though
03:50:18leorizebut yes, the proc that receive it will be the new owner
03:57:30shashlickThanks @leorize - will try at some point
03:57:49shashlickfor the near term, I need to figure out that | tee solution and upload to a paste service if build fails
03:58:39leorizewell, put every single command in a shell script
03:58:52leorize./script.sh 2>&1 | tee build.log
03:59:06leorizethen use curl to upload it to ix.io if build fails
04:00:29shashlicktee can append to the log too right?
04:00:40shashlickanyway I just have a nimble test command that is failing so it's easy that way
04:00:43leorizeyea, `tee -a`
04:03:02*cgfuh quit (Ping timeout: 276 seconds)
04:26:30FromGitter<gogolxdong> Is the Concentration translation the funny thing?
04:27:33*btop quit (Ping timeout: 245 seconds)
04:27:45FromGitter<zacharycarter> concentration camp is the funny thing
04:28:13FromGitter<zacharycarter> that's what labor / death camps were referred to as in the west during WWII
04:28:33FromGitter<zacharycarter> like Auschwitz = concentration camp
04:29:02FromGitter<zacharycarter> but I'
04:36:24*btop joined #nim
04:52:15*nsf joined #nim
04:56:04*rockcavera quit (Remote host closed the connection)
04:58:13*fjellfras joined #nim
05:16:46*narimiran joined #nim
05:17:59*juturnas quit (Remote host closed the connection)
05:18:11*fjellfras quit (Quit: Leaving)
05:23:41*absolutejam4 joined #nim
05:38:48*leorize quit (Ping timeout: 260 seconds)
05:41:50FromGitter<gogolxdong> yeah,I thought the translation wouldn't be the meaning I thought as you said.We will change if it is not proper.
05:46:30FromGitter<zacharycarter> what are you trying to say there? some type of Nim developer group?
05:54:49*laaron quit (Remote host closed the connection)
05:54:56FromGitter<gogolxdong> just a community name.
05:57:23*laaron joined #nim
06:01:42FromGitter<zacharycarter> `发展小组` might be better - google translate says in simplified Chinese this means 'Development Group'
06:01:49FromGitter<zacharycarter> but I'll let you tell me :)
06:02:17FromGitter<zacharycarter> or maybe `开发者小组` which google says = `Developer Group`
06:02:59owl_000wnim gui for windows. latest release (today) significantly reduces compilation time. previously `wnim` includes almost all the modules. now user can import widget by widget.
06:04:09Araqowl_000: great news!
06:05:21FromGitter<gogolxdong> yes, 开发组 means Development Group
06:06:30owl_000oviously, it is the most feature rich gui for nim.
06:10:10*solitudesf joined #nim
06:11:16owl_000one curious question-> when i compile, the output shows `hint: someModule [processing]`, what is the compiler actually doing.
06:12:55*laaron quit (Remote host closed the connection)
06:16:23*laaron joined #nim
06:20:04Araqit compiles the module
06:27:56*m712 joined #nim
06:32:11*alexander92 joined #nim
06:33:38*absolutejam4 quit (Ping timeout: 240 seconds)
06:36:23*PMunch joined #nim
06:55:19*laaron quit (Remote host closed the connection)
06:56:44*PMunch quit (Remote host closed the connection)
06:57:06*m712 quit (Ping timeout: 246 seconds)
06:57:33*PMunch joined #nim
06:58:36*Kaivo quit (Read error: Connection reset by peer)
06:58:42*laaron joined #nim
06:58:59*m712 joined #nim
06:59:36*Kaivo joined #nim
07:00:00*gmpreussner quit (Quit: kthxbye)
07:04:45*gmpreussner joined #nim
07:05:43*ng0 joined #nim
07:06:15*krux02 joined #nim
07:06:42*btop quit (Ping timeout: 245 seconds)
07:19:00*sealmove quit (Quit: WeeChat 2.5)
07:21:06*absolutejam4 joined #nim
07:22:52*arecaceae quit (Remote host closed the connection)
07:23:11*arecaceae joined #nim
07:27:20*floppydh joined #nim
07:33:45*sealmove joined #nim
07:34:43*btop joined #nim
07:39:27*btop quit (Ping timeout: 246 seconds)
07:57:21*gangstacat joined #nim
08:00:24sealmoveStill trying to wrap my head around Nim's structural typing. In C, if the field of a structure is a non-primative type, you have to refer to it with a point. In Nim you could either write `fieldName: A` or `fieldName: ref A` or maybe `fieldName: owned A`. I am having trouble understanding the differenc between `fieldName: A` and `fieldName: owned A`... Is it the same thing?
08:01:23Araq'owned A' is WIP
08:01:34PMunchThe difference between fieldName: A and fieldName: ref A has to do with how it is stored in memory and how it is passed around
08:01:36Araqand makes no sense without the 'ref'
08:01:59PMunchowned A is something that is used with `newruntime`
08:02:16PMunchCouldn't A be a type that is a ref in that case Araq?
08:02:16Araqneither C nor Nim have "structural" typing for objects, they use nominal typing
08:02:36AraqPMunch, sure but fromt the context I figured it doesn't
08:03:14PMunchsealmove, https://peterme.net/nim-types-originally-a-reddit-reply.html if you want a deeper explanation of ref vs. non-ref
08:06:38sealmoveThanks. If I would make the vast simplification: Use `fieldName: A` if data won't change much, use `fieldName: owned ref A` if data is going to move around, would that make somewhat sense?
08:07:14*absolutejam4 quit (Ping timeout: 240 seconds)
08:07:27sealmove(my bad, structural vs nominal is different from what I had in mind)
08:07:28PMunch`owned` should only be used when you are using `newruntime`
08:07:57Araqsealmove, you can't simplify it, it's the core of "systems programming"
08:08:18PMunchAnd it doesn't really have to do with how often you want to change something, more about how you want to change them
08:08:20Araqpointers vs values is essential
08:09:03sealmovePMunch, I am thinking of how to model structure of file formats. I think they are trees (for example png has fields which have fields).
08:09:27*btop joined #nim
08:10:33sealmoveJava does it with nested classes
08:11:42euantorAraq: If you're around, regarding https://github.com/nim-lang/Nim/pull/12105 - is the plan to merge that and then look at https://github.com/nim-lang/Nim/pull/12137 ? The second one fixes a couple of the OpenBSD CI errors that are repeated for several tests (all the `pthread` errors and the `time_t` error)
08:11:54sealmoveexample: https://pastebin.com/iyZDGrSW
08:12:11AraqI registered at sourcehut and enabled github integration
08:12:34Araqbut https://github.com/nim-lang/Nim/pull/12105 still lists "2 checks", not 3
08:12:41sealmoveowned refs kind of blur the line between pointers and values, no?
08:12:58Araqno, they do not
08:13:05*absolutejam4 joined #nim
08:13:10euantorOh, that's weird
08:13:12PMunchsealmove, just forget about owned
08:13:24euantorDid you enable it for both PRs and commits?
08:13:28PMunchAt least for now
08:13:46Araqeuantor, I think I did, yes
08:13:55PMunchDid you read the article I linked you? I feel anything I would write here would mostly just be a repeat of that..
08:14:16*btop quit (Ping timeout: 246 seconds)
08:14:30Araqsealmove, or actually, maybe they do that a bit
08:15:08euantorMight be worth dropping a message in #sr.ht if it stil ldoesn't work. The page should look similar to this: https://usercontent.irccloud-cdn.com/file/j3ie6k64/Capture.PNG
08:15:35sealmovePMunch, I will read for sure, seems very interesting article, thanks for posting
08:16:00Araqeuantor, it does look like that
08:16:38euantorHmmm. It should be working then. I've never tested the PR integration myself, perhaps it only works for new PRs or something?
08:17:04euantorI'll drop drew a message in the IRC channel and see if he can have a look
08:17:59PMunchsealmove, I wrote it originally as a reply to a Reddit question that was essentially the same thing you're asking about now :)
08:18:16PMunchYou can click through from the article to Reddit to see the original question and the following discussion
08:18:29*actuallybatman quit (Ping timeout: 276 seconds)
08:19:22euantorAlso, I assume there are two new WebHooks created on the repository in GitHub for `pull_request` and `push` pointing to `dispatch.sr.ht`?
08:20:00euantorhttps://www.irccloud.com/pastebin/QWILhCwH/
08:20:13euantorThat explains it. I'll push an empty commit and see if it triggers
08:23:01*ng0 quit (Quit: Alexa, when is the end of world?)
08:23:24euantorThat seems to have done the trick: https://builds.sr.ht/~araq/job/88649
08:26:33*gangstacat quit (Quit: Ĝis!)
08:27:30sealmoveI still can't see how value-based semantics can lead to faster code. I suspect it has to do with compiler optimizations and will leave it to that (too advanced for me probably)
08:28:04sealmoveThe fact that it's safer I understand
08:29:51*gangstacat joined #nim
08:34:17Araqsealmove, hardware caches play a big role for this
08:34:45Araqplus it's easier to reason about, much easier. both for compilers as well as for human beings
08:35:51sealmoveThis make sense, it's also my intuition
08:40:00*absolutejam4 quit (Ping timeout: 246 seconds)
08:44:50*absolutejam4 joined #nim
08:46:04*Vladar joined #nim
08:46:31PMunchsealmove, essentially it "costs" the computer some time to follow a pointer, especially when it's not something that hasn't been used in a while since it's out of the cache. The stack memory around your call is more likely to be in cache. So value-based types can be faster simply because they are more often cached, and you don't have to follow a pointer to get to them.
08:47:36PMunchThat being said if you want to pass a value-based object into a function it will have to be copied (I think Nim does some trickery here and actually passes it as a pointer to the previous stack frame, but I'm not entirely sure).
08:47:37*dddddd joined #nim
08:47:52PMunchBut in general value-based objects will lead to more copying of memory, which might in turn be slower.
08:47:59PMunchBut it's very dependent on workload
08:51:48AraqNim uses pointers for parameter passing if benefitial
08:52:51Araqbut copying can be an issue if you use too many value types and the values get bigger and bigger
08:53:27PMunchWhat is the maths on "beneficial"?
08:53:39PMunchIs it just anything bigger than a pointer?
08:53:44Zevv3 * sizeof(float)
08:53:49Zevvor something funny like that
08:54:18AraqI remember changing it to 4 * sizeof(float)
08:54:23Zevv:)
08:55:20Araqwe could make it a command line option so people can tinker with it and see if it makes a difference
08:56:16PMunchHmm, why sizeof(float)?
08:56:28PMunchAnd having it as a switch could be interesting
09:00:16euantorOpenBSD CI is working, though the build is still failing as expected: https://builds.sr.ht/~araq/job/88649 :)
09:00:55euantorIt doesn't send any emails on failure though, you have to add that to the job - I wasn't sure if that would be desired or not?
09:01:26Araqif I don't get an email, it doesn't exist
09:01:48euantorok, what email do you want it sending to?
09:02:21Araqmine
09:02:33euantorIf you don't want it exposed in the build file for some reason, it can be added as a build secret and then referenced
09:02:56euantorDocs: https://man.sr.ht/builds.sr.ht/triggers.md
09:07:16Araqeuantor, any insights what it implies for our CI times?
09:07:28Araqhow long does a build take?
09:08:07euantorIt took 42 minutes for the last build
09:08:31euantorTravis and APpVeyor are stil lrunning
09:08:38euantorSo it's faster than those at least
09:09:20livcdi just saw Araq fixed something for json
09:12:45Araqlivcd, opinions are welcome
09:13:00Araqit does fix the performance problem by an API addition
09:13:08Araqand usually we don't do that :P
09:13:28PMunchHmm, there are some weird issues on devel at the moment
09:14:50PMunchMy jsonschema module get's complaints about ambiguous calls between the forward declaration and the body it outputs. And in SDL I get an error about "named param already provided": http://ix.io/1UND
09:19:54AraqPMunch, report it properly please
09:24:11PMunchOh I thought these issues were known
09:27:14livcdAraq: but this is only in devel right ?
09:28:54Araqlivcd, it's not even merged
09:33:00PMunchhttps://github.com/nim-lang/Nim/issues/12174 https://github.com/nim-lang/Nim/issues/12175
09:33:15PMunchNot the best issues, but at least they are logged now
09:33:23PMunchI'll add more details if I get time to investigate further
09:34:05PMunchIs there a simple way to auto-bisect the Nim compiler?
09:34:40PMunchEssentially write a small test, then run git bisect with a script that rebuilds the compilers and uses the test as it's check
09:35:19AraqPMunch, dunno I don't bisect, I know all the changes and the cause for this regression
09:36:25PMunchCould possibly be useful for others that aren't as up to date on development to find what causes an issue
09:37:02Araqit's causes by the changes to compiler/sigmatch.nim
09:39:32*ng0 joined #nim
09:44:22Araqnarimiran, https://github.com/nim-lang/Nim/issues/12170 how can 'yaml' be an important package and yet fail?
09:48:42narimiran'yaml' tests are bad and don't catch stuff like that?
09:49:20Araqpossible :-)
09:55:33*laaron quit (Remote host closed the connection)
09:58:18*laaron joined #nim
10:11:29*abm joined #nim
10:21:52AraqPMunch, how does this nimlsp setup work?
10:21:57Araqimport "../../../Nim/nimsuggest/nimsuggest.nim"
10:21:57Araqimport messageenums
10:21:57Araqimport strutils
10:21:57Araqimport compiler / ast
10:22:01Araq^ ???
10:22:18Araqok, you used "./Nim" instead, but the same problem
10:50:59FromGitter<alehander42> gitter has threaded convos
10:51:01FromGitter<alehander42> now
10:51:13FromGitter<alehander42> i guess this totally wouldn't work with irc/etc
10:52:38FromGitter<kayabaNerve> Really? I don't see any GUI changes.
10:52:49FromGitter<alehander42> you https://next.gitter.im/
10:53:09FromGitter<alehander42> just visit this and enable it from the swich
10:53:20FromGitter<alehander42> no idea if people who haven't enabled it see it
10:53:32FromGitter<alehander42> @kabayaNerve
10:53:46FromGitter<kayabaNerve> Thanks
10:53:51FromGitter<alehander42> ah ok
10:54:07FromGitter<kayabaNerve> Still some bugs, unfortunately :P Chat text is black and it doesn't enable a cursor when clicked.
10:54:10FromGitter<kayabaNerve> No, I enabled it.
10:54:17Araqstill looking for feedback: https://github.com/nim-lang/Nim/pull/12165
10:54:23FromGitter<alehander42> hm, for me a possible issue is that mentions
10:54:24FromGitter<alehander42> dont work
10:54:29FromGitter<alehander42> but not too important
10:54:37FromGitter<alehander42> more like, autocomplete for mentions
10:55:00FromGitter<kayabaNerve> Yep. Still in beta
10:55:44FromGitter<alehander42> hm, would tyhis work
10:55:53FromGitter<alehander42> for e.g. {},[],4,5
10:56:17*oculux quit (Ping timeout: 245 seconds)
10:56:34FromGitter<alehander42> does parseJson finish when it can't consume a character, or only if it's whitespace?
10:57:52Araqit produces an error then
10:58:25*theelous3 joined #nim
10:58:53*oculux joined #nim
11:00:42FromGitter<alehander42> sounds logical
11:16:49*leorize joined #nim
11:27:52*ng0 quit (Ping timeout: 260 seconds)
11:34:49*theelous3 quit (Read error: Connection reset by peer)
11:37:58livcdAraq: i am going to try on my linux vm
11:39:04*couven92 joined #nim
11:41:57*ng0 joined #nim
11:52:06Araqlivcd: and then what?
11:58:44*nif quit (Quit: ...)
11:58:53*nif joined #nim
12:01:27PMunchAraq, sorry I was off to lunch
12:01:48*nif quit (Client Quit)
12:01:51PMunchDid you figure out how it works?
12:02:04*nif joined #nim
12:20:14AraqPMunch: I'm quite sure you f*cked it up ;-)
12:20:29PMunchHaha, how so?
12:20:35Araqno offense implied, I fuck up too, all the time
12:20:46PMunchNone taken, I fuck up plenty
12:21:08Araqwell you use both the Nim compiler nimble package
12:21:32Araqhave some custom stuff in nimlsppkg
12:21:43Araqand git submodules
12:21:59Araqand inconsistent import paths
12:22:07Araq ./Nim/ vs compiler/ast.nim
12:23:03Araqhowever the regression in jsonschema is my fault and I'm working on it
12:23:51PMunchWhere do you see "compiler/ast.nim"?
12:24:10PMunchBut yeah, the entire build system is a bit wonky
12:28:03livcdi thought you wanted some feedback :-)
12:28:28Araqlivcd: about the solution, not how it runs on your machine
12:28:39AraqI know it's faster on your machine too
12:28:45livcdit's actually not
12:28:55Araq:D
12:28:59Araqin theory it's much faster
12:29:04livcdunless i fucked up
12:29:11livcdalso not qualified to talk about the solution sorry
12:29:27Araq for x in parseJsonFragments(newFileStream("jomg.json")):
12:29:27Araq discard x
12:29:37*Araq renamed j to jomg.json
12:29:45Araqis the program to test
12:29:57Araqfor me it's 3 times faster
12:31:46livcdok fine i fucked up
12:43:08disruptekAraq: 2.0s -> 1.39s for me.
12:48:50*rockcavera joined #nim
13:06:58Araqdisruptek: nice
13:08:16AraqError: type mismatch: got <WrapsCancelParams, type WrapsCancelParams>
13:08:17Araqbut expected one of: …
13:08:39AraqPMunch: ^ afaict the compiler is right and it shouldn't compile
13:20:57PMunchIt has compiled fine since I wrote it without any issues..
13:22:08PMunchSince like 0.19.0 or something like that
13:22:59Araqwell see it for yourself, all candidates take a JsonNode as the first parameter and you don't give any
13:23:28Araqnim c src/jsonschema.nim
13:25:10PMunchWait, where are you now
13:26:30Araqin your jsonschema
13:26:56FromGitter<alehander42> when you open your jsonschema and find araq inside
13:28:20PMunchOh, that code haven't been run in a while
13:28:29PMunchThe types have become distinct since then
13:28:47PMunchIf you just do wcp.JsonNode.isValid instead it should work
13:30:39Araqwell give me something easy to reproduce, nimlsp is too hard to setup
13:31:07PMunchHmm, JsonSchema seems to work fine by itself..
13:31:21PMunchSo there has to be more to it that makes it fail..
13:31:33alexander92instantiation of something?
13:32:49*cgfuh joined #nim
13:33:06PMunchIt's strange, but I have to look at it tomorrow. Have to leave early from work today
13:33:08*PMunch quit (Remote host closed the connection)
13:48:04narimirananybody wants to add nim version to this benchmark? https://github.com/ixy-languages/ixy-languages
13:48:36*disruptek 🤣alehander42
13:50:10shashlick@leorize - using tee now makes all tests pass for some reason, non-zero return code doesn't fail
13:50:11*nsf quit (Quit: WeeChat 2.5)
13:50:37shashlickhttps://travis-ci.org/nimterop/nimterop/jobs/583665049#L9564
13:51:08sealmoveDo you guys use "unittest" module (outside of Nim compiler project)?
13:51:19disrupteksealmove: i like it.
13:51:27FromGitter<kaushalmodi> sealmove: I like unittest too
13:51:32shashlickat least i get full logs now but a failure is successful
13:53:00FromGitter<kaushalmodi> If I remember correctly unittest does not use stderr
13:57:54disruptek"ixy.rs is not production-ready. Do not use it in critical environments. DMA may corrupt memory.
13:58:13disruptekuhhhh
14:02:08FromGitter<mratsim> rust is memory safe :p
14:03:17disruptekabstinence is safe.
14:05:41Zevvthat's what she said
14:08:38leorizeshashlick: that's weird
14:08:58leorizeI think `tee`'s zero return value caused the prob
14:09:12shashlickmeh, just my fault - i was usign set -e and travis couldn't capture all logs
14:09:16shashlickwithout it all problems are gone
14:09:22leorizecouldn't you just read the travis log though? it seems to be quite adequete
14:09:27shashlicki could use set -o pipefail if i want to use tee
14:09:37shashlicktravis logs were getting truncated due to set -e
14:10:06leorizehttps://api.travis-ci.org/v3/job/583665049/log.txt
14:10:09leorizeseems full to me
14:11:49shashlickthat one was full because i used tee and it returned 0 so build succeeded
14:12:16shashlickoverall moral - don't use `set -e` with travis
14:13:15leorizeI think they even advised against it :P
14:15:49*owl_000 quit (Ping timeout: 246 seconds)
14:15:54shashlickat least i can stop grumbling about travis now
14:27:07*btop joined #nim
14:32:23*btop quit (Ping timeout: 268 seconds)
14:34:18*owl_000 joined #nim
14:37:53livcdfor me it's 5.05164909362793 vs 3.518282651901245 ./araq_fix
14:48:58sealmoveEmbarrassing question but when I run "nim c -r test tdefault_big_endian.nim" I get "Error: cannot open 'test.nim'".
14:49:38sealmoveWhat am I missing?
14:51:00disrupteka slash between test and tdefault_big_endian.nim?
14:52:49*vegax87 quit (Ping timeout: 250 seconds)
14:53:28*drewr quit (Ping timeout: 264 seconds)
14:55:48sealmovebut documentation shows "test" as an argument before the test file https://nim-lang.org/docs/unittest.html
14:56:28*fjellfras joined #nim
14:57:30disruptekin that example, test is a test.nim and the -r is running it with arguments "something" and "something else" which correspond to test names as declared `test "something": ...` and `test "something else": ...` in test.nim.
14:58:07disrupteki wish the compiler didn't add a .nim to try to find source.
15:00:54shashlick@disruptek - libevent wrapper - http://ix.io/1UP6/nim
15:01:01sealmovefrom the looks of it I would swear "test" is this example is not supposed to be a file (test.nim), but whatever
15:01:07sealmovein*
15:01:18*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
15:01:21disrupteki know, it's confusing for weak reasons.
15:01:34disruptekshashlick: sweet man, thank you!
15:02:03shashlickgetHeader() is working pretty nicely - anyone has other good challenges for wrappers? not the C -> Nim portion but the part of getting the library built and ready for wrapping
15:02:59shashlick`nim c -d:eventGit -d:eventStatic event.nim` => will download and build library from git and then statically link to it
15:03:18*laaron joined #nim
15:03:22disruptekyeah, it's cool.
15:03:25shashlickput your -d: stuff in a .cfg file and `import event` and you are good to go
15:04:48*floppydh quit (Quit: WeeChat 2.5)
15:07:04disruptekyou should really put that example on the main nimterop readme, no?
15:09:06*zyklon joined #nim
15:11:08*drewr joined #nim
15:13:16*uvegbot quit (Ping timeout: 264 seconds)
15:14:15*Tyresc joined #nim
15:15:54*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
15:16:12*laaron joined #nim
15:18:53FromGitter<awr1> @Araq i don't get why this is not a seperate test module
15:19:02FromGitter<awr1> https://github.com/nim-lang/Nim/blob/8c2e490a8cbf3de5a9d59b24ede7b9c01699caa3/lib/pure/json.nim#L1603
15:20:43alexander92sealmove, i use it too
15:21:31alexander92you just compile and run tests
15:21:45alexander92the example might be confusing, but you just nim c test.nim; ./test
15:21:52alexander92(or use nim c -r test.nim)
15:22:13alexander92(but i prefer to pre-build my tests..not sure if that is actually good)
15:22:57*absolutejam4 quit (Ping timeout: 268 seconds)
15:23:06alexander92ah, i am not sure about the subtest cases filtering, i dont recall using it
15:23:46sealmoveok thanks :-)
15:29:33FromGitter<awr1> @zacharycarter that bit about the translation reminds of a screwup with a certain finnish dev studio
15:30:10FromGitter<awr1> for city skylines they had some kind of thing in the game where players could build universities and there was like some kind of research tree in the game
15:30:40*Hideki_ joined #nim
15:31:09FromGitter<awr1> and like the last research thing was intended to be like "the ultimate answer of reality" or something along the lines
15:31:58FromGitter<awr1> anyway after finnish to english translation it had become "the final solution" which of course was...concerning
15:32:07FromGitter<awr1> they promptly fixed it once they realized though
15:37:16pbbHey, how can I parse an IEEE 754 float from an int?
15:37:26pbbIn C I would just change the type
15:39:43narimiranpbb: can you show us some example of some input and desired output?
15:40:14alexander92if it works in C, it should work in the same way for the nim c backend i guess?
15:40:18alexander92(using cast ?)
15:40:34FromGitter<awr1> do not use cast if you're trying to truncate
15:40:42FromGitter<kaushalmodi> I curious to see what's special about IEEE 754 float
15:40:51FromGitter<kaushalmodi> yes, I wouldn't use cast as the first option
15:41:01FromGitter<kaushalmodi> usually `float(someIntVar)` works
15:41:05FromGitter<awr1> just use float()
15:41:06FromGitter<awr1> yeah
15:41:15FromGitter<awr1> `cast ()` is a bitwise cast
15:41:51narimiranyeah, the way the question was asked, i'm not sure if he wants `float(myInt)` or `cast[float](myInt)`, that's why i asked for a clarification
15:42:17FromGitter<awr1> i forget if C implicitly converts between the two
15:44:26AraqC does value conversions
15:44:33Araqfloat(myint) in Nim
15:44:43AraqC's cast sometimes doesn't cast :-)
15:45:51alexander92ah
15:46:18alexander92good
15:46:40*Hideki_ quit (Remote host closed the connection)
15:46:48Araqnot really low level if you ask me, but whatever
15:48:59alexander92maybe til
15:49:49*btop joined #nim
15:50:00pbbnarimiran, alexander92: cast[float32](foo) worked fine
15:54:15Araqpbb, that's not what C does
15:56:38pbbAraq: what would be the nim equivalent to plain casting in C?
15:57:04narimiranpbb: once again, show us some example of what you're trying to achieve
15:57:46narimiranit is much easier to answer if we know exactly what you're trying to do (show input and output, don't use words for explanation)
16:00:10*rockcavera quit (Remote host closed the connection)
16:00:22*actuallybatman joined #nim
16:01:35pbbnarimiran: https://termbin.com/811x5
16:01:41shashlick@disruptek is still on a branch
16:01:48shashlickI'm closer to merging
16:02:45shashlickCheck out the updated readme
16:02:47pbbnarimiran: I have the int and I want the float. it should be the same in-memory representation.
16:02:48shashlickhttps://github.com/nimterop/nimterop/tree/v020
16:03:01narimiranpbb: thanks. `cast[float32](myInt)` is really the way to go then
16:03:16sealmovesuppose you have a type which has a field of ref to the same type (self-referencial type, a tree for example), and you have a proc that creates an object of this type and returns it. Is it possible to assign a pointer to itself for the self-referential field?
16:03:17*owl_000 quit (Quit: Leaving)
16:03:19pbbok, as I said it seems to work
16:03:23FromGitter<zacharycarter> @awr1 lol
16:04:12*btop quit (Ping timeout: 245 seconds)
16:05:05FromGitter<awr1> @pbb: nim has no perfect "equivalent" to casting to C by design because the meaning of "cast' in C is ambiguous.
16:05:07sealmovefor example if every node has a field called "root", then the actual root should have a pointer to itself
16:05:27sealmovehow to do that?
16:05:46FromGitter<awr1> nim clarifies and destroys the ambiguity by saying that "casting" is a bit-a-bit copy into a new type
16:06:14FromGitter<awr1> and characterizes "conversion" as a transformation between two types that is not necessarily a bit-for-bit copy
16:06:49pbbawr1: is the in-memory representation of float32 in nim always the same?
16:07:04pbbawr1: or is it dependent on the platform so that I would want a conversion in some cases?
16:07:45FromGitter<awr1> `float32` is generally linked to `float` in C. there is a special `cfloat` type though to support the possibility of an edge case
16:08:30FromGitter<awr1> and, to my knowledge, `float` in C is specified to IEEE 754
16:08:54FromGitter<awr1> just `float` (not `float32`!) in nim is slightly different
16:09:30FromGitter<awr1> b/c in 64-bit `float` is `float64` and in 32-bit `float` is `float32`
16:10:46sealmovehttps://pastebin.com/GSZDHB2g
16:11:57FromGitter<awr1> so in general, in-memory rep of float32 should be the same (so as long as the C spec is expected to behave), barring endianness concerns
16:12:14FromGitter<alehander42> sealmove, yes
16:12:29FromGitter<alehander42> it's not so pretty
16:12:34FromGitter<alehander42> but you just do it on two lines
16:12:38FromGitter<alehander42> result = Node(..)
16:12:41FromGitter<alehander42> result.root = result
16:13:05sealmove`result = Node(root: result, parent: nil)` ?
16:13:17FromGitter<alehander42> well, no, because result
16:13:39FromGitter<alehander42> is not still pointing to the new `Node()`
16:13:45FromGitter<alehander42> when you do `root: result`
16:13:57sealmoveI see...
16:14:00FromGitter<alehander42> another option
16:14:07FromGitter<alehander42> is `new(result)` on the first line
16:14:14FromGitter<alehander42> and then just `Node(root: result, ..)`
16:14:22FromGitter<alehander42> on the second
16:14:34FromGitter<alehander42> wait, no
16:14:51FromGitter<alehander42> this is wrong, the first one is fine
16:15:12FromGitter<alehander42> the `result.root = result` thing
16:15:34sealmoveresult = new(Node); result.root = result?
16:15:35*couven92 quit (Remote host closed the connection)
16:16:00FromGitter<alehander42> no, result = Node(..other fields maybe?); result.root = result
16:16:03*couven92 joined #nim
16:16:56sealmoveI see, thanks for help!
16:17:11FromGitter<alehander42> sorry for the confusion, np
16:21:02*Trustable joined #nim
16:27:57*Hideki_ joined #nim
16:31:12FromGitter<zacharycarter> ugh - the shadowmap example from bgfx is 3k lines long
16:31:35FromGitter<zacharycarter> need motivation to start haha
16:31:47FromGitter<zacharycarter> in other news, I found this today - https://github.com/exezin/exengine
16:40:06*couven92 quit (Read error: Connection reset by peer)
16:43:21*Hideki_ quit (Ping timeout: 246 seconds)
16:46:16*fjellfras quit (Quit: Leaving)
16:52:45*nsf joined #nim
16:55:07*rockcavera joined #nim
16:57:31sealmovezacherycarter: wow it looks nice
16:58:26*sealmove quit (Quit: WeeChat 2.5)
17:05:08FromGitter<awr1> the shadow mapping example i assume is CSM?
17:05:36FromGitter<zacharycarter> it has a few different implementations but directional light + csm is one of them
17:05:39*actuallybatman quit (Ping timeout: 240 seconds)
17:05:52FromGitter<zacharycarter> which is what I'd be using
17:06:13FromGitter<zacharycarter> I mean, truthfully, I wouldn't have to port 3k lines of code it's just digging through it and pulling out what I need
17:06:28FromGitter<zacharycarter> I probably will just wait till the weekend to do it
17:07:12FromGitter<zacharycarter> and then I'll need to refactor all of this into separate modules and start building a render pipeline
17:07:13FromGitter<awr1> i wish i had more time to work on my engine
17:07:44FromGitter<awr1> i've been busy with my last semester before i graduate and get my bachelors
17:07:46FromGitter<zacharycarter> I now wish I had written zengine with bgfx
17:07:52FromGitter<zacharycarter> instead of just OpenGL
17:08:02FromGitter<zacharycarter> it'd be worth resurrecting if I had
17:08:20*actuallybatman joined #nim
17:08:22FromGitter<zacharycarter> or maybe I should just bring it back anyway and replace all the opengl calls with bgfx calls
17:08:28*btop joined #nim
17:09:54FromGitter<zacharycarter> on the other hand, I probably already have most of what I was doing with zengine done, so I think I'll just continue down the path I am
17:10:14FromGitter<awr1> i've never used BGFX before
17:10:38FromGitter<awr1> i've looked a little at sokol_gfx though
17:10:48FromGitter<zacharycarter> sokol is nice
17:11:27FromGitter<zacharycarter> but the author is more focused on web than desktop
17:11:42FromGitter<zacharycarter> and I imagine the second version will focus on webgpu
17:12:28FromGitter<awr1> yeah i think i recall talking about it before with the author on twitter
17:14:59*blu joined #nim
17:16:22bluDoes Nim's async support async operations that don't fit into the underlying IO system (epoll, uring, etc.)? Via a threadpool or some other means
17:17:20*actuallybatman quit (Ping timeout: 276 seconds)
17:17:27*Hideki_ joined #nim
17:17:59*btop quit (Ping timeout: 276 seconds)
17:21:38*Hideki_ quit (Ping timeout: 240 seconds)
17:26:55FromGitter<mratsim> import threadpool and use spawn/^
17:28:23*shashlick quit (Ping timeout: 276 seconds)
17:31:05dom96spawn with ^ will block
17:31:38FromGitter<mratsim> @awr1 I think Nimbus `float` is always `float64`
17:31:39dom96you can do this as a workaround: https://github.com/dom96/nim-in-action-code/blob/master/Chapter3/ChatApp/src/client.nim#L40-L50
17:31:46FromGitter<mratsim> Nim*
17:32:04FromGitter<awr1> :/
17:32:23FromGitter<awr1> double precision is overrated tbh
17:32:40dom96this needs to be done to get spawn in async https://github.com/nim-lang/Nim/pull/11724
17:32:50dom96blu, ^
17:33:08FromGitter<mratsim> @pbb, casting int<->float works fine, tested here for a LUT table for fast exponentiation: https://github.com/numforge/laser/blob/master/laser/primitives/simd_math/exp_log_common.nim#L40-L43
17:33:09bludom96, thanks
17:33:21blureading it over now
17:34:56FromGitter<kayabaNerve> Float operations that aren't a BigDecimal are pointless :P
17:35:27FromGitter<kayabaNerve> (in response to double precision is overrated; I fully understand the value of float and double, but in my work, I've always needed BigDecimal precision)
17:35:38FromGitter<awr1> i remember reading somewhere that NASA used single precision floats + fixed point on guidance computers in space and it sufficed for them
17:35:48FromGitter<kayabaNerve> I'd also hate using float when doubles are available.
17:36:10FromGitter<kayabaNerve> I mean, my float work means any float variances between any computer breaks everything.
17:36:39FromGitter<kayabaNerve> That said, I also don't use floats for that exact reason. I just use ints with a human readable repr of X decimals.
17:36:40FromGitter<mratsim> Apparently physics need float64, hence why Nvidia always sell consumer GPU with gimped float64 compute and Quadro with very high markup
17:37:18FromGitter<kayabaNerve> Anything with a fixed amount of decimals can be represented as an int.
17:37:22FromGitter<awr1> but maybe this is just b/c i come from game/GPU dev where basically nothing uses floats
17:37:22FromGitter<mratsim> for ML/Data science you need float32 for training the model and float16 for using the model
17:37:27FromGitter<awr1> save for delta time
17:37:29FromGitter<awr1> https://randomascii.wordpress.com/2012/02/13/dont-store-that-in-a-float/
17:37:33FromGitter<awr1> *nothing use doubles
17:38:06FromGitter<mratsim> or you can use a mixed precision training with individual computation in float16 but accumulation across multiple computations in float32
17:42:33FromGitter<awr1> fp16 has use in gamedev, i've seen it used for lighting purposes
17:43:25FromGitter<awr1> i have heard it once used for stuff like UV coords in vertex data but i dont think it's very common, as opposed to quantized unorm/snorm
17:50:48Calinouusing doubles by default is a good idea, but floats should remain available for performance
17:51:23*Trustable quit (Remote host closed the connection)
17:51:51FromGitter<awr1> i have heard that it has lead to unnecessary performance issues in javascript
17:51:55FromGitter<awr1> who woulda thunk
17:52:22FromGitter<mratsim> in python they are fast because they are not boxed compared to integers :P
17:55:46*NimBot joined #nim
18:04:43*wildtrees joined #nim
18:05:05FromGitter<akavel> Hi! Is there a way to render Nim code from AST? In other words, to convert Nim AST "back" to Nim code?
18:05:22disrupteksomenimnode.repr
18:05:26*wildtrees quit (Max SendQ exceeded)
18:05:55FromGitter<akavel> But this will dump the tree-like representation, no?
18:06:01disrupteknah.
18:06:02*wildtrees joined #nim
18:06:16disruptekyou're thinking of treeRepr.
18:06:34FromGitter<akavel> ahhh, ok, awesome, I'll try it now...
18:07:01FromGitter<akavel> woohooo, yeah, thanks! :D :D <3 <3 <3
18:07:04FromGitter<akavel> cool :D
18:07:10disruptek:-)
18:07:26*wildtrees quit (Max SendQ exceeded)
18:08:11*wildtrees joined #nim
18:09:24*wildtrees quit (Max SendQ exceeded)
18:09:58ozHey, it looks like the Discord chat is out of sync with IRC and Gitter.
18:10:00*wildtrees joined #nim
18:11:54*wildtrees quit (Max SendQ exceeded)
18:12:42*wildtrees joined #nim
18:14:18*wildtrees quit (Max SendQ exceeded)
18:14:47FromGitter<zacharycarter> so is Godot nim dead?
18:14:54*wildtrees joined #nim
18:15:15Calinoulooks like it, maybe endragor stopped working on their project
18:16:06*wildtrees quit (Max SendQ exceeded)
18:16:16Araqit's not dead, but somebody needs to fork it so that PRs can be applied
18:16:20FromGitter<zacharycarter> ah
18:16:42Araqwe can add it to important_packages and ensure Nim doesn't break it
18:16:44*wildtrees joined #nim
18:17:00FromGitter<zacharycarter> meh
18:17:05FromGitter<zacharycarter> I was just curious
18:18:04*wildtrees quit (Max SendQ exceeded)
18:18:41*wildtrees joined #nim
18:19:36*wildtrees quit (Remote host closed the connection)
18:21:35FromGitter<mratsim> @akavel I use toStrLit
18:35:52*mal`` quit (Ping timeout: 245 seconds)
18:40:54*alexander92 quit (Ping timeout: 268 seconds)
18:42:09*sagax quit (Quit: Konversation terminated!)
18:42:13*laaron quit (Remote host closed the connection)
18:42:26*sagax joined #nim
18:43:25*mal`` joined #nim
18:45:22*laaron joined #nim
18:47:46*absolutejam4 joined #nim
19:14:20*Hideki_ joined #nim
19:19:32*Hideki_ quit (Ping timeout: 276 seconds)
19:33:50*laaron quit (Remote host closed the connection)
19:35:45*laaron joined #nim
19:38:42*actuallybatman joined #nim
19:46:41Zevvpff I hate this. I have this little macro helper proc that's clearly full of duplicated code, but still I can't get it any neater then this: http://ix.io/1UQH
19:47:03Zevvdoes anyone see the obvious thing I'm missing?
19:47:44Zevvit's not at all *important*, but it's just bugging me
19:53:43FromGitter<mratsim> foldIt?
19:54:18Zevvhmmm lemmesee
20:00:57FromGitter<brentp> what's wrong with this: ⏎ ⏎ ```type B* = distinct seq[uint64] ⏎ ⏎ proc `[]`*(x:B, i:SomeOrdinal): uint64 {.borrow.}``` ⏎ ⏎ ? [https://gitter.im/nim-lang/Nim?at=5d7952796e889c4bbddc8c0d]
20:01:22FromGitter<brentp> gives: ` Error: no symbol to borrow from found`
20:02:01FromGitter<brentp> `proc len*(x:B): int {.borrow.}` works as expected
20:04:16FromGitter<brentp> https://github.com/nim-lang/Nim/issues/3564
20:04:18Zevvfunny, you'd expect it to borrow from system.nim:1061
20:07:30FromGitter<Varriount> Zevv: I guess `n.kind == nnkInfix and n[0].eqIdent("|")` could be turned into a proc/template?
20:07:52FromGitter<Varriount> But really, it's fairly good looking, for macro code
20:08:08Zevv"for macro code" :)
20:08:15Zevvdifferent standards there of course, haha
20:08:26FromGitter<Varriount> And macro code usually looks like someone put a programming dictionary through a meat grinder.
20:09:11FromGitter<mratsim> @brentp borrow is broken with generics
20:10:19FromGitter<mratsim> https://github.com/nim-lang/Nim/issues/4121
20:10:48FromGitter<brentp> yeah. seems so.
20:10:50FromGitter<mratsim> macro code is like dom manipulation at the beginning of JS
20:11:02FromGitter<Varriount> Zevv: Case in point: The Nim compiler
20:11:19FromGitter<mratsim> we avoid distinct and ranges in our codebase at this point
20:11:25FromGitter<Varriount> Or really any programming language compiler
20:21:10*nsf quit (Quit: WeeChat 2.5)
20:27:36*Hideki_ joined #nim
20:32:31*Hideki_ quit (Ping timeout: 268 seconds)
20:37:31Araqdistinct and ranges are stable, *shrug*
20:37:44Araq.borrow is a bit weird but it was a bad idea to begin with
20:40:36*couven92 joined #nim
20:42:47*shashlick joined #nim
20:43:20Cadeyi'm getting "Error: unhandled exception: passNimFlags is not accessible [FieldError]" in https://github.com/Xe/within-terraform/tree/master/dyson with docker build -t xena/dyson:0.1.0 .
20:43:47*narimiran quit (Ping timeout: 245 seconds)
20:44:01Cadeyit gets that after failing to build the tempfile package
20:44:30shashlickDang keep getting disconnected
21:01:15AraqCadey: I dunno, but your exec "rm -rf " & folderName make me super nervous
21:02:26CadeyrmDir wasn't working the way i expected
21:06:24Araqit removes a directory plus its contents
21:06:35Araqjust like 'rm -rf'
21:06:58Cadeyi need to double check something, but it simply wasn't working
21:06:59Cadeysec
21:07:54Cadeynow it works like i expect
21:07:56Cadeywtf
21:08:11*Vladar quit (Remote host closed the connection)
21:11:57*ng0 quit (Quit: Alexa, when is the end of world?)
21:13:19disruptekprobably araq is crossing his fingers.
21:13:34Cadeythanks for catching that Araq
21:13:52Araqdisruptek: what do you mean?
21:14:51disrupteki mean you have a habit of making things work via spooky action at a distance.
21:16:16Araqthe machines feel my wrath
21:17:00disruptekcould be. slow code seems to run faster on my box, so i have some experience with this.
21:17:09Cadeythanks for your magic Araq
21:19:18Cadeyi need to figure out how to use nimscript embedded in a program
21:21:55Cadeynimscript in nimble is _so good_
21:23:08disruptekwhy?
21:24:29Cadeyit makes it simple to do basic building and deploying tasks in a way that's reasonably cross-platform
21:28:04Cadeyfor things like `npm install` or other errata
21:28:34Cadeynot to mention things like the version and other metadata are _actual variables_
21:29:17Araq:-)
21:29:53Cadeyit's first-class too
21:30:03Cadeyso it's _going to work_ without having to install extra stuff
21:30:26*btop joined #nim
21:30:47Cadeyi kinda wish the package story for nimble was a bit more robust than "hope people's git repos don't go down"
21:30:53Cadeybut that's a hard problem to solve in general
21:32:40dom96all it takes is for someone to create a service that forks all repos periodically
21:33:06dom96put it together with ipfs and you might not even need a server ;)
21:36:52*Hideki_ joined #nim
21:38:26Araqor: convince github that "delete repo" is an anti-feature
21:41:41*Hideki_ quit (Ping timeout: 265 seconds)
21:46:50*livcd quit (Remote host closed the connection)
21:48:23*solitudesf quit (Ping timeout: 276 seconds)
21:49:39*absolutejam4 quit (Ping timeout: 240 seconds)
21:50:55*absolutejam4 joined #nim
21:52:17*btop quit (Ping timeout: 276 seconds)
21:59:46*couven92 quit (Quit: Client Disconnecting)
22:12:06*absolutejam joined #nim
22:14:03*absolutejam4 quit (Ping timeout: 245 seconds)
22:20:54FromGitter<awr1> @mratsim fixing that issue is in fact a form of borrow checking
22:20:56FromGitter<awr1> :P
22:21:14FromGitter<mratsim> :D
22:21:39shashlick@Cadey - check out https://github.com/nim-lang/nimble/blob/master/src/nimblepkg/nimscriptwrapper.nim
22:21:51FromGitter<mratsim> @Araq they are stable but semcheck doesn’t like them
22:22:16*krux02 quit (Remote host closed the connection)
22:22:26*absolutejam1 joined #nim
22:23:23FromGitter<awr1> that reminds me i need to work on my nimscript RFC
22:23:48shashlickThat's just calling `nim e`, if you prefer pulling the whole compiler in check out https://github.com/Serenitor/embeddedNimScript
22:24:13FromGitter<awr1> my RFC concerns that
22:24:15*absolutejam quit (Ping timeout: 246 seconds)
22:24:18FromGitter<awr1> i haven't finished it though
22:24:55FromGitter<awr1> i think the situation with embedding nimscript into anapplication could be a bit better
22:28:48*thomasross_ joined #nim
22:28:48*thomasross quit (Killed (niven.freenode.net (Nickname regained by services)))
22:28:48*thomasross_ is now known as thomasross
22:36:22disrupteksometimes, when building my code which hashes a simple enum for use as a table key, i get this error: Error: unhandled exception: /home/adavidoff/git/Nim/lib/pure/collections/hashcommon.nim(33, 9) `length > counter` [AssertionError]
22:37:10disrupteki will dig deeper tomorrow. :-/
22:40:37shashlick@zacharycarter had done some work on embedded nimscript as well
22:42:03FromGitter<gogolxdong> How to generate a symbol in a proc to the top level of a module?
22:46:33*absolutejam1 quit (Ping timeout: 245 seconds)
22:48:43*solitudesf joined #nim
22:53:13*solitudesf quit (Ping timeout: 265 seconds)
23:03:34FromGitter<gogolxdong> generate a symbol and inject to the first level of the module
23:06:40FromGitter<gogolxdong> I'm trying to achieve this ⏎ ⏎ ```buildHtml tdiv:``` [https://gitter.im/nim-lang/Nim?at=5d797e0032e48c5df27bdbf1]
23:08:03*owl joined #nim
23:08:43FromGitter<gogolxdong> I'm trying to achieve this ⏎ ⏎ ```buildHtml tdiv: ⏎ input(value := localAreaName)``` ⏎ ⏎ := means declare a module level variable [https://gitter.im/nim-lang/Nim?at=5d797e7a460a6f5a1608d7ce]
23:16:23*fov60 joined #nim
23:17:59fov60is there a way to get fields as a list?
23:26:27*abm quit (Quit: Leaving)
23:33:39FromGitter<ShieldHero> it looks like the discord bridge is down since sunday, did anybody notice?
23:41:49FromGitter<zetashift> godot-nim could be forked but what's wrong with making a separate branch for it?
23:57:21*Tyresc quit (Quit: WeeChat 2.5-dev)
23:59:30FromGitter<mratsim> @fov60, you can iterate on the fields but you can’t get them as a list in a runtime proc, in a macro yes