<< 05-07-2017 >>
00:04:44*couven92 quit (Ping timeout: 260 seconds)
00:07:14*def-pri-pub joined #nim
00:26:02*yglukhov joined #nim
00:30:25*yglukhov quit (Ping timeout: 248 seconds)
00:40:10FromGitter<zacharycarter> @ephja is your glfw wrapper the most up to date one?
00:41:14*def-pri-pub quit (Quit: leaving)
01:20:51*aerx_ joined #nim
01:22:46*aerx quit (Ping timeout: 240 seconds)
01:42:22*jsgrant joined #nim
01:50:20*def-pri-pub joined #nim
01:54:33*chemist69 quit (Ping timeout: 255 seconds)
02:01:56*ipjk_ joined #nim
02:04:54*ipjk quit (Ping timeout: 255 seconds)
02:07:49*chemist69 joined #nim
02:21:49*Nobabs27 quit (Quit: Leaving)
02:45:28*ipjk_ quit (Quit: Leaving)
02:56:45*aerx joined #nim
02:58:27*aerx_ quit (Ping timeout: 240 seconds)
03:26:39*pilne quit (Quit: Quitting!)
04:02:15FromGitter<Araq> @indy9000_twitter what's wrong with the 'if' solution?
04:02:43*Vladar joined #nim
04:04:11FromGitter<Araq> we should support an outer 'if' though, so that you can write it outside the 'for' loop: if result.len >= src.len: for ...
04:06:21*rkjx joined #nim
04:10:46*rkjx quit (Client Quit)
04:11:46*stisa quit (Ping timeout: 276 seconds)
04:14:12*stisa joined #nim
04:16:15*rauss quit (Quit: WeeChat 1.8)
04:38:53*def-pri-pub quit (Quit: leaving)
05:12:09*salewski joined #nim
05:13:54salewskiIs there a good actual example for a non trivial nimble install script available? I think for
05:13:57salewskihttps://github.com/StefanSalewski/nim-gi
05:14:38salewskiI have to check if GTK libs are installed at all, and maybe also if so called developer files are installed?
05:15:40salewskiSo I may need a combination of nimscript and nimble? Or nawabs? Or nake?
05:16:29salewskiI have not too much confidence if that is really all explained in the Manning book.
05:17:19*Snircle quit (Quit: Textual IRC Client: www.textualapp.com)
05:18:36FromGitter<Araq> :-) nimble files are nimscript files (and nawabs uses nimble files too)
05:20:00FromGitter<Araq> checking for gtk libs should be done like so https://github.com/nim-lang/Nimble#external-dependencies
05:20:18salewskiAraq, do you know a good example? Trustables NiGUI may be good, but I think he does not still have one.
05:21:28salewskiOK, will read the external-dependencies section...
05:25:02salewskiBye.
05:25:06*salewski quit (Quit: WeeChat 1.4)
05:44:35*jsgrant quit (Ping timeout: 240 seconds)
06:14:21*tankfeeder joined #nim
06:38:53*Vladar quit (Remote host closed the connection)
06:40:14*nsf joined #nim
06:51:56*arnetheduck joined #nim
07:00:48*aerx quit (Quit: WeeChat 1.9)
07:19:12*Arrrr joined #nim
07:19:12*Arrrr quit (Changing host)
07:19:12*Arrrr joined #nim
07:30:54*Vladar joined #nim
07:31:12*nattefrost joined #nim
07:47:26*yglukhov joined #nim
07:49:32*yglukhov_ joined #nim
07:51:52*yglukhov quit (Ping timeout: 260 seconds)
07:53:24*couven92 joined #nim
07:53:37*yglukhov_ quit (Ping timeout: 248 seconds)
08:25:40*sz0 quit (Quit: Connection closed for inactivity)
08:29:13*yglukhov joined #nim
08:34:29tankfeederi have array[16, uint8]
08:34:53tankfeederhow to get uint64 from last 8 numbers ?
08:37:49flyxtankfeeder: cast[ptr uint64](addr(myarr[8]))
08:38:35*ShalokShalom joined #nim
08:40:44tankfeederok
08:55:50*gangstacat quit (Quit: Ĝis!)
09:09:07*v17d joined #nim
09:10:16*skrylar joined #nim
09:10:28skrylarBook
09:10:38skrylarBoop
09:10:40skrylarBleh
09:11:55FromGitter<Araq> hey
09:11:57*chemist69 quit (Ping timeout: 255 seconds)
09:12:18*chemist69 joined #nim
09:14:02*skrylar quit (Remote host closed the connection)
09:15:20dave24If I have toplevel code in modules A and B where A imports B, will the code in B always execute first?
09:16:43*yglukhov quit (Remote host closed the connection)
09:17:09FromGitter<Araq> yes
09:18:58dave24Araq: Ah, ok Thanks!
09:22:14*skrylar joined #nim
09:22:33skrylarAraq when did you forsake your IRC client :e
09:23:38*gangstacat joined #nim
09:25:18*Matthias247 joined #nim
09:25:27FromGitter<Araq> just for the time being as I'm on a different machine
09:26:49*xet7 quit (Remote host closed the connection)
09:28:14*yglukhov joined #nim
09:29:57skrylarAh well I looked in to libui and started making a few templates
09:39:52*Matthias247 quit (Read error: Connection reset by peer)
09:40:07FromGitter<Araq> great :-) I'm thinking about merging karax and libui somehow
09:42:16skrylarWhat is karax
09:44:11couven92@Araq what about wxWidgets?
09:47:15FromGitter<Araq> karax is a single page application library for Nim
09:49:52skrylarDidn't someone do wx?
09:50:01skrylarKinda bloated tho
09:50:12couven92skrylar, PMunch and I worked on it some time ago...
09:51:35*bjz joined #nim
09:51:52FromGitter<Araq> me too :-)
09:52:20couven92yeah, PMunch and I worked on your original port...
09:52:49FromGitter<Araq> I'm aware ;-) (it's good you didn't start from scratch)
09:53:03*yglukhov quit (Remote host closed the connection)
09:53:29FromGitter<Araq> for karax to work out, I likely need a full-blown browser component though :-( any idea of how well this works in wxWidgets?
09:54:36*arnetheduck quit (Ping timeout: 260 seconds)
09:55:06couven92@Araq, I don't know in wxWidgets, but in C# we use Eto.Forms for xplat GUI, and it has a Browser component... So it suggests that all major native GUI frameworks have a browser compontent that we could bind against...
09:55:24skrylarEh mixing single page with a desktop app seems silly. But there is libCEF
09:55:46FromGitter<TiberiumN> there's CEF for nim
09:55:48*yglukhov joined #nim
09:56:00skrylarWindows has I.E. over COM, mac has WebKit
09:56:02FromGitter<TiberiumN> https://github.com/jangko/nimCEF
09:56:18skrylarLinux has migraines :>
09:56:44couven92PMunch and I concluded that we should have sth. like wxWidgets but written natively in Nim... because wxWidgets is xplat it wraps EVERYTHING into its own type, which is annoying, since Nim can do that on its own anyways
09:56:56FromGitter<Araq> yay thanks for that. wasn't aware of nimCEF
09:57:32FromGitter<Araq> couven92: I came to the same conclusion but thought to port libui slowly from C to Nim as it's much less code than wxWidgets
09:57:51couven92Hmm... yeah sounds reasonable
09:58:01*adeohluwa joined #nim
09:58:14FromGitter<TiberiumN> couven92: also NiGui tries to do that, but it tries to be something similar to libui
09:58:15couven92What's the catch though? Does libui support a lot?
09:58:37FromGitter<Araq> libui's event handling is insufficient for nimEdit
09:58:58FromGitter<Araq> and after I realized that I lost interest in it
09:59:54FromGitter<Araq> note: SDL2's events were sufficient for nimEdit
10:00:00*yglukhov quit (Remote host closed the connection)
10:00:01FromGitter<Araq> that's quite remarkable
10:01:43FromGitter<Araq> but I love SDL, so I'm biased.
10:06:23couven92@Araq, btw, I managed to strip jquery and bootstrap in such a way that they can be inlined in the HTML header... So I could make a testament output... This means that the testament output includes source from jquery and bootstrap, but both are released under MIT license, so it should be fine from a legal point of view, right? And the overhead is minimal, it takes far more text to actually write out 1500 testament results in comparison to including
10:06:23couven92bootstrap and jquery...
10:08:31couven92I know you said to leave it be and don't bother, but I wanted to experiment with bootstrap anyways, so I just did it for fun regardless... So, I can do a PR for testament when I come this evening, so we can discuss it further on GitHub?
10:08:56FromGitter<Araq> how much JS/CSS lines are we talking about?
10:09:14couven92Not much in lines, since it's minified
10:09:51couven92but... let me check for byte size...
10:11:29FromGitter<Araq> oh my, minified?
10:11:35skrylarLibui event handling looks primitive but workable I guess
10:11:45FromGitter<Araq> how am I gonna review it then?
10:12:11couven92@Araq, it the exact copy from getbootstrap.com
10:12:39couven92so no actual changes, except for inlining the glyphicons font, instead of having url links
10:13:09couven92I hate JS as much as you (as CSS, too), so no, I am not gonna write some custom JS by hand! :P
10:16:27couven92so... if I calculated correctly now, Bootstrap+jQuery together is about 266KiB, and the HTML for displaying ALL 1500 testament results (excluding the bootstrap and jquery overhead) is 3.6MiB
10:16:38couven92so that's a ~7% overhead
10:17:18skrylarTime to go insolvent
10:17:22*skrylar quit (Quit: Leaving)
10:17:42*yglukhov joined #nim
10:17:44couven92PMunch mentioned Skeleton as a WebGUI framework with much less overhead... I could take a lot at that instead
10:18:19FromGitter<Araq> well feel free to create a PR, but I'm not gonna merge stuff I can't review.
10:19:09yglukhovAraq: could you pls take a look at https://github.com/nim-lang/Nim/issues/6055.
10:19:45yglukhovdo i understand correctly the tyPtr get sfPure on it, but on semType it should propagate sfPure to its child, but it doesnt?
10:20:39FromGitter<Araq> it never did that iirc
10:20:56FromGitter<Araq> workaround: Explicitly named 'object' and then create a ptr to it?
10:21:25FromGitter<Araq> you are not talking about the JS backend are you?
10:21:34FromGitter<Araq> cause I changed that recently for JS
10:22:29yglukhovAraq: hrrm. workaround doesn't work for some reason. m_type still emitted.
10:22:35yglukhovthats C
10:25:25FromGitter<Araq> type ⏎ B {.pure, inheritable.} = object ⏎ ⏎ ```a: int``` ⏎ ... [https://gitter.im/nim-lang/Nim?at=595cbe956ae41d5d33b99ce7]
10:25:29FromGitter<Araq> works for me
10:26:04FromGitter<Araq> as does
10:27:02FromGitter<Araq> A {.pure, inheritable.} = ptr object
10:27:15yglukhovAraq: ouch. i misplaced pragma :(
10:27:29yglukhovA = ptr object {.pure, inheritable.}
10:27:46FromGitter<Araq> yeah well, your variant needs to work too, so it's a valid issue
10:27:51FromGitter<Araq> just not a regression ;-)
10:29:29FromGitter<Araq> could be easy to fix, want to give it a shot?
10:30:17yglukhovwhy should Nim allow two nearly equivalent ways for doing the same thing?
10:30:52yglukhovmaybe my way should be just forbidden?
10:32:10FromGitter<Araq> I like your way better. less important stuff like pragmas should come later
10:39:45ArrrrWhy not the java way:
10:39:46Arrrr{.pure, inheritable.}
10:39:46ArrrrA = ptr object
10:42:43FromGitter<Araq> because that's worse. essential you are agreeing with me. the prefix is so hard to read that you put it into its own line
10:43:21FromGitter<Araq> it's like C's
10:43:23FromGitter<Araq> int
10:43:34FromGitter<Araq> foo(int a, int b) { ... }
10:44:11ArrrrBut consider big examples. Usually, a proc with a pragma, its declaration, will take two lines anyway
10:44:16ArrrrBut in a ugly way
10:45:00FromGitter<Araq> I disagree.
10:45:21Arrrrhttps://pastebin.com/303ijzKv
10:45:59FromGitter<Araq> I don't write code that way
10:46:15FromGitter<Araq> and I doubt there is a single line in the stdlib that looks like that.
10:46:48Arrrrhttps://github.com/nim-lang/Nim/blob/master/lib/system.nim#L1199-L1200
10:46:57ArrrrHave at you
10:47:12ArrrrI'd say there are a LOT of procs defined that way
10:47:28FromGitter<Araq> that is not the style you have shown
10:47:44ArrrrThat's only 1% different
10:47:49FromGitter<Araq> proc `@` * IDX, T (a: array[IDX, T]): seq[T] ⏎ {.magic: "ArrToSeq", nosideeffect.}
10:48:11FromGitter<Araq> also possible but came later :P
10:48:26ArrrrAnd some of these definitions takes >80 chars
10:48:57FromGitter<Araq> so?
10:49:41ArrrrIt goes against that rule that says no line can contain more than 80 characters
10:49:45ArrrrWhat about https://pastebin.com/jSv5rbn0
10:49:58FromGitter<Araq> you want prefix pragmas instead of suffix pragmas. 80 char limit is irrelevant
10:50:28ArrrrNo no, i don't want them in the same line as the proc/type definition
10:50:29FromGitter<Araq> what about it? I suggested that myself to Varriount multiple times
10:51:07FromGitter<Araq> it hasn't caught on because it's complex with forwarding / calling conventions
10:51:42FromGitter<Araq> proc foo = ⏎ {.cdecl.} # not an implementation detail, affects typing rules
10:52:13FromGitter<Araq> we could make '=' optional instead though
10:52:18FromGitter<Araq> proc foo
10:52:28FromGitter<Araq> {.cdecl.}
10:52:37FromGitter<Araq> other_body_stuff()
10:53:12FromGitter<Araq> stupid gitter loses my indentations
10:54:19FromGitter<Araq> the idea is that .} also acts as a '='
10:54:45*bjz quit (Quit: Textual IRC Client: www.textualapp.com)
10:55:06ArrrrBut it won't make the line look less polluted
10:55:35FromGitter<Araq> the line is already only pragmas (plus sometimes '=' )
10:55:55FromGitter<Araq> now you seem to change the topic again and complain about the pragma syntax instead?
10:56:48*bjz joined #nim
10:57:25ArrrrNo, the subject is still one line procs + pragmas looks ugly. And i don't know how replacing = with .} fixes that
10:59:18FromGitter<Araq> well that fixes what I consider ugly about it
10:59:51FromGitter<Araq> I still don't understand what you consider ugly about it
11:00:11ArrrrThis is what i understand:
11:01:06Arrrrhttps://pastebin.com/JyefKk4T
11:02:18FromGitter<Araq> https://pastebin.com/A7tsUers is what I meant
11:03:38FromGitter<Araq> 'before' works with today's Nim, try it
11:05:00ArrrrMmm, well, doesn't look that bad.
11:05:23*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
11:06:10FromGitter<Araq> yeah in fact, I don't see the need to remove the '=' here. it's ok.
11:06:13*skrylar joined #nim
11:23:42skrylarAraq what was the problem with libui's event dispatch?
11:24:39FromGitter<Araq> I didn't find a way to receive "key pressed event, corresponding unicode key: ..."
11:24:56skrylari did some hinky bs with rawEnv and rawProc that seems to work, and then it seems to work ok with a template over do notation. so "window.on_closing do (w: Window) -> cint: ..." works and does what i expect. although it does lead to either 1) wrapping a second closure to convert bool to cint or 2) trampoline to a signals/slots setup
11:25:42FromGitter<Araq> for wxWidgets? nice
11:25:46skrylarfor libui
11:26:00skrylarhaven't tried wrapping c++ or obj-c
11:26:04FromGitter<Araq> hmm? my libui wrapper supports closures already
11:26:17skrylarprobably
11:26:22skrylardo notation is nice there though
11:26:35FromGitter<Araq> if you say so
11:26:49skrylarmight steal a page from Red and poke a macro
11:27:07skrylarwhole thing about red's VID is you just write out the menu bar or something like normal and it boilerplates it for you
11:27:43FromGitter<Araq> the ugliness creeps in in the event handlers. the menu needs to do stuff
11:28:05skrylarideally should be using command model
11:28:18FromGitter<Araq> I don't know what that is
11:28:20skrylaractions go in command, buttons/menus are linked to said command. makes things clean, centralizes the enable/disable code
11:28:36skrylarqt does this for toolbars and some widgets but its not universal
11:28:51FromGitter<Araq> sounds like a reinvention of a 'proc' that you can call from multiple places
11:29:01skrylarit is but GUIs need more than that
11:29:07skrylar"can this button be pressed right now?"
11:29:18FromGitter<Araq> nilable proc :P
11:29:27*dexterk_ quit (Ping timeout: 240 seconds)
11:29:34skrylarok
11:29:42skrylarand the translation tag?
11:29:45skrylarand the undo step?
11:29:55*bjz joined #nim
11:30:33FromGitter<Araq> what about it?
11:31:24FromGitter<Araq> I click "file -> open" . what's there to undo? what's there to translate?
11:31:47*yglukhov quit (Remote host closed the connection)
11:31:58FromGitter<Araq> (I don't think you are talking about i18n, are you?)
11:32:49skrylari'm just talking about my experiences with desktop apps /shrug
11:33:25skrylarif you put all the stuff in command objects you can centralize a lot of stuff that big production apps do (namely REAPER) and such
11:34:26*yglukhov joined #nim
11:40:58FromGitter<Araq> sounds to me like the menu click handler is pretending to adhere to overdesigned interface
11:42:02FromGitter<Araq> but of course, since you gave it a noun it's abstract, powerful and scales to big apps :P
12:02:26*Snircle joined #nim
12:03:25*bjz_ joined #nim
12:05:33*bjz quit (Ping timeout: 268 seconds)
12:10:50*xet7 joined #nim
12:12:58skrylararaq nah it's just not useful for smaller apps
12:13:28skrylarthere's a good reason for the craziness but it only applies to *big* stuff.
12:14:20FromGitter<Araq> yes, big stuff requires interface lies, got it.
12:15:26skrylarshrugs if you say so
12:17:53FromGitter<ephja> @zacharycarter it's the most high level one, but it's not at 3.2 yet. what do you need?
12:37:48*adeohluwa quit (Quit: Connection closed for inactivity)
12:41:38*yglukhov quit (Remote host closed the connection)
12:46:10FromGitter<zacharycarter> @ephja hrm probably most of the 3.2 functionality
12:47:31FromGitter<ephja> @zacharycarter vulkan contexts?
12:47:46FromGitter<zacharycarter> probably not, I'm not playing with Vulkan quite yet
12:49:58FromGitter<zacharycarter> I've kind of been stuck in a rut as to how to proceed with my game / engine project
12:50:04FromGitter<zacharycarter> I'm thinking about giving CommonLisp a try
12:50:11FromGitter<zacharycarter> just to give my mind a break from Nim for a bit
12:50:53FromGitter<zacharycarter> I don't think FRAG is very usable in its current state, it has too many dependencies that people just getting into the project have a tough time dealing iwth
12:50:55FromGitter<zacharycarter> with*
12:51:04federico3yep
12:53:23*Matthias247 joined #nim
13:09:53*yglukhov joined #nim
13:10:36shmupdid you compare any other lisps before thinkin CL, zackharycarter?
13:14:11tankfeederzacharycarter, try picolisp
13:25:28FromGitter<zacharycarter> I did but CL seems to be pretty popular amongst the game dev community
13:26:38ArrrrYou should and then report your finds.
13:26:49FromGitter<zacharycarter> I will for sure
13:29:21*Vladar quit (Quit: Leaving)
13:33:28*Arrrr quit (Read error: Connection reset by peer)
13:35:41*nsf quit (Quit: WeeChat 1.7.1)
13:41:33*Snircle quit (Quit: Textual IRC Client: www.textualapp.com)
13:42:02*sz0 joined #nim
13:47:35*svara quit (Read error: Connection reset by peer)
14:03:51*xet7 quit (Remote host closed the connection)
14:06:31*xet7 joined #nim
14:09:58*nattefrost quit (Remote host closed the connection)
14:14:09*rauss joined #nim
14:32:22*rauss quit (Quit: WeeChat 1.8)
14:35:16*skrylar quit (Quit: My iMac has gone to sleep. ZZZzzz…)
14:35:18*jsgrant joined #nim
14:45:08*tankfeeder quit (Quit: Leaving)
15:00:02*ofelas joined #nim
15:01:19*oaao joined #nim
15:04:13*couven92 quit (Quit: Client disconnecting)
15:10:02*Trustable joined #nim
15:14:36*Arrrr joined #nim
15:19:56*nsf joined #nim
15:23:36*shashlick joined #nim
15:29:22*yglukhov quit (Remote host closed the connection)
15:37:54SentreenStupid question: Is there a proc that can resize (grow/shrink) seqs? The closest thing I can find is `add`
15:38:17FromGitter<zacharycarter> setLen
15:38:34SentreenThanks :)
15:39:00*yglukhov joined #nim
15:43:04*yglukhov quit (Ping timeout: 246 seconds)
15:48:50*aerx joined #nim
15:57:00shmuphttps://github.com/nim-lang/Nim/pull/6048#issuecomment-313146456
16:04:30FromGitter<Varriount> shmup: What happens when you run tests locally?
16:05:55shmupso this is new to me Varriount, but am I doing more than: koch test, from Nim root
16:07:16FromGitter<Varriount> That should run the tests - is the test from Travis failing for you locally?
16:08:00shmupI couldn't say from this output alone, unless it would still tell me short of loading uh.. tester.exe https://gist.githubusercontent.com/shmup/d08bf994448bb10636ad8167d64c5395/raw/34ddaf185d48c7a200e3945e74cebef716c3ac16/gistfile1.txt
16:11:06*vivus joined #nim
16:13:59FromGitter<Varriount> Ah, you need sqlite
16:14:50FromGitter<Varriount> shmup: I wonder if this is some weird bug.
16:15:35FromGitter<Varriount> shmup: Could you try making a small change/resubmitting the PR, to get Appveyor and Travis to rebuild?
16:19:08*shashlick quit (Ping timeout: 260 seconds)
16:28:10*shashlick joined #nim
16:28:45*yglukhov joined #nim
16:29:02shmupsure Varriount, in meantime I grabbed binary for sqlite3_32 and prec32.dll and now it doesn't error and just sits on C:\Users\jared.miller\git\nim\Nim\tests\testament\tester.exe all not launching/showing anything
16:33:15*yglukhov quit (Ping timeout: 240 seconds)
16:34:21*Snircle joined #nim
16:34:33shmuphere goes nothin
16:40:40*nsf quit (Quit: WeeChat 1.7.1)
16:50:05shmupjobs are still running but it is going to fail again, Varriount https://travis-ci.org/nim-lang/Nim/jobs/250427598#L2562
16:50:14shmupFUN4~
16:55:07*oaao quit (Read error: Connection reset by peer)
16:55:20*oaao joined #nim
16:56:20*oaao1 joined #nim
16:58:42*oaao quit (Read error: Connection reset by peer)
16:58:49*arecacea1 quit (Remote host closed the connection)
16:58:52FromGitter<ephja> Sentreen: You could sort the system module documentation by type and then go to the 'seq' section
16:59:26*arecacea1 joined #nim
17:10:56*onionhammer1 quit (Quit: WeeChat 1.0.1)
17:26:13*Trustable quit (Remote host closed the connection)
17:57:04koppehHey guys.
17:57:50koppehSo I'm on Windows and using MinGW + GCC. Kind of new to this.
17:58:25koppehCan't find GL/glu.h. My friend is saying I should add the headers to the include path.
17:59:00koppehC_INCLUDE_PATH environment variable I suppose should do..? But can I set that in nim.cfg or so? Or is there a more proper way?
18:02:08FromGitter<ephja> why is that header necessary?
18:03:50*Sentreen quit (Ping timeout: 268 seconds)
18:04:04koppehFor glew?
18:06:36FromGitter<Varriount> koppeh: You might verify that GL/glu.h is in your include directory first
18:06:37FromGitter<ephja> oh. right
18:08:36*pilne joined #nim
18:08:51koppehRight.. how would I check that, again?
18:10:53koppehI couldn't find anything in the choosenim mingw folder.
18:11:04koppehUnsure if that matters.
18:12:53koppehAnd `gcc --print-search-dirs` appears to only print directories within the mingw dir.
18:14:02koppehExcept.. it also searches in a folder that doesn't exist.
18:14:21FromGitter<ephja> then it's simply ignored
18:14:25*v17d quit (Ping timeout: 248 seconds)
18:14:31FromGitter<ephja> I would put the headers relative to one of the existing dirs
18:14:46koppehBut that means there might be a way to specify the windows SDK include folder.
18:17:01FromGitter<ephja> putEnv:"..." in nim.cfg?
18:17:25*jsgrant quit (Ping timeout: 276 seconds)
18:17:27*Sentreen joined #nim
18:18:55FromGitter<ephja> do you want to use GLEW in order to implement extension fallbacks?
18:19:36koppehI'm not quite sure. That part is being done by my friend.
18:19:51FromGitter<Varriount> koppeh: Windows has a search function?
18:20:07FromGitter<Varriount> Or if you have the right tools installed, the 'find' utility works too.
18:21:04koppehI'm in the process of searching.
18:21:31koppehThough what do you want me to search?
18:27:47FromGitter<zacharycarter> currently writing VUEjs tests or work
18:27:49FromGitter<zacharycarter> for work *
18:27:58FromGitter<zacharycarter> r= O<
18:28:22*tankfeeder joined #nim
18:31:13FromGitter<gokr> @couven92 Did you look at NiGui? libui seems to me to be stalled
18:40:33koppehgcc.options.linker="-I\"C:\\Program Files (x86)\\Windows Kits\\8.1\\Include\\um\""
18:40:51koppehI managed to do this, and it appears to be passed to gcc, but no effect..?
18:44:45*Sentreen quit (Ping timeout: 240 seconds)
18:47:37FromGitter<ephja> gcc.options.always?
18:50:31koppehI made it work, I think!
18:50:39koppeh--cincludes: "C:\\Program Files (x86)\\Windows Kits\\8.1\\Include\\um"
18:50:53koppehBut now it's missing kernelspecs.h??
18:53:26koppehOkay got that fixed by including /shared on top of /um, but...
18:53:34koppehNow I have all the errors.
18:54:39*ShalokShalom_ joined #nim
18:55:03koppehLike, all of the errors. More than my terminal has buffer for.
18:56:04*arnetheduck joined #nim
18:57:06koppehIn stdlib_system and stdlib_dynlib..?
18:57:27*Sentreen joined #nim
18:58:06*ShalokShalom quit (Ping timeout: 268 seconds)
19:00:57*nsf joined #nim
19:02:20*couven92 joined #nim
19:03:35*cyraxjoe quit (Ping timeout: 255 seconds)
19:04:19*cyraxjoe joined #nim
19:11:38FromGitter<ephja> I wonder if it will work with the windows headers if you are going to use GCC
19:12:30FromGitter<ephja> I need to implement the missing features so that glew won't ever be needed
19:14:39*yglukhov joined #nim
19:19:04*Arrrr quit (Quit: Leaving.)
19:21:14*salewski joined #nim
19:21:57salewskiDom, in https://github.com/nim-lang/nimble we have:
19:22:12salewskiSimilar to the install command you can specify a version range, for example:
19:22:30salewski$ nimble uninstall [email protected]
19:23:13salewskiFirst, I would not call the @0.5 a range, and second
19:24:05salewskiwhen we can uninstall a version range, than we can install multiple version of packages at the same time?
19:24:32salewskiAnd how do we use multple versions?
19:25:21FromGitter<indy9000_twitter> @Araq > we should support an outer 'if' though, so that you can write it outside the 'for' loop: if result.len >= src.len: for ... ⏎ ⏎ because that would be too much noise. Ideal is when the compiler can check for matching lengths implicitly using its type system.
19:25:43salewskiA few lines above you wrote:
19:25:49salewskiA branch, tag, or commit hash may also be specified in the place of head.
19:26:30salewskiWould be great if you can gave an example for that 3 cases -- for people like me with limited git knowledge.
19:30:47salewskiAnd, while I am still connected, a few lines above you wrote:
19:30:51salewskiversion in the git (or hg) repository, if the
19:31:41salewskiIt may be not obviuosly to every one that hg refers to Mercury. And the colon should be a period.
19:31:58salewskiBye, will continue reading...
19:40:12shmupthese are good criticisms but will possibly be lost here. open an Issue
19:58:07*Matthias247 quit (Ping timeout: 260 seconds)
19:58:12salewskiHave thought about the multiple installed versions, but do not get it. Docs tells us:
19:58:22*Matthias247 joined #nim
19:58:41salewskiSince there can be many versions of the same package installed, the path command will always show the latest version.
19:58:59salewski$ nimble path argument_parser
19:59:14salewski/home/user/.nimble/pkgs/argument_parser-0.1.2
19:59:48dom96salewski: when you want me to see a message make sure to write my full nickname, not just "dom"
19:59:48salewskiFirst I was thinking that multiple versions may be installable only for executables.
20:00:41salewskiOh yes, always forgett about that numbers :-)
20:00:53dom96salewski: also, PRs welcome.
20:01:33salewskiSo when I would install multiple versions of argument_parser, how would I import a specific one?
20:02:39salewskiWell, I just try to understand nimble... Read the nimble chapter in your book a year ago, but very fast.
20:04:16salewskiAnd as Amazon.de show september for release date, I thought starting with only docu.
20:04:27salewskiBye.
20:05:09salewskis/only/online
20:07:38salewskiOh, amazon.de just changed it: 30 July 2017. Great!
20:08:25salewskiWas september some days ago. So maybe I should read it better in the book then.
20:15:24FromGitter<Varriount> koppeh: You shouldn't mix Visual Studio headers with Mingw, you'll give it a stomachache
20:19:14dom96salewski: You specify the version in your nimble file
20:19:18dom96then compile using Nimble
20:19:22dom96'nimble c file'
20:22:58salewskidom96, the package developer specifies the version in the nimble file, not the user?
20:23:07dom96yes
20:23:30salewskiMy question was about the users. What I understood user can install multiple version of
20:23:52salewskifor example argument_parser. Do I understand that wrong?
20:25:28salewski(All what I read still now was above section "creating pakages")
20:26:15dom96yes, they can.
20:26:32dom96But when they're developing their own packages, they should have their own nimble package...
20:26:46dom96and thus able to choose the version inside the package's nimble file
20:29:29salewskiUnderstanding the github docu is not very easy for beginners. Under section "nimble path":
20:29:58salewskiSince there can be many versions of the same package installed,
20:30:39salewskiThat confused me a bit. But well, generally we want to use the latest version only.
20:31:02dom96what's confusing about it?
20:31:24salewskiWell, I will see if I can create a nimble file for nim-gi. It is not easy.
20:31:35dom96why not?
20:32:29FromGitter<ephja> must every client generate some of those things for every installation?
20:32:45salewskiWell, I gave the manual instructions at https://github.com/StefanSalewski/nim-gi
20:32:51FromGitter<ephja> in any case, I think that can be automated with nimscript now
20:33:33salewskiand now I have to put it into nimble, with additional checking if GTK dev files exists.
20:34:37FromGitter<ephja> can't the output files be part of the package? and then you re-generate those when necessary
20:34:44salewskiBut I will try, bye.
20:36:55*ofelas quit (Quit: shutdown -h now)
20:37:04*shashlick quit (Ping timeout: 246 seconds)
20:43:24*Jesin joined #nim
20:47:45salewskiephja, on my box gobject-introspection generates modules for 64 bit Linux with GTK 3.22 by default.
20:48:21salewskiI think I can generate files for olnder GTK releases. But not for 32 bit on this box.
20:48:29*Ven joined #nim
20:48:52*Ven is now known as Guest87924
20:49:09salewskiAnd not for windows or mac at all. So I think shipping prebuilt files is not a good idea.
20:49:45salewskiLater, when we have many users, they may be able to prowide tested files, which we may ship then.
20:56:56salewskiGood night.
20:56:59*salewski quit (Quit: WeeChat 1.4)
20:58:31*tankfeeder quit (Quit: Leaving)
20:59:52*xet7 quit (Quit: Leaving)
21:06:17*shashlick joined #nim
21:16:47*ShalokShalom_ quit (Remote host closed the connection)
21:16:57*Guest87924 quit (Read error: Connection reset by peer)
21:21:28*Ven_ joined #nim
21:33:26FromGitter<mratsim> How to build nimble in TravisCI? I use the following script but I get nimble file not found: https://github.com/mratsim/Arraymancer/blob/2d6a0c42112520d52e81073af5dc0e8ded1efdaa/.travis.yml (I just added ./koch nimble)
21:33:32*Ven_ quit (Read error: Connection reset by peer)
21:35:47*Ven_ joined #nim
21:46:16*shashlick quit (Ping timeout: 260 seconds)
21:47:08dom96mratsim: might wish to just try using choosenim at this point.
21:52:15*shashlick joined #nim
21:52:49subsetparkHow can I write a template, designed to be called inside an async proc, so that I don't get an `undeclared identifier` error when using the `await` keyword?
21:54:46dom96You can't.
21:54:56dom96You can use 'yield' instead of 'await' as a workaround
21:55:02dom96But it won't work the same as an 'await'
21:55:14dom96(it won't read the future, only yield it)
21:55:44subsetparkin this case I think that's ok, since I just am using it to kick control back to the event loop
21:55:48subsetparkRight?
21:57:21subsetparkhttps://www.irccloud.com/pastebin/41ADG9Iv/
21:59:48*onionhammer joined #nim
22:04:01subsetparkno, that seems to block...
22:05:49*rauss joined #nim
22:06:00subsetparkAnythoughts dom96 ?
22:07:06dom96that should work
22:07:30dom96But if `sleepAsync` fails the error will silently disappear
22:07:45*nsf quit (Quit: WeeChat 1.7.1)
22:09:36*rauss quit (Client Quit)
22:10:11subsetparkNo, it just loops in the `while true`
22:10:41subsetparkThe same code works fine when I don't have a template, if I write out the structure directly in the jester handler
22:13:16subsetparkhm, interesting - it's hitting the "isReady" block - it seems that it's there that it isn't yielding
22:13:36subsetparkDo I need to put an explicit yield there too, after body? The body block contains a `resp`
22:14:30subsetparkYeah, looks like putting a `break` after body makes that work
22:24:27*bjz_ quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
22:28:15*Jesin quit (Quit: Leaving)
22:31:05subsetparkI wonder why I need the break. the `resp`s are still in the `async` proc, so I would think they would still kick it out.
22:31:31subsetparkBut maybe going through the template first means the async can't do its magic...
22:32:59shashlickwhat's the main branch on github? Is it devel or master?
22:33:24subsetparki think master is the most recent release
22:36:31shashlickso pull requests should be made on devel?
22:42:54FromGitter<ephja> yes
22:49:21*jsgrant joined #nim
22:52:27*shashlick quit (Ping timeout: 240 seconds)
22:53:19*Jesin joined #nim
23:00:09*shashlick joined #nim
23:05:58*Ven_ quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
23:24:04*vivus quit (Quit: Leaving)
23:26:53*handlex joined #nim
23:36:18*Matthias247 quit (Read error: Connection reset by peer)
23:48:15*arnetheduck quit (Ping timeout: 240 seconds)
23:48:45*handlex quit (Quit: handlex)
23:50:44*handlex joined #nim
23:53:37*jsgrant quit (Ping timeout: 248 seconds)
23:57:19*ipjk joined #nim