<< 02-08-2019 >>
00:08:47*krux02 quit (Remote host closed the connection)
02:09:23FromGitter<Varriount> Not unless memory corruption is going on, or a library is unloaded
02:16:55*dddddd quit (Remote host closed the connection)
02:18:59shashlicki think loading nim dlls has finally caught up with me
02:19:30shashlicknone of the GCs help and nimRtl doesn't work either with threads
02:24:15*abm quit (Quit: Leaving)
02:36:17*laaron joined #nim
02:36:20skrylar[m]finally managed the gdk stuff, still needs testing though
02:42:31*r3c quit (Ping timeout: 260 seconds)
02:57:24*dgreen quit (Remote host closed the connection)
03:00:32*laaron quit (Remote host closed the connection)
03:01:06*laaron joined #nim
03:44:10*stefanos82 joined #nim
04:14:05*fjellfras joined #nim
04:15:14*altarrel_ quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)
04:19:42*snifftek joined #nim
04:33:28*nsf joined #nim
04:41:37*tiorock joined #nim
04:41:43*tiorock quit (Changing host)
04:41:43*tiorock joined #nim
04:41:43*rockcavera is now known as Guest24310
04:41:43*tiorock is now known as rockcavera
04:43:09*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
04:45:14*Guest24310 quit (Ping timeout: 272 seconds)
04:47:19*laaron joined #nim
05:28:18*solitudesf joined #nim
05:31:05*laaron quit (Remote host closed the connection)
05:34:28FromGitter<kayabaNerve> Should % nil be invalid?
05:35:08FromGitter<kayabaNerve> The JSON standard lib doesn't convert it to a JNull. It adds a literal null element which breaks if you try to print.
05:35:37FromGitter<kayabaNerve> Side note, what is up with nil, and especially nil/null fragmentation?
05:40:33*laaron joined #nim
05:52:22*rockcavera quit (Ping timeout: 272 seconds)
05:53:47*solitudesf quit (Ping timeout: 245 seconds)
06:07:58*nullnullnull quit (Quit: Leaving)
06:13:48FromGitter<Varriount> Probably an oversight
06:27:21*rockcavera joined #nim
06:27:22*rockcavera quit (Changing host)
06:27:22*rockcavera joined #nim
06:33:06*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
06:33:43*laaron joined #nim
06:41:05*solitudesf joined #nim
07:00:00*gmpreussner quit (Quit: kthxbye)
07:04:29*gmpreussner joined #nim
07:06:52*solitudesf quit (Ping timeout: 244 seconds)
07:35:08*krux02 joined #nim
07:53:10FromGitter<kayabaNerve> And I caused a compiler crash again :/ Turns out my async code is problemat ic.
07:53:19FromGitter<kayabaNerve> Or async is. One of the two.
08:13:43*floppydh joined #nim
08:14:41dom96what's your code?
08:19:41Mister_Magisterguys i want to make small webserver using nim. Are there some libs for receiving/sending http request and/or html template engine?
08:20:31leorizeasynchttpserver is a simple http server that you might like
08:20:55leorizethere's also jester which is a bit more advanced
08:22:32Mister_Magisterthanks!
08:23:56Mister_Magisterwhat about template engine? is there some?
08:24:10leorizefor an http template engine, karax can be used as a server-side thing
08:24:14Mister_Magisterhttps://flyx.github.io/emerald/
08:24:16leorizeor htmlgen
08:24:18leorizeor that
08:24:20Mister_Magisterfound this
08:28:54*stefanos82 quit (Quit: Quitting for now...)
08:32:46*jjido joined #nim
08:35:00FromGitter<kayabaNerve> dom96: I tried making a MWE ~24 hours ago and only caused an assert, not a crash.
08:35:11FromGitter<kayabaNerve> It's a deref immediately after an await with no interim var.
08:39:22*m712 joined #nim
08:39:50dom96Do you have a stack trace?
08:41:04FromGitter<bevo009> Hi all, noob question here - is stdout.write equivalent to echo?(without inserting a new line) ⏎ And does Nim support creating direct aliases? ⏎ eg in JS I used var echo = console.log;
08:44:24Mister_Magisteri'm gonna try using nim to make something on Azure sphere starter kit
08:46:21FromGitter<bevo009> Alternatively...is there any equivalent to C++ 'using' so I could drop the stdout. prefix?
08:47:38FromGitter<mratsim> stdout.write is more like a syscall to write
08:47:53FromGitter<mratsim> you can use templates for aliasing
08:48:18FromGitter<bevo009> Cheers, you got an example?
08:48:34FromGitter<mratsim> Nim has "using" and {.this: self.} (in experimental) for shortcuts but this: self is being phased out
08:48:44FromGitter<bevo009> I tried with procs and templates and failes...too noob
08:48:52FromGitter<bevo009> *failed
08:49:25FromGitter<mratsim> several example here: https://github.com/nim-lang/Nim/pull/11686
08:49:52FromGitter<mratsim> template myalias(): untyped = whatYouWantYoAlias
08:50:30FromGitter<bevo009> ahh I see where I went wrong...I used untyped in the wrong field
08:50:39FromGitter<bevo009> Where would you read the documentation on stdout? ⏎ I went here and clicked on source, and got a 404 msg ⏎ https://nim-lang.org/docs/io.html#stdout
08:51:08FromGitter<mratsim> stdout is just the file descriptor for the standard output
08:51:51FromGitter<bevo009> stdout.write then
08:52:24FromGitter<mratsim> https://github.com/nim-lang/Nim/blob/devel/lib/system/io.nim#L297
08:53:00FromGitter<gogolxdong> https://nim-lang-cn.org/ is working in proccess.
08:53:28FromGitter<mratsim> nice
08:55:01*Vladar joined #nim
08:56:21FromGitter<bevo009> @mratsim Just tried going off your example but... ⏎ ```template write(): untyped = stdout.write()``` ⏎ Error: type mismatch: got <int literal(42)> ⏎ but expected one of: ⏎ template write(): untyped [https://gitter.im/nim-lang/Nim?at=5d43fab545da450fec9039bf]
08:57:03FromGitter<mratsim> template write(x: untyped): untyped = stdout.write(x)
08:57:20*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
08:57:44FromGitter<bevo009> many thanks!
09:00:51FromGitter<bevo009> I just realised stdout.write displays a float as an integer, eg stdout.write 42.0 => 42
09:01:00*ng0 joined #nim
09:01:13FromGitter<alehander42> hey kout
09:01:18FromGitter<alehander42> wait, no
09:01:21FromGitter<bevo009> There is no new-line less version of echo right?
09:01:21FromGitter<alehander42> import jsconsole
09:01:25FromGitter<alehander42> jsconsole.log
09:01:29FromGitter<alehander42> for the javascript backend
09:02:09FromGitter<alehander42> but you can also write your own `proc kout*T (x: T) {.importc: "console.log", varargs.} ` (taken from karax~
09:02:41krux02I just made an alias implementation in a pure macro solution: http://ix.io/1Qlf
09:02:57*lritter joined #nim
09:03:43krux02limitations are given: It can't alias template/macro symbols that don't need arguments, as well as const or overloaded procs.
09:05:53krux02what that alias implementation does, it generates the template definiton that you would otherwise write manually for that procedure, automatically.
09:08:53FromGitter<mratsim> what you want to alias stdout.write?
09:09:03FromGitter<mratsim> i.e. here it's currying
09:10:31FromGitter<mratsim> I think the `typed` would prevent that but it would probably work if it's changed to untyped if I followed correctly
09:13:47FromGitter<bevo009> Well I'm just trying to get a version of echo that doesn't print a new line
09:14:09FromGitter<bevo009> echo 42 .0 => 42.0
09:15:51FromGitter<mratsim> The write displaying float as integers if they are exact is probably a quirk of Javascript, you can raise a bug on that
09:16:21FromGitter<bevo009> I've only gone through marin's book so far so I haven't really touched templates and macros yet
09:16:45FromGitter<bevo009> Sorry, what's stdout.write got to do with js?
09:17:52FromGitter<bevo009> I just used js as an example of how I did it there
09:18:13FromGitter<bevo009> This is running a .nim file in code runner or the console
09:22:18FromGitter<alehander42> you can write your own usually
09:22:48FromGitter<alehander42> using something like proc echoNoNewline*(x: varargs[untyped, `$`]) = .. code
09:22:54FromGitter<alehander42> (adapted from system.nim)
09:23:40FromGitter<alehander42> $ is ``` ⏎ `$` ⏎ ⏎ `````` [https://gitter.im/nim-lang/Nim?at=5d44011cd5f4d4442ff2600a]
09:23:57FromGitter<alehander42> this means you receive something that looks like a sequence of strings (it autoconverts all the args with $)
09:24:14FromGitter<alehander42> and you can iterate and stdout.write them one by one without newline
09:24:57FromGitter<alehander42> notice echo is guaraneed to be thread-safe, not sure if thats important for you in this case
09:25:22FromGitter<bevo009> My one won't accept more than one arg either, like this does
09:25:23FromGitter<bevo009> stdout.write 42, " 42 ", 42.0, "\n"
09:29:31FromGitter<bevo009> #alehander42 I tried your version, Error: expression expected, but found ']'
09:29:38FromGitter<bevo009> was I meant to modify it?
09:30:14FromGitter<alehander42> yes
09:30:51FromGitter<bevo009> Well I'm lost then
09:30:55FromGitter<alehander42> ```proc echoNoNewline*(x: varargs[untyped, `$`]) = ⏎ # iterate here ⏎ ``` [https://gitter.im/nim-lang/Nim?at=5d4402cf25944f489adf0726]
09:32:31FromGitter<bevo009> I tried this:
09:32:34FromGitter<bevo009> ```proc echoNoNewline*(x: varargs[untyped, `$`]) = ⏎ stdout.write(x)```
09:34:42FromGitter<bevo009> Error: invalid type: 'untyped' in this context: 'proc (x: varargs[untyped])' for proc
09:35:00*NimBot joined #nim
09:35:25FromGitter<alehander42> uh sorry, untyped proc args were forbidden outside of system.nim indeed i forgot
09:35:32FromGitter<bevo009> It looks like it should work, variable number of args, untyped, but...
09:37:12FromGitter<alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5d44044825944f489adf072b]
09:37:28FromGitter<alehander42> ok, this works: basically you need a template for the untyped stuff usually .. but
09:37:39FromGitter<alehander42> you can really iterate over varargs[untyped]
09:37:44FromGitter<alehander42> so you need a macro
09:40:16FromGitter<bevo009> echoNoNewline 42, " 42 ", 42.0 => 42 42 42
09:40:33FromGitter<bevo009> the float still presents as an integer
09:40:52FromGitter<bevo009> You need a $ in there somewhere?
09:41:23leorize!eval echo 42.0
09:41:26NimBot42.0
09:42:42FromGitter<bevo009> I'm not getting 42.0
09:43:37FromGitter<bevo009> I'm on 0.20.2
09:43:53FromGitter<bevo009> How does it work on nimbot?
09:44:22FromGitter<bevo009> Ohh...that's using echo...duh
09:44:45FromGitter<bevo009> Disregard
09:48:12FromGitter<kayabaNerve> dom96: Uhhhh yes, I have one. I posted it here and can check the logs for it.
09:48:38krux02I once implemented a PR that unified all of float printing to a sigle function. It got rejected for reasons I don't know anymore.
09:48:55FromGitter<bevo009> Curious why it's not part of the core language ⏎ Red lang for instance has print (echo) and prin (no new line) ⏎ Although 'ech' would be a bit weird :) ⏎ Some people must use it, like before user input [https://gitter.im/nim-lang/Nim?at=5d440707a96def4751a34a68]
09:50:03FromGitter<alehander42> well it rarely makes sense, but i agree
09:50:12FromGitter<alehander42> a custom endline character might be useful
09:50:28FromGitter<bevo009> Is that possible?
09:50:37FromGitter<alehander42> well, if you write your own helper
09:50:39FromGitter<alehander42> as this macro
09:50:48FromGitter<alehander42> so, whats the problem? 42.0 is "42"
09:50:49FromGitter<alehander42> ?
09:50:57FromGitter<bevo009> yep
09:51:26FromGitter<bevo009> That is exactly how stdout.write presents it also
09:51:30FromGitter<alehander42> hm, try to add $`arg `
09:51:31FromGitter<alehander42> instead
09:51:36FromGitter<bevo009> echo must stringify it
09:51:37FromGitter<alehander42> ```$`arg` ⏎ ⏎ ``` [https://gitter.im/nim-lang/Nim?at=5d4407a9dfbfc20f5a853769]
09:51:53FromGitter<alehander42> instead of `arg`
09:51:54leorize!eval stdout.write 42.0
09:51:56NimBot42
09:52:03FromGitter<kayabaNerve> `Error: internal error: yield in expr not lowered` ⏎ Commit: https://github.com/MerosCrypto/Meros/commit/cec34fa76a5f13331c005e81ba5b576349b311c2 ⏎ Mutation since I didn't push broken code: https://gitter.im/nim-lang/Nim?at=5d4159f45367476cc983fd09] ⏎ Stack: https://pastebin.com/QSdV0BVA [https://gitter.im/nim-lang/Nim?at=5d4407c3757b7b19c83f650b]
09:52:06FromGitter<kayabaNerve> dom96 ^^
09:52:07leorizeinteresting
09:52:14FromGitter<alehander42> !eval stdout.write 42.2
09:52:17NimBot42.2
09:52:44FromGitter<kayabaNerve> I did push a smaller problem code segment/an attempt MWE after that Gitter message. I just couldn't get a working MWE despite multiple occurrences.
09:53:08leorize!eval stdout.write $42.0
09:53:11NimBot42.0
09:53:24leorizethis is interesting
09:53:35leorizefor one, this is write() signature: https://github.com/nim-lang/Nim/blob/devel/lib/system/io.nim#L320
09:53:56leorizemeaning that `$` should be applied to every non-string params
09:54:16leorizebut produced a different result than $42.0 :/
09:55:57FromGitter<alehander42> hmmm
09:56:01FromGitter<alehander42> what did it produce
09:56:10FromGitter<alehander42> i see, so thats how you can do it as a function
09:57:36FromGitter<alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5d4409102f64b21887da45c5]
09:57:45FromGitter<alehander42> this seems to work ?
09:57:49FromGitter<alehander42> @bevo009
09:58:19leorize!eval stdout.write 42.0
09:58:22NimBot42
09:58:27leorizeit doesn't, and that's a problem
09:58:35FromGitter<alehander42> but this is a different function
09:58:54FromGitter<alehander42> in my example (the last snippet) it works
09:59:16leorize!eval stdout.write 42, "42", 42.0
09:59:19NimBot424242.0
09:59:27leorizeoh it works now :P
09:59:36FromGitter<alehander42> but i see what you mean
09:59:40leorizeah, I see, there's a float overload for write()
09:59:46FromGitter<alehander42> ah right
10:00:11FromGitter<Riderfighter> Hello everyone 👋
10:00:46FromGitter<bevo009> Winner!
10:02:35FromGitter<bevo009> Can't use write, but print works
10:03:50FromGitter<bevo009> Thank you Alexander and leorize
10:04:26FromGitter<bevo009> I have now got some macro and template examples to tweak also...very educational :)
10:10:15*vlad1777d joined #nim
10:25:10*fjellfras quit (Quit: Leaving)
10:25:39*vlad1777d quit (Ping timeout: 258 seconds)
10:27:53FromGitter<bevo009> !eval stdout.write 42.0
10:27:56NimBot42
10:28:02FromGitter<bevo009> !eval stdout.write 42.0, 42.0
10:28:05NimBot42.042.0
10:28:10FromGitter<bevo009> I think there is something wrong with stdout.write() ⏎ With one arg supplied it gives an integer: ⏎ stdout.write 42.0 => 42 ⏎ but with 2 args supplied, it's correct... ⏎ stdout.write 42.0, 42.0 => 42.042.0 [https://gitter.im/nim-lang/Nim?at=5d44103a41d5cd61b59df7b2]
10:29:44FromGitter<alehander42> yes leorize found out
10:29:49FromGitter<alehander42> its overloaded for a float argument
10:30:00FromGitter<alehander42> basically different overloads are called
10:30:05FromGitter<alehander42> when you call it with one float arg
10:30:09FromGitter<alehander42> and with several args
10:30:35FromGitter<bevo009> ahh cool, you guys are onto it
10:33:45FromGitter<bevo009> !eval stdout.write $42.0
10:33:48NimBot42.0
10:33:58FromGitter<bevo009> That'll do for now
10:38:09*go|dfish quit (Quit: SIGQUIT)
10:53:27leorizeanyone know how the HCR codegen works?
11:02:13*ng0 quit (Quit: Alexa, when is the end of world?)
11:05:09*jjido joined #nim
11:08:29FromGitter<alehander42> there should be a hcr doc somewhere
11:08:36FromGitter<alehander42> 1) the source docs
11:08:56FromGitter<alehander42> \+ you even some recorded talks about that, i remember visiting one
11:10:13FromGitter<alehander42> e.g. https://www.youtube.com/watch?v=7WgCt0Wooeo 29 minute
11:14:58*dddddd joined #nim
11:19:11*solitudesf joined #nim
11:30:46FromGitter<kayabaNerve> The 42 problem above reminds me of https://github.com/nim-lang/Nim/issues/10963
11:41:49FromGitter<ahcm> I/O seems a bit slow or what is this: $ for i in 1 2 3; time python3.7 counlines.py FILE ⏎ 728304 ⏎ python3.7 counlines.py FILE 0,40s user 0,09s system 98% cpu 0,498 total ⏎ 728304 ⏎ python3.7 counlines.py FILE 0,38s user 0,09s system 98% cpu 0,479 total ... [https://gitter.im/nim-lang/Nim?at=5d44217dd7fc954750e6d313]
11:42:03FromGitter<ahcm> $ for i in 1 2 3; time ./counlines FILE ⏎ 728304 ⏎ ./counlines FILE 0,70s user 0,10s system 99% cpu 0,801 total ⏎ 728304 ⏎ ./counlines FILE 0,69s user 0,09s system 99% cpu 0,787 total ... [https://gitter.im/nim-lang/Nim?at=5d44218aff59f961b4ff87cd]
11:51:00FromGitter<ahcm> found https://github.com/nim-lang/Nim/issues/3084, reading ...
11:55:36*abm joined #nim
12:00:59*stefanos82 joined #nim
12:03:26FromGitter<ahcm> Might be -O2 vs -O3, how can I change that cc flag for him?
12:17:28*nsf quit (Quit: WeeChat 2.4)
12:27:47*solitudesf quit (Ping timeout: 248 seconds)
12:29:49*kungtotte quit (Quit: WeeChat 2.5)
12:36:51FromGitter<alehander42> passC
12:37:06FromGitter<alehander42> or directly in nim.cfg
12:37:32FromGitter<alehander42> but --opt:speed
12:37:35FromGitter<alehander42> should enable -O3
12:58:05*laaron quit (Remote host closed the connection)
13:01:26*laaron joined #nim
13:05:09*laaron quit (Remote host closed the connection)
13:16:25*laaron joined #nim
13:16:51Zevv-d:danger
13:17:19*sagax quit (Remote host closed the connection)
13:26:48*nsf joined #nim
13:27:56*fjellfras joined #nim
13:28:17*fjellfras quit (Max SendQ exceeded)
13:28:46*fjellfras joined #nim
13:29:25*fjellfras quit (Max SendQ exceeded)
13:29:56*fjellfras joined #nim
13:30:35*fjellfras quit (Max SendQ exceeded)
13:30:52*MarderIII joined #nim
13:31:12*fjellfras joined #nim
13:34:41*clyybber joined #nim
13:54:32*clyybber quit (Quit: WeeChat 2.5)
13:56:18livcdvery generic question but what is now the perf difference between d:release and d:danger ?
13:57:49ZevvThe nim compiler itself is about 50% slower for me on d:release then on d:danger
14:02:45*fjellfras quit (Quit: Leaving)
14:03:36livcdand for other things?
14:03:45livcdfor compiler that's different no ?
14:04:01Zevvwell, the compiler is just a Nim program like any other
14:04:10Zevvbut of course it will depend on your own code.
14:04:14Zevvit's easy to measure :)
14:06:38FromGitter<GULPF> @bevo009 I opened a PR to fix your issue: #11874
14:07:42*sagax joined #nim
14:11:01CalinouTIL about -d:danger
14:13:52krux02@GULPF: I am sorry, but I have to reject that PR
14:14:35FromGitter<GULPF> Ok, why?
14:14:40krux02these overloads exist, so that writing floats to stdout (or other streams) doesn't create an intermediate object.
14:15:01krux02intermediate string objects are performance killers.
14:15:27FromGitter<GULPF> I understand that. But performance should not come before correctness.
14:15:49krux02was it incorrect then?
14:16:10krux02performance is correctness.
14:20:10FromGitter<GULPF> Incorrect in the sense that it's completely unexpected. Based on the comments in strmantle.addFloat it sounds like io.write for float wont even behave the same on different platforms.
14:22:38FromGitter<kaushalmodi> > Incorrect in the sense that it's completely unexpected. ⏎ ⏎ =1
14:22:41FromGitter<kaushalmodi> +1 :)
14:23:37livcdZevv: that's "concerning"
14:24:52krux02GULPF, you can provide a fix for the unexpected behavior, but you cannot provide a "fix" that causes a big performance regression just to fix a minor visual annoyance.
14:25:57Zevvlivcd: what is?
14:28:33FromGitter<arnetheduck> "yeah don't worry about it, we print a bit different things randomly, but nim is fine for your scientific computing, trust us"
14:30:31livcdZevv: is the -d:danger meant to be used for things running in production?
14:30:53ZevvI don't know. See https://github.com/nim-lang/Nim/issues/11860
14:31:44*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
14:32:34krux02livcd: The danger is a flag that Araq came up with. He likes it. I think it is a horrible thing to have. It was was -d:release used to be.
14:33:20krux02GulpF, kaushalmodi, arnetheduck: If my PR would have been accepted here that issue would have already been fixed: https://github.com/nim-lang/Nim/pull/7262
14:33:28livcdkrux02: that's why I am asking. Looks like -d:danger is not really meant to be used as d:release then :X
14:34:12krux02I replaced all my usages of -d:release with -d:danger now, because I did not have a problem with what -d:release meant before it got changed.
14:34:25krux02-d:release now means: all runtime checks on.
14:34:32Zevvsame here. I have a hard time explaining the customer why all my code is running at half speed at once
14:34:36krux02which is super slow and very unexpected.
14:35:10livcdZevv: you have a customer for your Nim code ?
14:35:26krux02I told him that I can't just change the meaning of things under the hood. There are people out there that rely on the behavior of it.
14:35:31Zevvlivcd: one and a half. the other one doesn't know I'm using nim
14:35:58FromGitter<arnetheduck> generally, languages are able to implement the checks that nim does with no more than 2-3% perf loss - I suspect nim's troubles are down to bad codegen, not that the checks can't be implemented efficiently
14:36:13Zevvkrux02: and I imagine distro packagers will just use -d:release as this is still the advertized way in the documentation - so the default prebuilt nims in various os's will just be slower and bigger
14:36:42krux02But that was his intention. He wanted to change the behavior of unteachable server developers that resisted to not use the -d:release even though it meant bad things for them.
14:36:43FromGitter<arnetheduck> the problem with -d:danger is that it complete changes the meaning of the program, specifically because of the way exceptions are used
14:37:56krux02please open an issue about this in the issue tracker. I already told Araq thit this change was a bad idea. But my word alone was aparently not enough to convince Araq that this change was a bad idea.
14:38:44Zevvkrux02: is https://github.com/nim-lang/Nim/issues/11860 not clear enough?
14:38:47FromGitter<arnetheduck> I think it's a good change - in fact, I'd remove -d:danger completely so the language can *focus* on generating good code instead of wasting space and testing on UB
14:39:25krux02very important for those issues are good undarstandable complications and examples. Please don't create an issues that just links to this conversation. I really dislike such issues. An issue should summarize the problem.
14:39:27FromGitter<arnetheduck> ie fix it by fixing -d:release to perform well, not by introducing alternative behaviors
14:40:22FromGitter<ahcm> -d:danger makes it run much faster, still slow in comparison
14:40:29livcdarnetheduck: i have a feeling that was the intention
14:40:52krux02Zevv: If you want an issue to be solved, you need to document it properly.
14:42:20krux02references to external sources are always good underline your point, but referenceing without having own words to describe what is wrong is just an issue that I don't want to deal with. Nobody who works at the compiler wants to deal with lazy written issues.
14:50:58*laaron quit (Remote host closed the connection)
14:52:11krux02@GULPF: generally speaking about pull requests. It is always recommended to create an issues report and a pull request. This way the PR can be closed, but the issue still remains.
14:54:16*laaron joined #nim
14:57:53*sammich quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)
14:59:02*sammich joined #nim
15:09:01*MarderIII quit (Ping timeout: 268 seconds)
15:13:10*nsf quit (Quit: WeeChat 2.4)
15:19:02*laaron quit (Remote host closed the connection)
15:21:28*floppydh quit (Quit: WeeChat 2.5)
15:21:30*laaron joined #nim
15:32:00*ng0 joined #nim
15:37:30*jjido joined #nim
15:38:24*laaron quit (Remote host closed the connection)
15:41:57*shashlick quit (Remote host closed the connection)
15:42:23*shashlick joined #nim
15:42:46*laaron joined #nim
15:46:48FromGitter<bevo009> @krux02 Re: "these overloads exist, so that writing floats to stdout (or other streams) doesn't create an intermediate object... intermediate string objects are performance killers." ⏎ Be aware though, Nim prints correctly when the first decimal place isn't a 0 ⏎ !eval stdout.write 42.1
15:46:59FromGitter<bevo009> !eval stdout.write 42.1
15:47:02NimBot42.1
15:47:12FromGitter<bevo009> !eval stdout.write 42.0
15:47:14NimBot42
15:52:08*Jesin quit (Quit: Leaving)
15:54:43*JustASlacker joined #nim
15:56:21*Jesin joined #nim
15:58:14*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:07:19*laaron quit (Remote host closed the connection)
16:10:14*laaron joined #nim
16:12:14*solitudesf joined #nim
16:14:43Zevvcan anyone elaborate on the exact limitations of the vm considering object variants? I'm trying to do things that do not work, and I here say that this is not supposed to be possible, but the VM works with NimNodes all the time, which are just object variants
16:21:05lqdev[m]dang recursive dependencies (or the lack of them), I must do strange workarounds to get practically anything to work
16:34:51*dgreen joined #nim
16:34:55*Trustable joined #nim
16:37:08*shashlick quit (Remote host closed the connection)
16:55:53*MarderIII joined #nim
16:56:58*JustASlacker quit (Quit: Leaving)
17:13:51*jjido joined #nim
17:20:50*drewr quit (Quit: ERC (IRC client for Emacs 26.2))
17:24:30*drewr joined #nim
17:24:56*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
17:32:39*jjido joined #nim
17:44:07FromGitter<zacharycarter> is there a link to the rationale behind `-d:danger`?
17:45:12Zevvnot publicly afaik. it just happened
17:45:12FromGitter<zacharycarter> never mind I found the issue
17:45:20Zevvlink?
17:45:33FromGitter<zacharycarter> well - I guess this explains at least the logic behind it - https://github.com/nim-lang/Nim/commit/721534119000c2bd53cc72b531726a6104381222
17:45:48FromGitter<zacharycarter> although you probably also have to be a mind reader to fully understand
17:46:58FromGitter<zacharycarter> sudden increases in binary size and decreases in performance aren't good things right before a 1.0 release
17:47:19Zevvsudden increases in diagnostics and reliability are, though
17:47:28FromGitter<zacharycarter> is the only impact that people need to start using `-d:danger` vs `-d:release`? Also `-d:danger` sounds bad
17:47:35FromGitter<zacharycarter> like it's something you shouldn't be doing
17:47:36ZevvI guess that's the point
17:48:10FromGitter<zacharycarter> so basically to have the `like C` performance Nim touts everywhere, you have to do bad things with it
17:48:10ZevvI think I like arnetheducks point: it's supposed to suck - which should put the weight on improving performance of the stuff which is still enabled with using d:release
17:49:25FromGitter<zacharycarter> it's sound reasoning but it's also not going to be a very marketable selling point for the language
17:50:50FromGitter<zacharycarter> but I guess it's better than telling people they will be using a more performant language when in reality they are sacrificing safety to do so
17:52:05FromGitter<zacharycarter> in other Nim related news - I fixed my tile rendering issue I was facing a week or so ago -
17:52:12FromGitter<zacharycarter> (https://files.gitter.im/nim-lang/Nim/BU1q/image.png)
17:52:16FromGitter<zacharycarter> so I guess I'm back to working on a 3d rts
17:54:12FromGitter<zacharycarter> starting with the map editor - the texture is actually a 2d texture array so I can easily switch b/w textures
17:54:16FromGitter<zacharycarter> inside of the fragment shader
17:54:59FromGitter<zacharycarter> (https://files.gitter.im/nim-lang/Nim/AA2e/image.png)
17:55:04FromGitter<zacharycarter> and I already have the nimscript scripting / hot reloading system working for game code
17:55:41FromGitter<zacharycarter> so time to start adding nuklear gui to the mix in the editor
18:09:45*nsf joined #nim
18:15:16*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
18:15:56*laaron joined #nim
18:19:17FromGitter<zacharycarter> so regarding `newruntime` - if you want to depend on old code that isn't written with `newruntime` in mind - will you be able to?
18:19:26FromGitter<zacharycarter> I imagine no?
18:21:31disruptekdepends upon how old it is.
18:22:33FromGitter<zacharycarter> hmm okay
18:25:11*enthus1ast quit (Ping timeout: 250 seconds)
18:38:20*enthus1ast joined #nim
19:01:14*ehmry quit (Remote host closed the connection)
19:08:27*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
19:13:21*laaron quit (Quit: ZNC 1.7.1 - https://znc.in)
19:20:51*laaron joined #nim
19:33:10*jjido joined #nim
19:42:59*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
20:02:19*nsf quit (Quit: WeeChat 2.4)
20:30:28*stefanos82 quit (Quit: Quitting for now...)
20:31:19*Kaivo quit (Quit: WeeChat 2.5)
20:39:50*jjido joined #nim
20:50:53FromGitter<deech> Can anyone familiar with the compiler offer some guidance? After a function call the compiler immediately tries to compile time evaluate compile time nodes or ones that have no side effects (https://github.com/nim-lang/Nim/blob/devel/compiler/semexprs.nim#L873), which allows compile time evaluation of nodes that don't even typecheck! The proper behavior I think is to do it code gen. time but I can't immediately see where
20:50:53FromGitter... to place that call.
20:55:46FromDiscord_<telara{a> Hello guys, this it the Nim book, "Nim in Action" - Dominik Picheta, for 2017.
20:55:46FromDiscord_<telara{a> https://cdn.discordapp.com/attachments/371759389889003532/606953306799210496/unknown.png
20:56:21FromDiscord_<telara{a> That section say, that the limit for an uint16 is 65,5350 (?)
20:56:53FromDiscord_<telara{a> The real limit is, 65,535
20:56:53FromDiscord_<telara{a> https://cdn.discordapp.com/attachments/371759389889003532/606953588853571614/unknown.png
20:58:11FromDiscord_<telara{a> This is from ~ 3 years ago ahahah. Anyway, I want say publish that.
20:58:23FromDiscord_<telara{a> This is from ~ 2 years ago ahahah. Anyway, I want say publish that.
20:59:35*ng0 quit (Quit: Alexa, when is the end of world?)
21:05:40*drewr quit (Ping timeout: 276 seconds)
21:10:27*Vladar quit (Remote host closed the connection)
21:16:07*kungtotte joined #nim
21:18:25*drewr joined #nim
21:18:38FromDiscord_<Bit> looks like a typo considering the comma placement
21:24:48*MarderIII quit (Quit: Leaving)
21:26:02FromDiscord_<abeaumont> the same applies to 8-bit and 32-bit cases
21:28:47*solitudesf quit (Ping timeout: 245 seconds)
21:32:16*kungtotte quit (Quit: WeeChat 2.5)
21:39:04*Trustable quit (Remote host closed the connection)
21:41:09disrupteki'm really tired of forgetting to `import tables` have having no table semantics on table types imported from elsewhere. it's insane.
21:41:11dom96yes, it's not even a typo
21:41:26dom96it's a an error that was introduced when the book was converted into print
21:44:00dom96and yeah, it sucks. I actually wanted to use this as a neat little reference recently and got confused :(
22:03:34*kungtotte joined #nim
22:04:55disruptekwhat's it called when you make a bad decision, followed by a bad decision, followed by a bad decision, followed by a decision that would have been bad if it hadn't followed all those other bad decisions?
22:05:08FromDiscord_<telara{a> Again, why ythis guys are not equals?
22:05:08FromDiscord_<telara{a> https://cdn.discordapp.com/attachments/371759389889003532/606970753295974401/unknown.png
22:15:55*rockcavera quit (Remote host closed the connection)
22:17:02FromGitter<Varriount> Ugh, is it me, or is java the only language where it's "normal" to encounter 500 line tracebacks
22:19:01disruptekthat's a "feature" you get "for free".
22:21:51disruptekso `quote do` is compile-time only? 🙄
22:27:28*Summertime quit (Quit: Sunsetting.)
22:28:49*Summertime joined #nim
22:34:44FromGitter<Varriount> disruptek: Yes
22:34:51FromGitter<Varriount> It's a macro helper
22:36:18disruptekknow of any example of crafting nim code at runtime?
22:37:05disrupteki dunno why i should have to do at compile time what i wish to do at runtime.
22:45:23*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
23:10:24*snifftek quit (Remote host closed the connection)