<< 13-08-2018 >>
00:07:29*yglukhov[i] joined #nim
00:12:06*yglukhov[i] quit (Ping timeout: 272 seconds)
00:24:11*yglukhov[i] joined #nim
00:28:34*yglukhov[i] quit (Ping timeout: 256 seconds)
00:29:45*erratic quit (Quit: this server has gone to sleep)
00:53:05*erratic joined #nim
00:53:46FromGitter<kayabaNerve> vosper: nimcrypto is the superior lib but Na does try to keep things inside the library. I haven't used it but I can try to get a demo up for you if you want.
00:59:10vosperkayabaNerve: That would be helpful, thanks
00:59:35*vosper left #nim ("Textual IRC Client: www.textualapp.com")
01:07:09*jdhorwitz_ joined #nim
01:16:53FromGitter<kayabaNerve> vosper: Doesn't appear to work on devel 0_o I'm switching to 0.18
01:18:30FromGitter<kayabaNerve> My VM is maxed out lol.
01:48:28*BitPuffin quit (Ping timeout: 260 seconds)
01:53:09FromGitter<kayabaNerve> vosper Done: https://gist.github.com/kayabaNerve/074a4528a8da0d63f835f63c9484ed56
01:53:16FromGitter<kayabaNerve> Can someone on IRC tell me if he's still here lol
01:53:41*wildlander quit (Quit: Konversation terminated!)
01:58:28FromGitter<kayabaNerve> Oh. C seems to do better in the benchmarks for that gist. Just an FYI ;)
02:39:57*yglukhov[i] joined #nim
02:44:21*yglukhov[i] quit (Ping timeout: 240 seconds)
02:55:56*leorize joined #nim
02:57:08*data-man joined #nim
03:05:27*cspar quit (Ping timeout: 240 seconds)
03:09:44*sielicki joined #nim
03:09:56*sielicki quit (Remote host closed the connection)
03:11:49*cspar joined #nim
03:21:02*jdhorwitz_ quit (Quit: Connection closed for inactivity)
03:24:58*ftsf quit (Ping timeout: 244 seconds)
03:25:28*nullrouted joined #nim
03:27:16*nullrouted quit (Remote host closed the connection)
03:58:55*dddddd quit (Remote host closed the connection)
04:06:03*endragor joined #nim
04:27:11*ftsf joined #nim
04:37:28*leorize quit (Ping timeout: 256 seconds)
04:43:01*stefanos82 joined #nim
05:08:44*nsf joined #nim
05:45:57*miran joined #nim
05:51:50*leorize joined #nim
06:23:59*arecaceae quit (Remote host closed the connection)
06:24:18*arecaceae joined #nim
06:24:35*xet7 joined #nim
06:24:56*Trustable joined #nim
06:36:47*p3pp3rb0x joined #nim
06:37:03*p3pp3rb0x quit (K-Lined)
06:39:05*yglukhov[i] joined #nim
06:47:41*Exagone31320 joined #nim
06:49:22*Exagone31320 quit (Remote host closed the connection)
06:55:43leorizeAraq: is the poll backend of ioselectors tested?
06:57:34leorizeI tried to use it on Linux and tasyncRecvLine crashed
07:24:21FromGitter<craigglennie> @kayabaNerve sorry, I went offline. Thanks for posting that, the thing that I am missing is how to turn a user-supplied password of some variable length into an appropriately-sized key. From looking at the libsodium docs I imagine that I would use one of the password hashing functions (argon2 or scrypt, in the docs, but maybe also pbkdf2 from nimcrypto would work?). Unfortunately the nim lobsodium bindings seem to
07:24:21FromGitter... be missing the password hashing functions...
07:25:11*PMunch joined #nim
07:25:40FromGitter<kayabaNerve> @craigglennie KDFs take Password + Salt.
07:26:39FromGitter<gogolxdong> Is there any Ethernet frame CRC library?
07:26:47FromGitter<craigglennie> @kayabaNerve Okay, but that's fine, right? I take a user password, sprinkle some salt, get my key, and then use it with crypto_secretbox_easy?
07:26:55FromGitter<kayabaNerve> It is missing a scrypt binding; I'm guessing it's due to the version; it should be easy as hell to implement nimcrypto
07:27:10FromGitter<kayabaNerve> Yeah but the other user then most know the password + salt
07:28:39FromGitter<craigglennie> @kayabaNerve: ahh, in this case the user is encrypting it for themselves. A string is encrypted on the users machine (in the nim program) and decrypted in a browser by the same user. I thought the salt could be stored alongside the encrypted message
07:29:27FromGitter<kayabaNerve> The salt in KDF is used locally
07:29:32FromGitter<kayabaNerve> To store the passwords locally
07:29:43FromGitter<kayabaNerve> The password is used in the box
07:30:08FromGitter<kayabaNerve> If you solely want to get a valid length, do a hash squared or cubed.
07:30:31FromGitter<kayabaNerve> AKA SHA512(SHA512(password)). Don't use SHA512. Just giving an example.
07:32:45FromGitter<craigglennie> Ahh ok. Does nimcrypto implement AES-CBC? I see that the new WebCrypto API supports it in the browser. I think maybe it's in the bcmode module, but I'm not sure
07:33:06FromGitter<kayabaNerve> The is AES
07:33:10FromGitter<kayabaNerve> No idea about the CBC variant
07:33:43FromGitter<craigglennie> With the example you gave, my issue still seems to be the hash function - where do I get a good one for encryption? Not from nim-lobsodium AFAIK
07:34:32FromGitter<kayabaNerve> It's the Rjindael part of nimcrypto
07:34:35FromGitter<kayabaNerve> nimcrypto :P
07:34:45FromGitter<kayabaNerve> Pick which one you want; make sure you get a 32 byte output.
07:34:54FromGitter<kayabaNerve> *32 long string of A-F characters.
07:36:37FromGitter<gogolxdong> How to echo this ⏎ ⏎ ``` var buf = EnetBuffer(data: addr sendbuf, dataLength: payload.len) ⏎ echo buf.data``` [https://gitter.im/nim-lang/Nim?at=5b7135045ec2bc174fefdb2a]
07:37:31FromGitter<kayabaNerve> @gogolxdong What part do you want to print? The address or whatever sendbuf is?
07:37:44FromGitter<kayabaNerve> If the second, you need to say what type sendbuf is
07:37:48FromGitter<gogolxdong> content of sendbuf.
07:37:58FromGitter<gogolxdong> ` var sendbuf : array[46, byte] ⏎ `
07:38:48FromGitter<craigglennie> @kayabaNerve I have no idea which one I want, that's the thing. I don't know anything about cryptography and I'm trying to follow the mantra of "don't roll your own". I thought the libsodium stuff was supposed to make the whole thing somewhat dummy-proof, but now I am off learning about hashing functions when all I want to do is take a string and a password and encrypt the string somewhat securely... *sigh*... :)
07:39:14FromGitter<kayabaNerve> `echo buf.data[]` will call `$` on the array[46, byte]. You may want to do a for loop over buf.data[] though
07:39:28FromGitter<kayabaNerve> @craigglennie I can recommend one if you give me a second
07:39:46FromGitter<craigglennie> That would great :)
07:42:25FromGitter<kayabaNerve> You can use just nimcrypto with Rjindael (AES) OR BCMode (AES-CBC and others)
07:43:48*Vladar joined #nim
07:45:21FromGitter<kayabaNerve> You could also use RIPEMD128^2 + LibSodium but 128 bits is really bad security
07:46:03FromGitter<kayabaNerve> I would use: ⏎ BCMode > Rjindael > RIPEMD128^2 (RIPEMD128 of RIPEMD128 of data) + LibSodium
07:46:13FromGitter<kayabaNerve> Seriously. 128 bits is bad.
07:47:24FromGitter<gogolxdong> actually Error: type mismatch: got <pointer> of echo buf.data[]
07:48:13FromGitter<kayabaNerve> The third lets you use LibSodium/lets you distribute a password of "pass"/requires a computer to try 16^32 passwords before it's guaranteed to get in. ⏎ ⏎ Hell, the squared isn't needed and doesn't help here. I've been working on spam filters for too long...
07:49:00FromGitter<gogolxdong> Did dereference of pointer syntax change?
07:49:15FromGitter<craigglennie> @kayabaNerve Ok, thanks. I need to stick with something that there's a JS library for, so that I can decrypt. If nimcrypto supports AES-CBC with bcmode then I *think* I can decrypt in-browser. I'll see if I can get that to work. As an aside, nimcrypto seems to have an particular enthusiasm for really short variable names... Is "inp" = input and "oup" = output? What's wrong with "input" and "output"... but anywho...
07:49:15FromGitter... Thanks for the help
07:49:34FromGitter<kayabaNerve> @gogolxdong Iterate over the array. ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b71380e988005174ed0afec]
07:49:42FromGitter<kayabaNerve> No idea why you have that error. Just trying to fix it.
07:49:53FromGitter<kayabaNerve> @craigglennie Good luck
07:59:20FromGitter<gogolxdong> I know dereference pointer syntax, but it fails.
08:03:45*gmpreussner_ joined #nim
08:04:44*gmpreussner quit (Ping timeout: 256 seconds)
08:06:50Araquse a pointer to an array?
08:13:06FromGitter<gogolxdong> any issue?
08:18:49FromGitter<gogolxdong> @Araq do you have library to calculate Ethernet FCS ?
08:20:27*xet7 quit (Quit: Leaving)
08:21:19Araqsorry, no
08:21:58FromGitter<andreaferretti> @timotheecour @mratsim In fact, I just added back support for static dimensions in neo under the package `neo/statics`
08:22:28FromGitter<andreaferretti> as it turns out, using static types as a wrappers for dynamic ones is easier than the other way round
08:22:46FromGitter<andreaferretti> if only because sometimes dimensions are not known at compile time
08:23:19FromGitter<andreaferretti> so some form of dynamic vectors/matrices/tensors are needed anyway
08:23:25*yglukhov[i] quit ()
08:23:37FromGitter<andreaferretti> and the simplest way to make them interoperable with the static ones
08:23:54FromGitter<andreaferretti> is just to have the static types be distinct types over the dynamic ones
08:24:18FromGitter<andreaferretti> this way you get static typing when you want it (and you know dimensions at compile time)
08:24:38FromGitter<andreaferretti> but you can always switch back to dynamic types
08:25:08FromGitter<andreaferretti> there is a small overhead in that metadata is needed even if dimensions are known at compile time
08:25:08*yglukhov[i] joined #nim
08:25:12FromGitter<andreaferretti> but it seems acceptable
08:29:48*ftsf quit (Ping timeout: 244 seconds)
08:30:32*miran quit (Ping timeout: 272 seconds)
08:32:43*xet7 joined #nim
08:35:43*ftsf joined #nim
08:36:24FromGitter<gogolxdong> Is the magic number of CRC32 random for its implementation , why there is `c = (c >> 1)^(0xEDB88320);` and calculate a different CRC table with others provided directly?
08:39:20FromGitter<gogolxdong> It seems that there are many different purpose CRC32 algorithm.
08:41:15FromGitter<kayabaNerve> @andreaferretti @gogolxdong and I were just using your libsodium wrapper
08:41:39FromGitter<kayabaNerve> @gogolxdong Backdoor for the NSA.
08:41:48FromGitter<kayabaNerve> (joking; no idea)
08:42:43FromGitter<gogolxdong> Is it the same thing?
08:43:46*Colti19 joined #nim
08:44:13*Colti19 quit (Remote host closed the connection)
08:45:11FromGitter<kayabaNerve> @andreaferretti Sorry; confused you. I'm pretty tired.
08:45:32FromGitter<kayabaNerve> @gogolxdong The person? No. I'm just tired. I'll go to bed now... 4am. Bit late.
08:46:41FromGitter<gogolxdong> no , I mean is the crc in libsodium and Ethernet FCS the same , BTW there is no CRC in libsodium which I checked.
08:51:03*miran joined #nim
08:59:40FromGitter<gogolxdong> cannot send Ethernet frame due to FCS.
09:05:22*leorize quit (Ping timeout: 272 seconds)
09:11:03*TheLemonMan joined #nim
09:11:45TheLemonMangogolxdong, different constants = different polynomials
09:12:22*Karasu joined #nim
09:13:21*NimBot joined #nim
09:14:15*Karasu quit (Remote host closed the connection)
09:14:54ehmryI would prefer a libsodium compatible Nim library to a wrapper, nacl/libsodium has side-effects when it tries to reads the RNG file and does some memory security stuff that probably isn't useful
09:15:24ehmryits a frustrating library to use if you aren't running unix or windows
09:17:50ehmrythe actuall crypto routines should be copy-pasted though
09:36:06FromGitter<gogolxdong> @TheLemonMan what does it mean?
09:37:10TheLemonManwikipedia is your friend here
09:38:09FromGitter<gogolxdong> I mean how do I know which constants Ethernet FCS use?
09:41:31*xet7 quit (Quit: Leaving)
09:44:14TheLemonManthe 802.3 spec
09:44:17FromGitter<tim-st> @Araq if the following is fixed, it will increase VM time and memory (maybe quite much) I think: https://github.com/nim-lang/Nim/issues/8603#issuecomment-412461315
09:45:36Araqtim-st: I doubt 'when' is used that often that it makes a difference
09:45:54*cspar quit (Ping timeout: 272 seconds)
09:46:22FromGitter<tim-st> ok, but I often see code like `when false: do this which takes forever and eats ram`
09:50:15TheLemonManAraq, do you think that removing optLineDir from preInitProc may have adverse side effects I'm not aware of?
09:50:56Araqtim-st: the body is not affected, only the condition of the when
09:53:11*xet7 joined #nim
09:54:07FromGitter<tim-st> That's the problem: the body should be checked too, if it contains incorrect code
09:55:12Araqbut not for 'when false'
09:55:25FromGitter<drslump> hi @tim-st , I tried to explain the problem in this comment https://github.com/nim-lang/Nim/issues/8603#issuecomment-412465389
09:56:57FromGitter<tim-st> @drslump I read your comment, but I'm pretty sure the nim compiler can find out every [incorrect] example I posted here:https://github.com/nim-lang/Nim/issues/8603#issuecomment-412461315
09:57:44*Venusaur13 joined #nim
09:57:56*Venusaur13 quit (Remote host closed the connection)
09:58:02FromGitter<tim-st> I think at this time it's already known for the nim compiler that `typedesc` is `typedesc` and `typedesc` != Expression
09:58:14FromGitter<drslump> a work around if your intention is to make sure your code is correct even if it's "disabled" in a specific compilation (i.e. different platform) is to extract the logic into normal functions and just have function calls in the `when` statement
09:59:10FromGitter<tim-st> yes, that could work
10:01:32FromGitter<drslump> it's not really a problem of the compiler being able to figure out the semantics of a when block, it's a language design decision. Those blocks are not resolved unless the condition that guards them is true, this simplifies many use cases. Otherwise the compiler would not know if when it fails to validate a block it's actually an error that should be reported or it's just that in a specific platform some library doesn't
10:01:32FromGitter... have a given symbol for instance.
10:01:35FromGitter<gogolxdong> This is a crc32* in Nim Nim/tests/manyloc/keineschweine/dependencies/enet/enet.nim
10:01:52FromGitter<gogolxdong> don't know how to use it
10:05:08*ng0 joined #nim
10:05:36FromGitter<tim-st> @drslump hm, not sure what can be done about it, at least I expected that `when -1:` isnt true, and that it can be find out that `int != bool`
10:06:59FromGitter<drslump> yes, that's a the bug I was trying to fix, in `when <expr>:` the `<expr>` must be fully valid (constant and producing a bool)
10:06:59FromGitter<gogolxdong> it uses ⏎ ⏎ ``` TENetBuffer*{.pure, final.} = object ⏎ data*: pointer ⏎ dataLength*: csize``` ⏎ ⏎ which data is a pointer to hex bytes array, conflicts with the dereference symbol [] as for array indexing. [https://gitter.im/nim-lang/Nim?at=5b715843179f842c972fcfaa]
10:08:28FromGitter<gogolxdong> since it's a wrapping of C library, not sure how to change this.
10:08:31FromGitter<tim-st> @drslump And additionally *some* checks are done in `when`-blocks and some are not, that is surprising to me
10:11:47FromGitter<gogolxdong> got SIGSEGV: Illegal storage access. (Attempt to read from nil?)
10:13:11FromGitter<drslump> @tim-st yeah, I think the syntax is unfortunate, it should be more explicit about what's going on, otherwise feels like another flow-control construct more but it has completely different semantics
10:14:06FromGitter<gogolxdong> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b7159ee67eb9f60fed37d82]
10:15:38*yglukhov[i] quit (Ping timeout: 265 seconds)
10:16:02FromGitter<tim-st> @drslump ok, but then at least it should be possible to check the syntax of the condition that would already be good
10:16:53*yglukhov[i] joined #nim
10:17:07zacharycarter[m]Araq: should I create an issue for - https://gist.github.com/zacharycarter/f61be4127da3ec6efc9f3c2348b6469a - only the first set of ident|type is parsed into varargs?
10:21:30Araqzacharycarter[m]: that used to compile, right?
10:22:02zacharycarter[m]Not sure - the old code wasn't using varargs[untyped]
10:22:07zacharycarter[m]it was using callsite
10:22:25zacharycarter[m]and the ident|type sets were passed in via a tuple
10:22:27Araqwell but it used to *parse*
10:22:37Araqand now the parser errors.
10:22:51zacharycarter[m]https://github.com/fowlmouth/allegro5/blob/master/src/al.nim#L361-L404
10:23:23zacharycarter[m]let me try with an older version of Nim and see what varargs get parsed
10:29:16*TheLemonMan quit (Quit: "It's now safe to turn off your computer.")
10:37:22*lunaaa joined #nim
10:39:12*lunaaa quit (Remote host closed the connection)
10:55:42*dddddd joined #nim
11:00:48ng0haven't looked at your irc bot code again, but does it support NickServ authentication of itself? Turns out our drupal irc bot quit its job earlier than we expected, and now I need something without a search, generated from a database (I'm still hoping for plain text log format) and output as html with pagination.
11:10:03*Vladar quit (Remote host closed the connection)
11:15:22ng0furthermore, would you accept a patch for or be willing yourselves to implement support for postgresql in nimbot?
11:18:37FromGitter<alehander42> @Araq I finally wrote down https://github.com/nim-lang/Nim/pull/8627
11:19:36FromGitter<alehander42> honestly it would be really cool to just have a compiler switch where for the js target `string` is represented as js string and <=> with cstring
11:21:43FromGitter<alehander42> currently the whole cstring vs string thing leads to a big mess in nim->js code
11:30:23*Vladar joined #nim
11:35:27*bigpet7 joined #nim
11:37:12*bigpet7 quit (Remote host closed the connection)
11:48:17*miran quit (Ping timeout: 248 seconds)
12:01:56FromGitter<dom96> Ng0: postgresql? What for?
12:04:10ng0We have many years of logs (almost a decade now) for a collection of communities. The idea is to go bot -> postgresql -> static html output, because we have too much past logs and it is easier to work with in cases of future migration.
12:06:50ng0ah. the logs already are in a DB, that's the other reason ;)
12:07:28ng0I wasn't around in '09 otherwise I would have had my opinions about a Drupal integrated ircbot back then..
12:17:03zacharycarter[m]Araq: 0.16.0 parsed this as - `[id, PJoystick]` - so it looks passing in something like - `id:PJoystick, stick,axis:cint, pos:cfloat, button:cint` as `varargs[untyped]` has never been supported
12:26:33*gregf joined #nim
12:28:27*gregf quit (Remote host closed the connection)
12:40:44*nsf quit (Quit: WeeChat 2.2)
12:46:51*ng0 quit (Quit: Alexa, when is the end of world?)
12:53:50*dgwana joined #nim
12:56:02*dgwana quit (Client Quit)
13:05:12FromGitter<Varriount> @alehander42 But JavaScript strings and Nim strings have different behaviors.
13:07:21*endragor quit (Remote host closed the connection)
13:17:54*noonien joined #nim
13:23:56*Guest4399626 joined #nim
13:25:33*francisl joined #nim
13:25:42*endragor joined #nim
13:25:43*Guest4399626 quit (Remote host closed the connection)
13:30:13*endragor quit (Ping timeout: 260 seconds)
13:36:35Araqzacharycarter[m], it parsed before and now it doesn't, as far as I can tell, so please create an issue
13:36:51Araqthe fact that it never sem'checked is not what's concerning me
13:39:28FromGitter<alehander42> @Varriount well, you have to give an example of such a critical difference, for most common edge cases one can just tackle this in a `when` overrided method
13:39:54FromGitter<alehander42> mutability is the huge difference, but I don't think I use or depend on string mutability anywhere in my jsbackend code
13:42:28*francisl quit (Ping timeout: 260 seconds)
13:43:03*francisl joined #nim
13:45:55*ng0 joined #nim
13:54:59*wildlander joined #nim
14:06:08*ninsei joined #nim
14:08:11*ninsei quit (Remote host closed the connection)
14:10:39*Compu11 joined #nim
14:12:30*Compu11 quit (Remote host closed the connection)
14:14:41FromGitter<alehander42> btw what happend with https://forum.nim-lang.org/t/1793 I guess the move to immutable string plans never materialized?
14:24:19*nsf joined #nim
14:25:41AraqI implemented immutable strings and got no performance differences in a number of benchmarks
14:25:42FromGitter<zah> If we had immutable strings under a flag, the string type in JS could be mapped to the native JS string
14:26:09Araqand the new move semantics are more elegant anyway, ymmv
14:26:37Araqwe ensure there is only one owner of the string so mutation is safe
14:26:49Araqwhy disallow it then?
14:27:40*data-man quit (Remote host closed the connection)
14:27:47Araqwe can't map it to JS's string, JS uses utf-16, Nim's strings are utf-8
14:29:39Araqunless you want to be more hand-wavy about it than we usually are :P
14:32:04FromGitter<zah> well, it’s an improvement in JS-only codebases (read electron and web apps).
14:33:39FromGitter<alehander42> is there a lot of code in the js'supported stdlib that reallistically depends on the internals of the encoding? (except unicode.nim and the cstring string conversion functions?)
14:36:25Araqwell there is lots of code that does 'result = ""; for ...: result.add()' and relies on strings being mutable
14:38:21Araqthough I guess 'add' can be mapped to JS's strings
14:41:23Araqin the longer run when wasm is a thing Nim's strings can be more efficient than JS's
14:41:39Araqand we don't have to butcher their semantics
14:45:45FromGitter<alehander42> currently the problem is mostly that cstring are very pervasive in js environments, all api-s, libs etc require one to `cstring()` his nim strings and most importantly all the object keys are `cstring`, and they are literally everywhere
14:47:03FromGitter<alehander42> so one can't reasonably use `string` without `cstring()` all the time which is also slow (nim can autoconvert strings to cstring, I am not talking about code pollution: on the other hand those autoconvertions are a bit magical and I don't like them in some cases)
14:47:36FromGitter<alehander42> the other consistent option is to just use cstring everywhere, but then one loses all the stdlib string stuff
14:47:51Araqthat's what I do
14:48:20Araqthe stdlib string stuff is slow(er) for JS anyway and produces bigger code
14:48:21FromGitter<alehander42> well maybe reimplementing most stdlib string functions would be simpler indeed
14:48:31Araqthat's what Karax did
14:48:45FromGitter<alehander42> but in this case we should just put it in the stdlib
14:49:03FromGitter<alehander42> and not have each jsbackend lib redefine them again
14:50:51AraqI can move jstrutils to the stdlib
14:54:49FromGitter<alehander42> ok, I can PR with some of my own cstring util equivalents
14:55:31Araqalso JS strings can be null and Nim's strings cannot be 'nil' anymore
15:00:08FromGitter<alehander42> well, `cstring` can be nil, js strings by definition are never nil :P
15:02:16FromGitter<alehander42> btw nim's strings are still nil in js backend
15:02:24FromGitter<alehander42> so this is actually broken a bit ?
15:02:50FromGitter<alehander42> (or at least null is not threated as an empty string )
15:03:22Araqthe JS codegen doesn't know about the changed Nim string semantics yet
15:03:39Araqtried to fix it but got strange regressions and the produced JS code is a pita to debug
15:05:49FromGitter<alehander42> anyway, this shouldn't matter for the cstring utils (except that they might need some `nil` checks)
15:06:25FromGitter<alehander42> hm, fixing this would be cool
15:15:18dom96well, this spam is seriously not going away and seems to be getting worse https://i.imgur.com/zbBkkM2.png
15:16:15Araqdom96, gah, any suggestions?
15:16:35dom96Go to #freenode and ask them to fight this
15:17:28FromGitter<alehander42> is this in our channel :O
15:17:55*alehander42 joined #nim
15:18:12*Vladar quit (Remote host closed the connection)
15:21:26ng0they already do
15:22:09*alehander42 quit (Ping timeout: 252 seconds)
15:37:58*yglukhov[i] quit (Remote host closed the connection)
15:44:10*BitPuffin joined #nim
15:44:49*cryptocat1094 joined #nim
15:53:48*yglukhov[i] joined #nim
15:59:52*cornfeedhobo quit (Quit: ZNC - http://znc.in)
16:18:53*cornfeedhobo joined #nim
16:20:09*PMunch quit (Quit: Leaving)
16:44:59*chachasmooth0 joined #nim
16:45:48*chachasmooth0 quit (Remote host closed the connection)
16:58:53*Guest29805 joined #nim
16:59:12*Guest29805 quit (K-Lined)
17:00:28FromGitter<arnetheduck> https://julialang.org/blog/2018/08/one-point-zero
17:06:39*issyl017 joined #nim
17:07:43*issyl017 quit (Remote host closed the connection)
17:07:56*timvisher24 joined #nim
17:10:55*timvisher24 quit (Remote host closed the connection)
17:21:38*data-man joined #nim
17:28:59*nsf quit (Quit: WeeChat 2.2)
17:39:37*icyphox quit (Ping timeout: 248 seconds)
17:41:08*Entropic joined #nim
17:56:06*ltr_ joined #nim
18:02:33Araqarnetheduck: I think most of us already read it
18:05:15stefanos82Araq: about the spam you mentioned before...Nim is in great condition I have to admit. In ##C they are posting some huge d*ks for pictures lol
18:06:16*cods22 joined #nim
18:08:49*tigermousr4 joined #nim
18:08:52*tigermousr4 quit (Remote host closed the connection)
18:10:25*cods22 quit (Remote host closed the connection)
18:16:36*thomasross quit (Remote host closed the connection)
18:17:01*thomasross joined #nim
18:25:10*ketralnis joined #nim
18:27:45*ketralnis quit (Remote host closed the connection)
18:28:32*endragor joined #nim
18:31:37*jgh- joined #nim
18:32:20jgh-hey, what does `asyncCheck` do? The most explanation it seems to get in the docs is "If you do not care for the result of a Future then you should use the asyncCheck procedure instead of the discard keyword."
18:32:58*endragor quit (Ping timeout: 260 seconds)
18:37:17*natrys joined #nim
18:41:05Araq'asyncCheck' is like 'await' that then trows away the awaited value. I think
18:41:57jgh-I see.
18:42:48AraqasyncCheck can also be used in an non-.async context, it's one bridge between the async and non-async world
18:43:08*jgh- quit (Remote host closed the connection)
18:45:08*Loki15 joined #nim
18:47:47*jgh- joined #nim
18:48:27jgh-ok. I'll dig more into it, thanks.
18:50:12*Loki15 quit (Read error: Connection reset by peer)
18:50:12*jgh- quit (Client Quit)
19:03:56*Tojil joined #nim
19:04:12*Tojil quit (K-Lined)
19:13:46*seni joined #nim
19:17:29*odc_ joined #nim
19:17:42*zielmicha__ joined #nim
19:18:28*Jesin quit (Remote host closed the connection)
19:20:07*Sargun_ joined #nim
19:25:17*epictek[m] quit (*.net *.split)
19:25:17*unclechu[m] quit (*.net *.split)
19:25:23*codevance[m] quit (*.net *.split)
19:25:24*dyce[m] quit (*.net *.split)
19:25:24*odc quit (*.net *.split)
19:25:24*zielmicha_ quit (*.net *.split)
19:25:26*Sargun quit (*.net *.split)
19:25:26*Jesin joined #nim
19:27:57*themagician quit (Ping timeout: 240 seconds)
19:29:04*nsf joined #nim
19:31:44*vivus joined #nim
19:32:48*epictek[m] joined #nim
19:33:04*miran joined #nim
19:34:31*codevance[m] joined #nim
19:34:39*dyce[m] joined #nim
19:34:45*unclechu[m] joined #nim
19:40:39FromGitter<zetashift> https://pastebin.com/pV9GpfWB why does this give me an undeclared identifier error for in method ready and method onTextEntered?
19:40:52FromGitter<zetashift> line 12 and 17
19:42:26*Cthalupa quit (Ping timeout: 256 seconds)
19:45:01*Cthalupa joined #nim
19:52:18*Cthalupa quit (Ping timeout: 260 seconds)
19:57:49*Cthalupa joined #nim
20:04:10*cspar joined #nim
20:17:47FromGitter<zetashift> whoops undeclared identifier error for story on line 12 and textbox for line 17* english is hard
20:19:32*vivus quit (Quit: Leaving)
20:22:51*yglukhov[i] quit (Ping timeout: 240 seconds)
20:26:51*OPK2 joined #nim
20:29:44*jhutchins14 joined #nim
20:29:48*OPK2 quit (Remote host closed the connection)
20:32:17*yglukhov[i] joined #nim
20:32:29*jhutchins14 quit (Remote host closed the connection)
20:57:14*miran quit (Ping timeout: 272 seconds)
21:01:05*cryptocat1094 quit (Quit: WeeChat 2.2)
21:02:38*felco25 joined #nim
21:03:12*felco25 quit (Remote host closed the connection)
21:06:54*nsf quit (Quit: WeeChat 2.2)
21:09:27*NimBot joined #nim
21:19:47*noonien quit (Quit: Connection closed for inactivity)
21:24:15*Trustable quit (Remote host closed the connection)
21:30:32FromGitter<zetashift> I guess the gdobj macro doesn't allow fields to be initialized outside of methods?
21:30:42FromGitter<zetashift> putting the prompts and story in the init method worked
21:32:26FromGitter<Clyybber> Are you using pragmagics godot bindings?
21:33:12FromGitter<zetashift> yea
21:43:33*ZetaDot joined #nim
21:44:08ZetaDotdummy question. How can I install jester? On the github page it says “Jester 0.3.0” So I cannot use it on a stable nim?
21:44:25ZetaDot*Starting with Jester 0.3.0, a devel version of Nim is required.
21:45:59FromGitter<kayabaNerve> You'd need to get an old version; I do believe dom backported the security updates to 0.2.0 which will work on jester
21:46:38*natrys quit (Ping timeout: 260 seconds)
21:46:58FromGitter<kayabaNerve> Yep. https://github.com/dom96/jester/releases/tag/v0.2.1
21:47:19FromGitter<kayabaNerve> I don't believe that has any of the 0.4.0 security updates though...
21:48:04ZetaDotso when will the 0.4.0 be usable?
21:48:25dom960.2.1 does have security fixes
21:48:37FromGitter<Clyybber> @zetashift Is it usable with godot devel?
21:49:03FromGitter<kayabaNerve> ZetaDot it's on the releases page right now
21:49:11FromGitter<zetashift> @Clyybber don't think so according to: https://github.com/pragmagic/godot-nim/issues/32
21:49:59zacharycarter[m]Araq: I'm sorry - I think there was some misunderstanding / miscommunication regarding the gist I shared earlier - I'm not so much concerned with the parsing and eventual compilation - this example SHOULD fail
21:50:35FromGitter<zetashift> @Clyybber man it's late I thought you were talking about Nim devel :P
21:50:44FromGitter<zetashift> it should work on godot devel I could give it a whirl if you'd like
21:50:50zacharycarter[m]what I was concerned with - is the fact that there are multiple pairs of idents/types specified, yet only the first pair is turned into arguments
21:51:10zacharycarter[m]why is godot-nim using nake to begin with?
21:51:29FromGitter<zetashift> to generate the godot api I think?
21:51:44zacharycarter[m]yeah - but nake isn't necessary anymore
21:51:47zacharycarter[m]use nimble + nimscript
21:52:13zacharycarter[m]no new project should be using Nake IMO
21:52:34FromGitter<Clyybber> @zetashift If you would like to test it that would be amazing
21:53:12zacharycarter[m]fowl isn't even acdtive anymore
21:53:16zacharycarter[m]active*
21:55:08FromGitter<zetashift> I dunno godotnim started late 2017
21:55:22FromGitter<zetashift> Also don't have any problems with the nake setup
21:59:56zacharycarter[m]fowl was last active several years ago
22:00:29zacharycarter[m]I've just heard several complaints in the last few days regarding nake and godot
22:00:57zacharycarter[m]I think yuiry wrote the bindings for godot, and I think he's a nake user - so that's probably why
22:01:07zacharycarter[m]IMO - the less dependencies the better
22:01:43FromGitter<zetashift> true!
22:04:00*natrys joined #nim
22:04:23FromGitter<zetashift> @Clyybber it runs
22:04:30zacharycarter[m]the less godot the better IMO too - but I'll shush about that
22:04:48FromGitter<zetashift> haha :P
22:05:04FromGitter<zetashift> that's just preferences, the learning resources are solid though and that's what I currently need
22:05:27FromGitter<Clyybber> zacharycarter Is yuiry an alias for pragmagic?
22:05:37AlexMaxIs there any way to tell what I'm doing wrong when my nim configuration file is invalid and I get "SIGSEGV: Illegal storage access. (Attempt to read from nil?)"?
22:05:44FromGitter<Clyybber> @zetashift Thank you
22:05:47AlexMaxThat's....not very descriptive
22:06:02zacharycarter[m]Oh pragmagic made the godot nim bindings? Nevermind then - that is not Yuriy
22:06:11FromGitter<Clyybber> Ah ok
22:06:22*francisl quit (Quit: francisl)
22:06:25zacharycarter[m]Also - it's more than just preferences - I don't like Godot / the community / the authors
22:06:33FromGitter<Clyybber> zacharycarter Why?
22:06:41zacharycarter[m]pragmagic is euantor but I thnk there are multiple devs associated w/ that account
22:06:52FromGitter<Clyybber> Interesting
22:06:55zacharycarter[m]They make nonsensical decisions / bold proclamations and don't back them up with any sound reasoning
22:07:18FromGitter<Clyybber> zacharycarter Could you provide an example?
22:07:23zacharycarter[m]sure
22:07:49zacharycarter[m]this is a great one - https://godotengine.org/article/why-does-godot-use-servers-and-rids
22:08:47*zacharycarter[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/RSliMjWqrLmbOJmVOhRwczQV >
22:09:33FromGitter<Clyybber> Hmm, ok thats a valid critique
22:09:43zacharycarter[m]the entire article is basically defending their ridicuous Server / RID abstraction with the essence that - it would be too hard to do otherwise while maintaining usiblity
22:09:45zacharycarter[m]which plenty of other game engines have done
22:09:48*stefanos82 quit (Quit: Quitting for now...)
22:09:53zacharycarter[m]it's just holes like that all over the place
22:10:19zacharycarter[m]not to mention - most of the community seems to be 12 year olds with zero programming experience
22:10:25zacharycarter[m]not that it's different with other game engines- but they don't become contributors
22:10:34FromGitter<Clyybber> Haha
22:10:40zacharycarter[m]just go look at the C++ source code forgodot
22:10:42zacharycarter[m]it's a hell hole
22:12:12FromGitter<Clyybber> What I dont like is the variant system
22:12:39FromGitter<Clyybber> But thankfully you dont have to deal with that
22:12:50FromGitter<Clyybber> unless you are using GDNative
22:13:04FromGitter<Clyybber> and it affects you in terms of performance
22:13:07AraqAlexMax, can you gist your config?
22:13:57zacharycarter[m]GDScript might as well die - it serves no purpose
22:14:09zacharycarter[m]but you can't convince the core devs of godot that
22:14:39zacharycarter[m]it's like telling your userbase - well you won't be able to code a game with this runtime because it sucks
22:15:38zacharycarter[m]so learn this syntax and write everything you think is asinine in this language, and then write all your code that needs to actually perform in another language
22:15:49FromGitter<Clyybber> Yeah I think moving away from GDScript would be a good decision
22:15:53zacharycarter[m]which begs the question - why do I even write anything in GDScript in the first place?
22:15:54AlexMaxAraq: Wait a minute, I think i was reading the wrong guide - I was on the niminst page
22:16:14zacharycarter[m]well, they don't think so :P
22:16:33*natrys quit (Ping timeout: 248 seconds)
22:16:34AlexMaxWhat options does the myproject.nim.cfg file accept?
22:16:35zacharycarter[m]the thing is - when your user base is 12 year olds trying to make vidya games
22:16:39AlexMaxSame as the compiler itself?
22:16:45Araqzacharycarter[m], because you want quick development iterations and C++'s compile times are considered too bad
22:17:00AraqAlexMax, all of the command line options, pretty much
22:17:00FromGitter<Clyybber> zacharycarter Then you gotta choose nim
22:17:38AlexMaxAraq: Do I get the ability to script my config? Like if I want a different config between windows and *NIX and mac?
22:17:50Araqof course.
22:17:53FromGitter<Clyybber> zacharycarter It just bothers me that the GDNative bindings(and the whole engine) are a bit hold back by supporting GDScript
22:17:58Araq@if windows: ... @end
22:18:00FromGitter<Clyybber> But maybe that will get better
22:18:16Araqbut it's probably better to use a myproject.nims nimScript file
22:18:25Araqthe old config system is kinda deprecated
22:18:42zacharycarter[m]Araq: So do what UE4 did and add visual scripting instead of just some asinine scripting language that you need to now learn an entirely new syntax / api for, so you can write less performant code
22:19:10Araqzacharycarter[m], Nim's VM is available as an API now anyway
22:19:12zacharycarter[m]I'd rather use lua than learn another language
22:19:30Araqsorry but Lua is pretty terrible to program in.
22:19:35*Cthalupa quit (Ping timeout: 272 seconds)
22:19:39FromGitter<Clyybber> There are some ultra experimental lua bindings for godot
22:19:40Araqevery typo leads to a nil bug
22:20:03Araqand they merged lists with hash tables (WTF?)
22:20:14AlexMaxI don't think they "merged" them
22:20:14FromGitter<Clyybber> Araq Yeah, but with a good IDE like zerobrane its pretty easy
22:20:20zacharycarter[m]like - don't make me learn your bs proprietary scripting language to use your product
22:20:24AlexMaxmore "emulated lists with tables"
22:20:31zacharycarter[m]it sucks in every sense of the word
22:20:54AlexMaxalso, lua has a few weird parts to it, but it's one of my favorite embeddable scripting languages
22:20:57*data-man quit (Quit: Konversation terminated!)
22:20:57Araqdon't lecture me, AlexMax, I studied its C code before I started to write Nim
22:21:09AlexMaxAraq: Sorry, I didn't mean to offend D:
22:21:29Araqok, no worries
22:21:42zacharycarter[m]Araq: if lua is terrible to program in, how bad do you think GDScript is?
22:21:43zacharycarter[m]one is half-baked
22:21:43FromGitter<Clyybber> Araq I think lua is very elegant, in that it has only one real container type
22:21:43zacharycarter[m]and only usable in a single application
22:21:44*Cthalupa joined #nim
22:21:55zacharycarter[m]GDScript is like a bastardized python
22:22:39AraqClyybber: I think that's not elegant at all, it's a hack. "Everything must be as simple as possible, but not simpler" Lua is too simple.
22:23:06zacharycarter[m]I hate Lua too - but I mean, at least it's a convention
22:23:29zacharycarter[m]like there's plenty of C++ game engines already using Lua as a scripting language, and succeeding with it
22:23:43Araqand its stack based API is terrible too, but at least it doesn't expose its GC implementation to the clients (hello Python, Ruby...)
22:23:46zacharycarter[m]and I don't dislike Unity's approach of baking in a managed runtime
22:24:16zacharycarter[m]but come on - force me to learn a new scripting language for your game engine? This is why UE3 was such a disaster and UE devs just went C++ or visual scripting
22:24:17FromGitter<Clyybber> Araq The only downside of that hack is performance really
22:24:27FromGitter<Clyybber> and the usual downsides of dynamic typing
22:24:37Araqno, it's also error prone
22:24:47FromGitter<Clyybber> Yeah
22:24:49Araqyou can't use the hash and list parts at the same time
22:24:57FromGitter<Clyybber> You can
22:25:06FromGitter<zetashift> I think GDScript is alright. And Godot is working on a VisualScript implementation. I think Godot received it's fair amount of backlash for GDScript and that's why they added GDNative and C#
22:25:10FromGitter<Clyybber> Araq
22:25:12Araqunless you can ensure that e.g. '_n' is not used as a key because that would screw up the list part
22:25:32Araqwell that was for Lua 5.x
22:25:33AlexMaxno kidding, every time I've ever had to do something with Lua in C, I've had to keep comments about what's on the stack in comments, like I was writing assembly.
22:25:34FromGitter<Clyybber> Araq What list part do you mean? Do you mean the array part?
22:25:39Araqmaybe they fixed that one...
22:25:54Araqyeah array part, list part, no difference for me
22:26:41AlexMaxIf Lua is bad to embed, and Python is worse, what's good?
22:26:59FromGitter<Clyybber> AlexMax Nothing
22:26:59zacharycarter[m]> <@freenode_FromGitter:matrix.org> <zetashift> I think GDScript is alright. And Godot is working on a VisualScript implementation. I think Godot received it's fair amount of backlash for GDScript and that's why they added GDNative and C#
22:26:59Araqit's not bad to embed.
22:27:00zacharycarter[m]So get rid of GD script then :P
22:27:02FromGitter<zetashift> Personally if I ever get to git gut in Nim I'd like to write bindings to Kha: http://kha.tech/ using their C api: https://github.com/Kode/Kore / https://github.com/Kode/Kha/tree/master/Backends/KoreHL/KoreC
22:27:32zacharycarter[m]I don't think that will work
22:27:37FromGitter<zetashift> probably not
22:27:38Araqit's not bad to embed, they got that part right. unfortunatley the *language* is bad
22:27:41FromGitter<zetashift> A man can dream
22:27:45zacharycarter[m]isn't kha based on that actionscript equivalent?
22:27:49AlexMaxOh? I misunderstood you then. I thought you said that you didn't like the stack-based C API?
22:27:58zacharycarter[m]haxe or whatever?
22:28:07FromGitter<zetashift> well the base is C and they (indeed) use haxe on top of it
22:28:28AraqAlexMax, the C API is bad, but better than most others
22:28:30zacharycarter[m]meh - just ditch haxe
22:28:43FromGitter<zetashift> exactly Nim would be a way better fit :D
22:28:47AlexMax*nod*
22:29:03FromGitter<Clyybber> Araq I think lua doesn't use _n anymore to store the length
22:29:45Araqyeah, they now use a hidden field and have the # operator to retrieve the length
22:29:47zacharycarter[m]> <@freenode_FromGitter:matrix.org> <zetashift> exactly Nim would be a way better fit :D
22:29:48zacharycarter[m]I see what you're saying - go from whatever language to Nim instead of C - yeah that'd be interesting but would probably require a complete rewrite of haxe I imagine
22:30:26zacharycarter[m]I'd just write your game / engine in Nim tbh
22:30:48zacharycarter[m]unless you're doing 3d - you shouldn't need something like godot anyway
22:31:02FromGitter<zetashift> I sortof am aren't I. I finished the SDL2 tutorial in Nim 2 weeks ago
22:31:19FromGitter<zetashift> and now tinkering with Godot agian seeing which I like more
22:31:19FromGitter<Clyybber> Now try vulkan
22:31:21FromGitter<Clyybber> jk dont
22:31:38FromGitter<zetashift> I'd like to finish a game ty
22:31:55FromGitter<Clyybber> Haha me too. Now I cant
22:32:11zacharycarter[m]I'm working on a new re-write of frag - I hope to produce something comparable to Lumix engine or Ethereal engine
22:32:35zacharycarter[m]what do you want to do? game programming or engine programming? - you don't need to learn an engine to learn how to write a game
22:32:56zacharycarter[m]IMO some of the more complex games are rougelikes - written without any type of engine at all for the most part
22:33:41zacharycarter[m]using a game engine often means you need someone to make graphics / sounds / etc and you will get lost in just learning the engine and staring at a lack of resources
22:33:56FromGitter<zetashift> I got to putting the @ on screen, moving it around and generating the map with nim + pdcurses
22:33:57zacharycarter[m]but you can certainly make a worthwhile game with just a terminal and programming knowledge and creativity
22:34:13zacharycarter[m]and forego having to learn a gam eengine
22:34:21FromGitter<Clyybber> ncurses is ironically a blessing
22:34:46zacharycarter[m]https://github.com/zacharycarter/nimrl
22:34:48FromGitter<zetashift> than I gave up cause I think way too OO and I have trouble with composition in Nim
22:34:59zacharycarter[m]render that with OpenGL
22:34:59FromGitter<zetashift> that map generator is smmmooootthh
22:35:24zacharycarter[m]https://github.com/zacharycarter/mapgen
22:35:34zacharycarter[m]you can do a lot without a real game engine
22:35:49FromGitter<Clyybber> zacharycarter I hate when roguelikes that use tiles, do it in an engine or a custom terminal
22:36:00FromGitter<Clyybber> Because then you cant play over ssh
22:36:22AlexMaxdcss does both
22:36:35zacharycarter[m]well that's the king of RLs
22:36:36FromGitter<Clyybber> It has tiles and they are fantastic
22:36:43AlexMaxwow holy cow
22:36:44FromGitter<Clyybber> CDDA does it like that too
22:36:47zacharycarter[m]but it didn't start that way
22:36:49AlexMaxHow did nim find my clang installation?
22:36:54AlexMaxwithout me having to do anything
22:36:56AlexMaxthat's pretty slick
22:36:57FromGitter<Clyybber> Magic
22:37:01zacharycarter[m]are you on OSX?
22:37:04AlexMaxNo, Windows
22:37:10zacharycarter[m]no idea then
22:37:24FromGitter<zetashift> nim is packed with a trojan horse
22:37:33FromGitter<Clyybber> Also terminal roguelikes can be run through lolcat...
22:38:17zacharycarter[m]I really want to make a MUD in Nim
22:38:18*Taylor_ joined #nim
22:38:27AlexMaxman does anybody play muds anymore heh
22:38:28zacharycarter[m]but no time for that
22:38:41FromGitter<Clyybber> AlexMax I would
22:38:47FromGitter<Clyybber> If others would
22:38:48*Taylor_ quit (Client Quit)
22:38:51zacharycarter[m]definitely
22:39:07*PrimHelios quit (Quit: Leaving)
22:39:38*PrimHelios joined #nim
22:39:49zacharycarter[m]https://github.com/shawncplus/ranviermud
22:40:41*xet7 quit (Quit: Leaving)
22:40:51AlexMaxoh, clang must've put itself in my PATH
22:40:55AlexMaxdream world shattered
22:54:14AlexMaxhrm, is there shorthand to `doc` a library installed by nimble?
22:57:17AlexMaxI did go into the .nimble directory and did it myself
23:03:12FromGitter<Clyybber> How would I get the function pointer of a nim function for using it as a callback for a c API?
23:04:08*seni quit (Quit: Leaving)
23:10:14*salios joined #nim
23:10:19*salios quit (Remote host closed the connection)
23:23:17zacharycarter[m]> <@freenode_FromGitter:matrix.org> <Clyybber> How would I get the function pointer of a nim function for using it as a callback for a c API?
23:23:18zacharycarter[m]I think you would pass it as a pointer to the C function
23:23:28*r3m19 joined #nim
23:23:56zacharycarter[m]and I think you have to annotate the proc with the `{.cdecl.}` pragma
23:24:05zacharycarter[m]see - https://forum.nim-lang.org/t/1801
23:24:24AlexMaxyay, I got my little sdl2 + opengl program working in windows
23:24:38AlexMaxunfortunately, there is a problem
23:24:51AlexMaxhidpi scales the resolution up
23:25:03AlexMaxso the resolution number I pass SDL2 is a lie, a sweet little lie
23:26:15zacharycarter[m]AlexMax: sounds like a SDL2 issue
23:26:30*r3m19 quit (Remote host closed the connection)
23:27:28AlexMaxzacharycarter[m]: It is a shortcoming in SDL2, but I've solved it before.
23:27:35AlexMaxhttps://bugzilla.libsdl.org/show_bug.cgi?id=3281
23:28:19AlexMaxwell, not solved...more like "worked around"
23:28:32zacharycarter[m]well whatever you did previously should be doable in Nim I would imagine
23:29:03AlexMaxAt least when doing C/C++, you have to pass an .rc file that points to a mainfest.xml file that tells Windows not to scale the window, that the application will handle scaling
23:29:18AlexMaxhttps://github.com/chocolate-doom/chocolate-doom/blob/master/src/manifest.xml
23:29:24zacharycarter[m]why can't you do that with Nim?
23:29:39AlexMaxMaybe I can. Not sure how, though.
23:31:10zacharycarter[m]looks like a huge / gigantic hack to make SDL2 behave properly on windows
23:31:18zacharycarter[m]is there any strong reason you're using SDL2? Are you developing for mobile?
23:31:39zacharycarter[m]if you're just using SDl2 for desktop windowing - you may want to checkout GLFW3
23:31:59AlexMaxI'm used to it, and I've used it before.
23:32:04zacharycarter[m]https://github.com/chocolate-doom/chocolate-doom/blob/master/src/setup-res.rc.in - looks sketchy
23:32:51zacharycarter[m]Maybe use GLFW3 then? If you're not taking advantage of the SDl2 drawing / other APIs - I don't see much sense in using SDl2
23:33:00zacharycarter[m]I generally only opt for SDL2 when I need to target mobile
23:33:05zacharycarter[m]and then there's CSFML and Allegro
23:33:13zacharycarter[m]which are much lighter weight than SDL2
23:33:31zacharycarter[m]and both support mobile
23:33:34FromGitter<zetashift> really even CSFML?
23:33:34zacharycarter[m]and OpenGL drawing contexts
23:33:59zacharycarter[m]if all you need is an opengl context sure - I generally don't even use OpenGL - I use BGFX
23:34:42zacharycarter[m]most of the content of these libraries is just cross platform system / threading / windowing / etc code
23:34:45FromGitter<zetashift> bgfx is probably better now that Apple is gonna stop supporting OPENgl
23:35:10zacharycarter[m]yeah - another point I raised with the Godot devs and was told off on
23:35:24zacharycarter[m]have fun writing your own Metal / Vulkan backends 😇
23:36:19AlexMaxBuilding bgfx on Windows is untested.
23:36:22AlexMaxwelp
23:36:29zacharycarter[m]wut?
23:36:37zacharycarter[m]no it's not
23:36:37AlexMaxfrom nim-bfgx's readme
23:36:49zacharycarter[m]don't use those bindings
23:36:50AlexMaxI meant to put a > in front of that
23:37:18zacharycarter[m]I'll have more up to date ones along with CI - https://github.com/zacharycarter/bgfx.nim - shortly
23:37:35zacharycarter[m]those bgfx bindings don't work - the author tried to assimilate the C++ api and failed
23:37:48zacharycarter[m]I just target the C99 api - but my bindings need updating as well
23:39:12zacharycarter[m]I'll work on that now
23:39:13FromGitter<zetashift> @zacharycarter iirc they're gonna target Vulkan and Metal through MoltenVK
23:39:53zacharycarter[m]I'm sure they will
23:40:15zacharycarter[m]and I'm sure they'll do something to completely screw it up
23:41:50zacharycarter[m]when you have a dev team working on a game engine claiming they won't do things because they're too complicated
23:42:04zacharycarter[m]but they're things that every other modern game engine exhibits
23:42:11zacharycarter[m]I start to lack trust in the team developing said game engine
23:42:16zacharycarter[m]even if I don't know how to do it
23:44:37FromGitter<zetashift> you really don't like the way they handle the engine huh
23:44:42zacharycarter[m]nope
23:45:09zacharycarter[m]a game engine should not be run by its community
23:45:33zacharycarter[m]most game engine users are not game engine programmers - not that I am either
23:45:40*francisl joined #nim
23:45:56zacharycarter[m]but it seems like there is a lot of considerations for people that have no game programming experience / very little - weaved into godots design decisions
23:46:09zacharycarter[m]in fact the core devs apologize for it a lot of the time
23:46:37zacharycarter[m]so who are we trying to appeal to here - people that want to make games? or 12 year olds that think gamemaker is too hard?
23:46:45zacharycarter[m]godot seems like it wants to appease the latter
23:47:56FromGitter<zetashift> I have no idea about the inner workings of the community I mainly use godot cause I can use Nim with it
23:48:16FromGitter<zetashift> and the docs were pretty noice
23:48:46zacharycarter[m]yeah - I mean godot has been around for a while and was commercial before it went OS so it's had a lot of hours poured into it
23:48:58FromGitter<zetashift> nice*, however I do hope to get some 3D stuff up, I don't know if I'll use Godot
23:49:03FromGitter<zetashift> https://armory3d.org/ is looking interesting
23:49:23FromGitter<zetashift> but if frag has an alpha by than I'll try it the spartan way why not
23:49:29zacharycarter[m]:)
23:50:01zacharycarter[m]armory is interesting - you have to wonder why they chose blender as an IDE though
23:50:12zacharycarter[m]I'm not sure if anything capable yet has been built w/ armory
23:51:09FromGitter<zetashift> no definitely not anything capable. That's why it's interesting for now. The dude is posting some neat stuff with it on twitter; https://twitter.com/luboslenco
23:52:47zacharycarter[m]yeah - hopefully it gets beyond that stage, I think it's definitely a better option vs Godot
23:53:03AlexMaxoof, getting a lot of errors trying to `nim doc glm`
23:53:40AlexMaxhttps://paste.ee/p/jLCwS
23:53:47FromGitter<zetashift> using haxe though ;(
23:54:48zacharycarter[m]yeah - but I have a feeling that was the only way to attack blender as an editor
23:55:02zacharycarter[m]I don't want my modeling program mixed with my game editor though
23:56:14zacharycarter[m]I mean - https://github.com/Kode/Kha
23:56:17zacharycarter[m]look at the code distribution there
23:56:23zacharycarter[m]it's just a haxe wrapper on top of C
23:56:58zacharycarter[m]so just sub kha with bgfx
23:57:01zacharycarter[m]and bgfx is probably more battle tested / capable than kha
23:57:07zacharycarter[m]considering a AAA graphics dev develops BGFX
23:57:11zacharycarter[m]and he knows what he's doing
23:57:37zacharycarter[m]also commercial games are built on BGFX
23:57:38FromGitter<zetashift> oh yea they fill the same niche
23:57:54FromGitter<zetashift> both got alot of attention when Apple announced their stuff about opengl
23:58:16FromGitter<zetashift> only thing was when I looked at the bgfx samples I understood absolutely nothing
23:58:39zacharycarter[m]bgfx predates armory by quite a bit
23:59:00zacharycarter[m]bfgx does more than just abstract away the APIs
23:59:10zacharycarter[m]it also does draw call bucket sorting
23:59:24zacharycarter[m]and other optimizations behind the hood when draw calls are made
23:59:53zacharycarter[m]it also has a C API which makes it very consumable