00:14 <aiwakura> related to my previous question: is there a better way, currently, to test for union(enum) equality?

00:14 <aiwakura> http://paste.debian.net/plain/1014834

00:41 <andrewrk> aiwakura, would you expect it to codegen to the eql function that you have created?

02:07 <aiwakura> andrewrk, yeah - two union(enum) are equivalent if both the enum value and the payload value are the same

02:08 <aiwakura> is there an alternative definition?

02:45 <andrewrk> yes. for example in the zig compiler the function to test if 2 comptime values are equal is much more nuanced

02:46 <andrewrk> the existence of == would only serve to be a footgun

02:56 <aiwakura> Can you give me an example of an ambiguous case?

03:09 <andrewrk> for example if you have a pointer field, and you want to dereference the pointer to determine equality

03:10 <andrewrk> this is the same problem as struct equality

03:11 <aiwakura> A pointer is a value - its value is the address

03:12 <aiwakura> It seems you don’t want people to think that == performs a deep comparison

03:13 <aiwakura> What about a @compare builtin?

03:14 <aiwakura> With the same semantic I’m defining in the example

03:14 <aiwakura> (No deep comparison)

03:17 <andrewrk> we could provide it in the standard library

03:17 <andrewrk> with reflection

03:18 <aiwakura> So we don’t need to build the switch cases by hand?

03:18 <aiwakura> Yeah I like that

03:57 <GitHub15> [zig] andrewrk pushed 1 new commit to master: https://git.io/vxLml

03:57 <GitHub15> zig/master 50e25f6 Andrew Kelley: add missing docs for setAlignStack builtin

04:00 <andrewrk> 0.2.0 release is locked and loaded for tomorrow morning

04:00 <andrewrk> hopefully we can get in the news

04:19 <clebermatheus[m]> Hello

04:36 <dtz> \o/

04:41 <aiwakura> yay!

05:49 <zwhadok_> THIS IS A FREENODE BREAKING NEWS ALERT!! Hitechcg AND opal ARE GOING AT IT RIGHT NOW WITH A LOT OF FIGHTING AND ARGUING WOW YOU DON'T WANT TO MISS THIS!! TYPE /JOIN ## TO SEE THE ACTION...AGAIN TYPE /JOIN ## TO SEE THE ACTION!!

05:49 <zwhadok_> clebermatheus[m] jdufault Tobba return0e meena tridactyla DuClare ltr_ _whitelogger lorde_kumamon[m] alandipert commander monteslu aiwakura MajorLag minus l1x walac jacobdufault dtz mal`` ragge m6w6 lqd cgag Vins

05:50 <aiwakura> wat

06:44 <lorde_kumamon[m]> IRC spam

10:57 <jfo> MajorLag: good point. I've added a not at the end that this is the case linking to andrewrk's og post about the rationale. I've also added a whole section about the error inference, which I think makes a good climax to the post. there's a lot to cover and I've left a lot out I wanted to get to, but I think this is pretty good as a really rudimentary intro! I hope people get something out of it :)

10:58 <jfo> also I've called out please join #zig on irc if you have any questions because I guess we can expect to see some support requests incoming post 0.2.0 anyway

10:58 <jfo> aiwakura: thanks!

11:29 <Hejsil> jfo, Nice post. I like your recent addition of the error sets and why they are useful

11:31 <Hejsil> But it's missing that 'comptime' flare :)

13:09 <jfo> Hejsil: I think I'm going to cover that in another post, there's so much ground to cover! but I want to reimplement the seekforward functionality using a stack and then making that generic with comptime. was hoping to get that into this post but it got so long and I ran out of time...

13:15 <GitHub119> [zig] andrewrk pushed 1 new commit to master: https://git.io/vxLdN

13:15 <GitHub119> zig/master f073923 Andrew Kelley: Release 0.2.0

13:17 <jfo> ooooooh snappppp

13:17 <jfo> boom

13:38 <jfo> https://github.com/Homebrew/homebrew-core/pull/25320

14:07 <cgag> just getting set up again after not playing with zig in quite a while, built ok but build.zig test is failing: `0x23ed94 in ??? (build): return error.InvalidArgs;`

14:07 <cgag> is that expected at the momentr?

14:34 <andrewrk> cgag, probably not. can you give more context?

14:38 <cgag> https://www.irccloud.com/pastebin/XSDlrrAW/

14:39 <cgag> almost certainly related to this from cmake, i was trying to follow the wiki about building llvm

14:39 <cgag> in ~/local, which seemed to go fine, then copied the command from the readme and set -DCMAKE_INSTALL_PREFIX=$HOME/local

14:43 <andrewrk> cgag, it looks like a conflict between the llvm you've built and system libs

14:43 <andrewrk> what linux distro is this?

14:43 <cgag> is there something I need to set somewhere to force it to prefer $home/local

14:43 <cgag> fedora 27

14:44 <andrewrk> I thought that CMAKE_PREFIX_PATH would do that

14:48 <andrewrk> I posted the 0.2.0 release notes to /r/programming. https://www.reddit.com/r/programming/new/

14:58 <jordyd> Does Zig use semantic versioning?

14:59 <jordyd> For the project itself I mean

14:59 <andrewrk> jordyd, yes, starting at 1.0.0

14:59 <andrewrk> until then, minor bumps can break everything

14:59 <jordyd> Good to know

14:59 <jordyd> I like those labelled loops btw

15:00 <andrewrk> glad to hear :)

15:05 <cgag> Hmm, I had CMAKE_INSTALL_PREFIX set, added CMAKE_PREFIX_PATH=$HOME/local/lib and cmake exits cleanly, but make itself is failing:

15:05 <cgag> https://www.irccloud.com/pastebin/zubyVwjP/

15:06 <cgag> seems strange that only libclangFrontend is a problem

15:08 <jordyd> Is there support--or plans for support--for examining the AST programmatically

15:09 <andrewrk> jordyd, this is intentionally not possible

15:10 <andrewrk> so that someone reading zig code can be confident that this is never done in the code they are looking at

15:10 <jordyd> Well, I don't mean for the purpose of code generation

15:11 <jordyd> For a Frama-C-like use case

15:11 <andrewrk> oh this is IDE integration type of thing?

15:11 <jordyd> For the purpose of static analysis, verification, etc.

15:12 <andrewrk> the current plan is to have a language server that provides deep introspection into a compilation, for the purposes of static analysis, IDEs, and things like this

15:12 <andrewrk> and it's planned to only implement this in the self hosted compiler, not the c++ compiler

15:12 <jordyd> Neat

15:14 <jfo> cgag: thanks for your comment and please lmk if anything is unclear to you in how I've written that post

15:17 <cgag> will do once I get this libclang nonsense sorted out. Maybe I'll give up on this ~/local stuff and just install libclang where cmake expects it to be

15:18 <jfo> good luck

15:21 <jordyd> Is there any kind of Doxygen-Like thing?

15:23 <andrewrk> jordyd, https://github.com/zig-lang/zig/issues/21

15:23 <andrewrk> this is another thing that will only be in the self hosted compiler