=== heathkid|2 is now known as heathkid === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [05:14] awafaa: pretty much === chihchun is now known as chihchun_afk === doko_ is now known as doko [11:13] suihkulokki: thanks for the confirmation, so in your view what langs are high/medium/low priority? [11:16] awafaa: c,c++,java/go,gchi,erlang,ocaml/everything-else, intentionally leaving out purely interpreted languages that seem to generally Just Work (perl, python, etc). [11:16] awafaa: And probably missing a lot of random favourites from others. :P [11:17] infinity: who asked you?! :P [11:17] infinity: i was thinking more of what's the priority for langs that need porting? [11:17] awafaa: I figured our relationship was intimate enough that I could interject. I have photos to prove it. [11:17] infinity: you seen my outfit for the next connect? [11:17] awafaa: I'm not sure I want to... [11:18] awafaa: So, with a Canonical hat on, I'd say golang would be a high priority, taking that hat off, a proper ghci port would make a lot of nerds happy. [11:18] infinity: guess what? you're going to https://plus.google.com/u/0/103092666279088875227/posts/aCtNPvQdHt1 [11:18] awafaa: a native erlang port wouldn't go amiss, but I've also not (yet) heard people complaining about the performance impact of not having it. [11:19] awafaa: Dude, if your junk can maintain that contraption in place through an entire evening, more power to you. [11:19] infinity: so far the only high priority one i have is go [11:21] awafaa: There are also stragglers, like fpc, that don't have a ton of users but would probably also take someone with the right intersection of skills (pascal and ARM, in fpc's case) an afternoon to fix. [11:22] awafaa: So, not a priority at all, but perhaps a fun project for someone. [11:22] awafaa: my personal opinion is that it's quite low - but it would be still good to have someone assigned to working on porting/optimizing more esoteric languages [11:23] awafaa: Oh, and v8/node! I understand there's an AArch64 port sitting on tip/trunk/head/whatever somewhere, but that's of little use while distros are still shipping an old stable libv8 and node to match. [11:23] yeah, problem is getting resources to assign to do the work - I have very little to play with so I have to be very picky with what I put it on [11:24] infinity: you need to take that up with joyent [11:24] chasing the long tail and all - it should be made sure at least the languages work so nobody doesn't skip buying and armv8 server because they happen to have a legacy freepascal app as part of their system [11:24] suihkulokki: I doubt anyone would make a purchasing decision based on pascal, to be fair. [11:24] But nodejs, definitely. And saying "there will be a new upstream release some day with support" doesn't cut it. [11:25] suihkulokki: I agree, but I have to prioritise - I have the list of things that need work, I just need to prioritise it [11:25] awafaa: What does joyent have to do with it? [11:27] infinity: they have an exceptional amount of say in node - they just canned the CLA, but still have a lot to say [11:27] awafaa: "porting" nodejs to a new arch is trivial. libv8, on the other hand, is painful. And last time I asked about backporting aarch64 support to libv8-3.14 (ie: the version most everyone ships), I was told it was "too much effort", and thus no one cared. Which seems like a pretty lousy thing to tell potential customers. :P [11:29] And sure, some day there will be a node/v8 combo based on 3.20 or 3.22 or something, but that doesn't help me today. [11:30] it's in node.js master - * v8: upgrade to 3.24.35.22 [11:31] suihkulokki: So, maybe I'll see that release before 16.04 ... I'd still love to see a backport to 3.14 for 14.04 [11:32] But, *shrug*... [11:33] they don't seem to have release schedule [11:33] Anyhow, I'm not going to lose any sleep over it. Just a bit disappointing. [11:34] oddly I'd assume something as hipster and trendy as node being released monthly or so [11:34] They don't do the hipster release thing. [11:34] They also maintain stable branches for effin' ever. [11:34] Which I appreciate, as a stuffy conservative engineer. [11:35] But their reluctance to cut a new stable out of master is also irksome. :P [11:37] I think theyll still manage to release faster than awafaa is able to find someone to do the backport of v8 :F [11:38] almost definitely :/ [11:40] Perhaps a fair point. [11:41] awafaa: So, discounting the node/v8 argument, and understanding that you already have people prioritising golang (or, you might do), I'd say that ghci is pretty widely-used and often-ignored. [11:42] infinity: ocaml is ported [11:43] infinity: i've been trying to work out what actually uses ghci? and when yoiu guys were porting ghc, why do a half arsed effort? :P [11:43] hrw: I know, I was listing priorities period, before he narrowed it to discussing unported things. :P [11:43] ok [11:43] awafaa: Colin's GHC port isn't native, it's more just a dirty hack. [11:43] http://www.chiark.greenend.org.uk/ucgi/~cjwatson/blosxom/2014/04/15 [11:44] ah, that's not what all the noise said - thanks for the clarification [11:48] https://ghc.haskell.org/trac/ghc/wiki/Platforms === dk is now known as Guest61140 === chuck__ is now known as zul === mhaberler_ is now known as mhaberler