Newest at the top
| 2025-01-22 11:43:49 +0100 | __monty__ | (~toonn@user/toonn) toonn |
| 2025-01-22 11:42:59 +0100 | j1n37 | (~j1n37@user/j1n37) j1n37 |
| 2025-01-22 11:42:58 +0100 | j1n37- | (~j1n37@user/j1n37) (Ping timeout: 272 seconds) |
| 2025-01-22 11:41:28 +0100 | xff0x | (~xff0x@2405:6580:b080:900:55d6:5b69:524c:1857) |
| 2025-01-22 11:40:52 +0100 | hellwolf | retiring from this thread, I guess I got lost somewhere |
| 2025-01-22 11:40:21 +0100 | <hellwolf> | huh, why plan9 was brought up. |
| 2025-01-22 11:39:28 +0100 | <homo> | kuribas 1. for clean bootstrap something other than ghc must be able to convert ghc language into c language, 2. practically plan9 is so different that you'll need to rewrite generated c code by hand |
| 2025-01-22 11:38:18 +0100 | alfiee | (~alfiee@user/alfiee) (Ping timeout: 245 seconds) |
| 2025-01-22 11:37:04 +0100 | tzh | (~tzh@c-76-115-131-146.hsd1.or.comcast.net) (Quit: zzz) |
| 2025-01-22 11:36:54 +0100 | tnt1 | (~Thunderbi@user/tnt1) tnt1 |
| 2025-01-22 11:36:38 +0100 | tnt1 | (~Thunderbi@user/tnt1) (Ping timeout: 245 seconds) |
| 2025-01-22 11:36:29 +0100 | <homo> | just to port microhs to plan9 I'll still have to rewrite combinator virtual machine to align with plan9, and on top of that I'll need to get rid of all posix and ffi dependencies in haskell code |
| 2025-01-22 11:35:51 +0100 | <merijn> | kuribas: Only kinda |
| 2025-01-22 11:35:00 +0100 | j1n37 | (~j1n37@user/j1n37) (Ping timeout: 276 seconds) |
| 2025-01-22 11:34:24 +0100 | j1n37- | (~j1n37@user/j1n37) j1n37 |
| 2025-01-22 11:34:20 +0100 | <homo> | merijn and practically microhs "not being haskell" is irrelevant if all I want is to have compiler for lazy statically-typed functional programming language on plan9, even if ghc was on plan9 I would unlikely be able to build more than 1% of packages on stackage, because plan9 is neither posix nor microsoft windows |
| 2025-01-22 11:34:18 +0100 | <kuribas> | Can you not compile ghc to C, and bootstrap from there? |
| 2025-01-22 11:34:12 +0100 | alfiee | (~alfiee@user/alfiee) alfiee |
| 2025-01-22 11:30:33 +0100 | EvanR | (~EvanR@user/evanr) (Ping timeout: 252 seconds) |
| 2025-01-22 11:30:24 +0100 | tnt2 | tnt1 |
| 2025-01-22 11:30:23 +0100 | tnt1 | (~Thunderbi@user/tnt1) (Ping timeout: 245 seconds) |
| 2025-01-22 11:29:14 +0100 | j1n37 | (~j1n37@user/j1n37) j1n37 |
| 2025-01-22 11:29:09 +0100 | tnt2 | (~Thunderbi@user/tnt1) tnt1 |
| 2025-01-22 11:27:51 +0100 | j1n37 | (~j1n37@user/j1n37) (Ping timeout: 276 seconds) |
| 2025-01-22 11:25:34 +0100 | <homo> | the only thing that can be worse than ghc in terms of quality is zig, because to bootstrap zig you need to rebuild zig 0.10.0 32 times and zig 0.11.0 16 times, all those 48 are slightly different from each other, but broken enough to need step-by-step source code adjustments after every build |
| 2025-01-22 11:24:39 +0100 | <merijn> | Even UHC is questionable, since it presumably can't compile large parts of Hackage |
| 2025-01-22 11:24:21 +0100 | <merijn> | I mean, at this point anything that's not GHC isn't Haskell in any practical sense, tbh. |
| 2025-01-22 11:23:47 +0100 | <homo> | merijn don't bother, like I said earlier my goal is to get haskell to plan9 and I am satisfied with microhs |
| 2025-01-22 11:23:42 +0100 | <merijn> | Who cares, how often do you think new platforms get bootstrapped? |
| 2025-01-22 11:23:40 +0100 | <geekosaur> | you want to make life hard for yourself, go right ahead. don't demand everyone else bow to your insistence |
| 2025-01-22 11:23:30 +0100 | <merijn> | "It takes time and is annoying to do"? |
| 2025-01-22 11:23:22 +0100 | <merijn> | homo: So? What's the problem with bootstrapping via ancient versions? |
| 2025-01-22 11:22:44 +0100 | <homo> | geekosaur thank goodness I'm not stakeholder, but that was reply to suggestions such as building from ancient ghc or doing cross-compiling |
| 2025-01-22 11:22:32 +0100 | <geekosaur> | C doesn't have these problems because C doesn't change that redically |
| 2025-01-22 11:22:28 +0100 | <merijn> | homo: ok, so suppose we spend 5 or so man-years to make GHC 9 be compilable with GHC 4. Then what? What exactly is the benefit of that massive time investment? |
| 2025-01-22 11:21:38 +0100 | <geekosaur> | no, ghc is not going to be downgraded so you can build 9.12 with ghc 4 |
| 2025-01-22 11:21:14 +0100 | <geekosaur> | - to a lesser extent, pedagogical - teaching |
| 2025-01-22 11:20:56 +0100 | fmira | (~user@user/fmira) fmira |
| 2025-01-22 11:20:55 +0100 | <geekosaur> | - industry users (yes, they very much exist _and_ drive ghc development) |
| 2025-01-22 11:20:48 +0100 | zmt01 | (~zmt00@user/zmt00) (Ping timeout: 272 seconds) |
| 2025-01-22 11:20:39 +0100 | <geekosaur> | - PL researchers |
| 2025-01-22 11:20:35 +0100 | <geekosaur> | the primary stakeholders are: |
| 2025-01-22 11:20:31 +0100 | j1n37 | (~j1n37@user/j1n37) j1n37 |
| 2025-01-22 11:20:27 +0100 | <geekosaur> | you say this like you think you're ghc's primary stakeholder |
| 2025-01-22 11:19:36 +0100 | fmira | (~user@user/fmira) (Remote host closed the connection) |
| 2025-01-22 11:16:54 +0100 | swamp_ | (~zmt00@user/zmt00) zmt00 |
| 2025-01-22 11:16:42 +0100 | j1n37 | (~j1n37@user/j1n37) (Ping timeout: 252 seconds) |
| 2025-01-22 11:15:27 +0100 | tromp | (~textual@92-110-219-57.cable.dynamic.v4.ziggo.nl) |
| 2025-01-22 11:12:55 +0100 | <homo> | and 7.x gap got solved last month by Googulator, it's not funny that ghc 7.x can't build ghc 7.x, literally same version of itself |
| 2025-01-22 11:09:44 +0100 | alfiee | (~alfiee@user/alfiee) (Ping timeout: 252 seconds) |