Newest at the top
2024-04-29 22:41:17 +0200 | <tomsmeding> | the worst part about accelerate's nofib is that compiling the test suite takes ages |
2024-04-29 22:38:48 +0200 | <tomsmeding> | https://packdeps.haskellers.com/reverse/QuickCheck |
2024-04-29 22:37:54 +0200 | <shapr> | then I can clone all the repos and use ripgrep to see how many modules depend on Hedgehog or QuickCheck |
2024-04-29 22:37:26 +0200 | <tomsmeding> | iirc people have done that before |
2024-04-29 22:37:23 +0200 | <shapr> | I want to find all direct depends on QuickCheck or Hedgehog, and extract the repository value |
2024-04-29 22:37:20 +0200 | <tomsmeding> | download all the sources and grep |
2024-04-29 22:37:02 +0200 | <shapr> | I wish I could write a SQL query against hackage |
2024-04-29 22:36:49 +0200 | tri | (~tri@ool-18bbef1a.static.optonline.net) (Ping timeout: 272 seconds) |
2024-04-29 22:36:30 +0200 | <tomsmeding> | and as you may have gathered I don't have a good suggestion for you either |
2024-04-29 22:36:21 +0200 | <shapr> | too bad, gotta pick another few candidates |
2024-04-29 22:36:07 +0200 | <tomsmeding> | I don't think so |
2024-04-29 22:36:00 +0200 | tomsmeding | runs `cabal run nofib-interpreter -f nofib` |
2024-04-29 22:35:48 +0200 | <shapr> | My real goal is to find a big pile of PBTs, I'm thinking accelerate isn't what I want |
2024-04-29 22:35:19 +0200 | <tomsmeding> | doesn't matter anyway if the problem is not there |
2024-04-29 22:35:10 +0200 | <shapr> | I could be wrong |
2024-04-29 22:35:08 +0200 | <tomsmeding> | I haven't tried recently, I confess, but when I was doing my master thesis I never got that to work |
2024-04-29 22:35:05 +0200 | <shapr> | I get tix values out of doctest, so I think it runs? |
2024-04-29 22:34:55 +0200 | <tomsmeding> | well that's good I guess |
2024-04-29 22:34:50 +0200 | <tomsmeding> | doctest works for you ?! |
2024-04-29 22:34:44 +0200 | <shapr> | yes, doctest runs quickly |
2024-04-29 22:34:35 +0200 | <tomsmeding> | doesn't `cabal test` also run doctest |
2024-04-29 22:34:22 +0200 | <shapr> | https://gitlab.haskell.org/ghc/ghc/-/issues/15751 |
2024-04-29 22:34:13 +0200 | <shapr> | there's a bug report that says it does this, but it's not the desired result |
2024-04-29 22:33:57 +0200 | <shapr> | I think so |
2024-04-29 22:33:53 +0200 | <tomsmeding> | on master? |
2024-04-29 22:33:47 +0200 | <shapr> | if that's what I get from accelerate "cabal test -f nofib" |
2024-04-29 22:33:44 +0200 | <tomsmeding> | I'm _quite_ sure it's not supposed to do that |
2024-04-29 22:33:28 +0200 | <tomsmeding> | nofib-accelerate in the interpreer? |
2024-04-29 22:33:24 +0200 | <tomsmeding> | what test suite are you running |
2024-04-29 22:33:18 +0200 | <shapr> | 2.5 hours |
2024-04-29 22:33:12 +0200 | <shapr> | now it's up to using 60 GB of RAM |
2024-04-29 22:32:18 +0200 | tri | (~tri@ool-18bbef1a.static.optonline.net) |
2024-04-29 22:29:39 +0200 | <tomsmeding> | (which I why I responded :p) |
2024-04-29 22:29:33 +0200 | <tomsmeding> | I don't myself, but I have a highlight on 'accelerate' here on irc |
2024-04-29 22:29:19 +0200 | <shapr> | neato |
2024-04-29 22:29:11 +0200 | <tomsmeding> | I share my office at the university with two PhDs who work on accelerate almost full-time :p |
2024-04-29 22:28:35 +0200 | <shapr> | tomsmeding: oh, I didn't know you did your master's thesis on accelerate |
2024-04-29 22:27:01 +0200 | waleee | (~waleee@h-176-10-144-38.NA.cust.bahnhof.se) |
2024-04-29 22:26:51 +0200 | <shapr> | I can't do much in <15 minutes |
2024-04-29 22:26:38 +0200 | <tomsmeding> | which is not anything deep; egraphs are just more complex, I also cannot write a red-black tree in <15 minutes |
2024-04-29 22:26:02 +0200 | <tomsmeding> | I cannot at all say the same for egraphs |
2024-04-29 22:25:54 +0200 | <tomsmeding> | I can write a passable UF implementation in <15 minutes |
2024-04-29 22:25:26 +0200 | <dolio> | Trickier than union-find? |
2024-04-29 22:25:08 +0200 | <tomsmeding> | yes |
2024-04-29 22:24:58 +0200 | flounders | (~flounders@24.246.176.178) |
2024-04-29 22:24:55 +0200 | <monochrom> | We also need a monoid-good group. :) |
2024-04-29 22:24:03 +0200 | <tomsmeding> | aren't we in there? |
2024-04-29 22:23:59 +0200 | <tomsmeding> | :D |
2024-04-29 22:23:54 +0200 | <tomsmeding> | egraphs are trickier, but also more powerful |
2024-04-29 22:23:54 +0200 | <monochrom> | Yeah I'm being sarcastic. Oh we missed the opportunity to form a function-programming-good group too! Oh wait... >:) |