2024/11/10

Newest at the top

2024-11-10 21:48:08 +0100 <tomsmeding> I see `( B.hs, interpreted )` and I read "interpreted" -- is that not accurate?
2024-11-10 21:47:57 +0100 <geekosaur> unless you tell it otherwise with *
2024-11-10 21:47:50 +0100 <geekosaur> did it really load it interpreted? if ghci finds object code it uses it
2024-11-10 21:47:41 +0100 <haskellbridge> <sm> TARGET arg and --repl-* opts, eg
2024-11-10 21:47:27 +0100 <geekosaur> afaik you can't
2024-11-10 21:47:24 +0100 <tomsmeding> I just did that there -- `:m B`
2024-11-10 21:47:18 +0100 <tomsmeding> so why can I load a module interpreted but non-*, then? :p
2024-11-10 21:47:10 +0100 <geekosaur> because object code doesn't contain symbols for non-exported names
2024-11-10 21:46:55 +0100 <geekosaur> which is the only way you can see non-exported names
2024-11-10 21:46:52 +0100 <tomsmeding> ah
2024-11-10 21:46:45 +0100merijn(~merijn@128-137-045-062.dynamic.caiway.nl) merijn
2024-11-10 21:46:41 +0100 <geekosaur> * means "compiled to bytecode so it can be debugged"
2024-11-10 21:46:40 +0100 <tomsmeding> with * I can see the internal definitions, without * I cannot
2024-11-10 21:46:32 +0100 <tomsmeding> yes it is
2024-11-10 21:46:27 +0100 <geekosaur> * is not related to scope
2024-11-10 21:46:26 +0100 <haskellbridge> <sm> tomsmeding, you do have some control with cabal repl args/opts don't you
2024-11-10 21:46:18 +0100 <tomsmeding> just keep my scope as-is, thank you!
2024-11-10 21:46:04 +0100 <tomsmeding> surely recompilation can be separate from _scope_?
2024-11-10 21:45:47 +0100 <geekosaur> I think the only recourse you have there is `-fobject-code`. otherwise, yes, if it recompiles it it'll do so as source, not object
2024-11-10 21:45:33 +0100 <tomsmeding> perhaps then cabal tends to pass a "topmost" module :)
2024-11-10 21:45:15 +0100 <haskellbridge> <sm> no, yes, it picks the first one on the command line
2024-11-10 21:45:10 +0100 <tomsmeding> and also, I have no control over what `cabal repl` passes to ghci
2024-11-10 21:44:56 +0100down200(~down200@shell.lug.mtu.edu) (Client Quit)
2024-11-10 21:44:48 +0100 <tomsmeding> geekosaur: what if I want my scope list to not be `*B` but `B`? I cannot achieve `:r` keeping `B` even if I list B first, because then it will become `*B`.
2024-11-10 21:44:29 +0100 <geekosaur> (if B needs A, it'll load it but not bring it in scope)
2024-11-10 21:44:20 +0100 <tomsmeding> how do I load a bunch of modules in ghci, but then preserve my current scope list on `:r`?
2024-11-10 21:44:14 +0100 <geekosaur> don't list it, or list it second
2024-11-10 21:44:06 +0100 <tomsmeding> geekosaur: okay, so how do I not do that?
2024-11-10 21:43:53 +0100 <geekosaur> I do not understand what point you are trying to make. Line 13 clearly shows you specifying it, as the first module which makes it "blessed" as you put it
2024-11-10 21:43:32 +0100 <haskellbridge> <sm> maybe it picks the alphabetically first ?
2024-11-10 21:43:31 +0100 <tomsmeding> ghci has decided here that A is the blessed "main" module or something
2024-11-10 21:43:17 +0100 <tomsmeding> if I `:m` so that I have only Prelude, then `:r` still makes the module scope `*A`
2024-11-10 21:42:57 +0100 <tomsmeding> no, only A
2024-11-10 21:42:52 +0100 <geekosaur> it brings into scope anything named on the command line
2024-11-10 21:42:30 +0100 <tomsmeding> it should _recompile_ it, not _bring it in scope_
2024-11-10 21:42:16 +0100 <tomsmeding> can ghci please not load A in my face here ?!
2024-11-10 21:42:05 +0100xdminsy(~xdminsy@117.147.71.147) (Read error: Connection reset by peer)
2024-11-10 21:41:54 +0100 <geekosaur> (consider that you don't want ghci to unload a module being used with `-interactive-print`
2024-11-10 21:41:51 +0100 <tomsmeding> geekosaur: sm: https://paste.tomsmeding.com/yOsBMz9q
2024-11-10 21:41:49 +0100down200(~down200@shell.lug.mtu.edu) down200
2024-11-10 21:41:00 +0100 <tomsmeding> it chooses the "main" (?) module, and always loads that in
2024-11-10 21:40:52 +0100down200(~down200@shell.lug.mtu.edu) (Quit: ZNC - https://znc.in)
2024-11-10 21:40:47 +0100 <haskellbridge> <sm> what's the module ?
2024-11-10 21:40:40 +0100 <tomsmeding> oh, but it also happens with ghci plain without _anything_ else.
2024-11-10 21:40:32 +0100Jeanne-Kamikaze(~Jeanne-Ka@142.147.89.228) Jeanne-Kamikaze
2024-11-10 21:39:55 +0100 <tomsmeding> perhaps I should have mentioned: this is with `cabal repl`.
2024-11-10 21:39:31 +0100 <tomsmeding> it happens with plain ghci without any config and has happened since ghc 8.* ages; probably before that but I wasn't around in the haskell world then :p
2024-11-10 21:39:25 +0100 <geekosaur> that's what I was wondering
2024-11-10 21:39:05 +0100 <haskellbridge> <sm> is that caused by a .ghci file somewhere ?
2024-11-10 21:38:17 +0100peterbecich(~Thunderbi@syn-047-229-123-186.res.spectrum.com) peterbecich