2025/03/02

Newest at the top

2025-03-02 21:31:33 +0100Jonno_FTW(~come@user/jonno-ftw/x-0835346) (Ping timeout: 244 seconds)
2025-03-02 21:26:16 +0100peterbecich(~Thunderbi@syn-047-229-123-186.res.spectrum.com) peterbecich
2025-03-02 21:19:35 +0100krei-se(~krei-se@p5dea1a2e.dip0.t-ipconnect.de) krei-se
2025-03-02 21:15:07 +0100bitdex(~bitdex@gateway/tor-sasl/bitdex) bitdex
2025-03-02 21:13:54 +0100 <geekosaur> (if you can use ANSI I've even detected them in simple shell scripts. ANSI standardization was a boon, once most older terminals died out)
2025-03-02 21:12:03 +0100alfiee(~alfiee@user/alfiee) (Ping timeout: 268 seconds)
2025-03-02 21:07:51 +0100 <geekosaur> sadly I think there's still stuff like emacs terminal that doesn't use ANSI
2025-03-02 21:07:08 +0100alfiee(~alfiee@user/alfiee) alfiee
2025-03-02 21:06:19 +0100 <tomsmeding> likely :)
2025-03-02 21:05:58 +0100 <monochrom> Perhaps s/no way to detect/no way I add extra logic to detect/ :)
2025-03-02 21:00:48 +0100caconym(~caconym@user/caconym) caconym
2025-03-02 21:00:01 +0100caconym(~caconym@user/caconym) (Quit: bye)
2025-03-02 20:59:52 +0100ljdarj(~Thunderbi@user/ljdarj) ljdarj
2025-03-02 20:57:49 +0100 <tomsmeding> https://github.com/haskell/haskeline/pull/197
2025-03-02 20:57:19 +0100Jonno_FTW(~come@user/jonno-ftw/x-0835346) Jonno_FTW
2025-03-02 20:56:59 +0100 <tomsmeding> yes, there is, ANSI sequences have a well-defined set of terminator characters
2025-03-02 20:56:46 +0100 <tomsmeding> > However, the prompt string must indicate the end of each escape sequence, since there's no way for Haskeline to detect it automatically.
2025-03-02 20:55:29 +0100Jonno_FTW(~come@user/jonno-ftw/x-0835346) (Ping timeout: 248 seconds)
2025-03-02 20:55:08 +0100cheater_cheater
2025-03-02 20:55:04 +0100cheater(~Username@user/cheater) (Ping timeout: 252 seconds)
2025-03-02 20:52:41 +0100cheater_(~Username@user/cheater) cheater
2025-03-02 20:50:06 +0100 <monochrom> I think yes.
2025-03-02 20:44:47 +0100 <int-e> (If README.md was included, would hackage pick it up?)
2025-03-02 20:44:24 +0100pavonia(~user@user/siracusa) siracusa
2025-03-02 20:43:25 +0100 <int-e> https://github.com/haskell/haskeline/blob/master/haskeline.cabal#L37 ...oh no README.md there
2025-03-02 20:42:55 +0100 <tomsmeding> but point there! If the entry point to your software is a git repository, I might look around there, but if it's a haskell library, I expect to be pointed to stuff from hackage :)
2025-03-02 20:42:27 +0100 <tomsmeding> I'm completely fine with github wikis; ace.js, the editor that the playground uses, has a whole bunch of info on a github wiki
2025-03-02 20:42:06 +0100 <int-e> Still... nobody uses github wikis. It's not a thing.
2025-03-02 20:41:53 +0100 <tomsmeding> the readme which is not shown on hackage :p
2025-03-02 20:41:20 +0100 <int-e> Well, I guess the README points there too.
2025-03-02 20:40:48 +0100Everything(~Everythin@195.138.86.118) Everything
2025-03-02 20:40:08 +0100 <int-e> So yeah I'd consider this hard to discover.
2025-03-02 20:39:39 +0100 <int-e> Which I found because https://github.com/haskell/haskeline/issues/72 mentions a wiki :P
2025-03-02 20:39:39 +0100 <tomsmeding> I'll make a small PR
2025-03-02 20:39:14 +0100 <tomsmeding> it would be nice if that's also in the haddocks. :)
2025-03-02 20:38:43 +0100 <int-e> there's https://github.com/haskell/haskeline/wiki/ControlSequencesInPrompt
2025-03-02 20:36:15 +0100michalz(~michalz@185.246.207.203)
2025-03-02 20:36:14 +0100 <tomsmeding> I would consider all this together insufficient documentation for haskeline to accept \STX :p
2025-03-02 20:35:33 +0100 <tomsmeding> perhaps this? https://git.savannah.gnu.org/cgit/bash.git/tree/y.tab.c#n8330
2025-03-02 20:34:42 +0100 <monochrom> Oh interesting.
2025-03-02 20:33:56 +0100 <tomsmeding> so perhaps bash accepts the same bytes, but the \[ \] parsing _is_ bash-specific?
2025-03-02 20:33:11 +0100notdabs(~Owner@2600:6c40:4300:950:2d43:42a2:e2d9:67c9)
2025-03-02 20:32:54 +0100ash3en(~Thunderbi@2a03:7846:b6eb:101:93ac:a90a:da67:f207) (Quit: ash3en)
2025-03-02 20:32:54 +0100 <tomsmeding> https://superuser.com/a/301355 "The bash-specific \[ and \] are in fact translated to \001 and \002 at y.tab.c:7640."
2025-03-02 20:32:49 +0100notdabs(~Owner@2600:6c40:4300:950:2d43:42a2:e2d9:67c9) (Remote host closed the connection)
2025-03-02 20:32:01 +0100 <tomsmeding> that's zsh, which has its own magic
2025-03-02 20:31:24 +0100 <geekosaur> cf. https://unix.stackexchange.com/questions/734343/ansi-escape-sequences-in-ps1-cause-incorrect-lengtā€¦
2025-03-02 20:31:07 +0100 <tomsmeding> iirc readline considers anything between \[ ... \] to be zero-width
2025-03-02 20:30:53 +0100 <geekosaur> I believe that's actually from readline
2025-03-02 20:30:52 +0100 <tomsmeding> my terminal passes 0x1d on ctrl-]