Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 251 252 253 254 255 256 257 258 259 260 261 .. 5022
— 502,152 events total
2020-09-27 16:30:26 → kritzefitz joins (~kritzefit@212.86.56.80)
2020-09-27 16:30:43 × Tesseraction quits (~Tesseract@unaffiliated/tesseraction) (Ping timeout: 260 seconds)
2020-09-27 16:30:45 × Rudd0 quits (~Rudd0@185.189.115.103) (Ping timeout: 240 seconds)
2020-09-27 16:31:11 <maerwald> sm[m]: https://github.com/hasufell/stack2cabal/runs/1173082195?check_suite_focus=true#step:8:321 have you seen this on windows?
2020-09-27 16:31:26 <Guest_7> im a mac user
2020-09-27 16:31:26 <sm[m]> just a drive-by comment, I don't know what y'all are talking about
2020-09-27 16:32:12 <Enrico63> sm[m], but you can know what I am talking about :D
2020-09-27 16:33:01 <maerwald> sm[m]: because I remember you claimed windows support is excellent and I'm seeing xxx build failures whatever I try to build :)
2020-09-27 16:33:36 <maerwald> "could not execute ld" sounds good
2020-09-27 16:34:05 <p0a> Excellence is relative
2020-09-27 16:34:23 <Squarism> Anyone know if people started doing production stuff in Asterius?
2020-09-27 16:34:38 <sm[m]> maerwald: it doesn't ring a bell (and I don't see the "details above" it refers to)
2020-09-27 16:35:00 × Guest_7 quits (026317aa@host-2-99-23-170.as13285.net) (Remote host closed the connection)
2020-09-27 16:35:20 <sm[m]> maerwald: are you sure you're quoting me in context :)
2020-09-27 16:35:39 <maerwald> ofc not, I want to nudge you into helping me :p
2020-09-27 16:36:10 <sm[m]> better to just ask me then, I'll probably try
2020-09-27 16:36:37 → cole-h joins (~cole-h@c-73-48-197-220.hsd1.ca.comcast.net)
2020-09-27 16:36:54 <sm[m]> any luck googling for the error ?
2020-09-27 16:37:48 <sm[m]> (& are you using a gitlab worker with github CI ? seemed so)
2020-09-27 16:38:10 <maerwald> yeah
2020-09-27 16:38:39 <sm[m]> that sounds cunning
2020-09-27 16:40:16 → mirrorbird joins (~psutcliff@h85-8-41-6.cust.a3fiber.se)
2020-09-27 16:42:12 × Rudd0^ quits (~Rudd0@185.189.115.98) (Ping timeout: 256 seconds)
2020-09-27 16:42:25 → Rudd0 joins (~Rudd0@185.189.115.103)
2020-09-27 16:44:25 <maerwald> what? no
2020-09-27 16:44:49 <sm[m]> maerwald: btw I did have a look earlier since you cc'd me.. I wasn't logged in and couldn't see everything, but now I am.. and still having trouble finding the workflow that generated this - got a link ?
2020-09-27 16:45:30 <maerwald> sm[m]: https://github.com/hasufell/stack2cabal/blob/9dd4fc1a538dd2aa0a1d91c19f590b959f9a0b40/.github/workflows/haskell.yml
2020-09-27 16:45:35 → Dolly joins (585fd1fd@ti0203q160-5312.bb.online.no)
2020-09-27 16:45:45 <sm[m]> thanks
2020-09-27 16:45:46 <maerwald> I'm messing around with flags and constraints randomly
2020-09-27 16:46:00 <sm[m]> that'll surely work! :)
2020-09-27 16:47:49 <maerwald> overall, haskell + windows is an awful experience
2020-09-27 16:48:12 <maerwald> https://github.com/haskell-hvr/regex-posix/issues/4
2020-09-27 16:48:31 <sm[m]> oh I see.. when it says see the build log above, it's actually appearing a few lines below
2020-09-27 16:48:38 <[exa]> so if I run mmap from the mmap package on a pipe, it silently mmaps it as an empty file and doesn't report any kind of error. Is there any best way to detect that this has happened?
2020-09-27 16:49:40 <[exa]> ad pipe I mean something that comes out of `mkfifo`
2020-09-27 16:50:02 → sdx23 joins (~sdx23@unaffiliated/sdx23)
2020-09-27 16:50:22 <sm[m]> maerwald: one thing that comes to mind that's easy to try, specify shell: bash (or whatever the syntax is) so that bash is used on windows too. Otherwise it's in powershell or something
2020-09-27 16:50:40 <sm[m]> it can remove some variability
2020-09-27 16:50:40 → bitmapper joins (~bitmapper@159.2.173.158)
2020-09-27 16:51:16 <sm[m]> I guess it runs in WSL if you do that
2020-09-27 16:51:34 <maerwald> imo... there's not a single CI solution that has made it easy to *develop* a CI config
2020-09-27 16:51:44 <maerwald> it's all commit-push-wait non-interactive
2020-09-27 16:52:14 → Rudd0^ joins (~Rudd0@185.189.115.98)
2020-09-27 16:52:59 × Dolly quits (585fd1fd@ti0203q160-5312.bb.online.no) (Remote host closed the connection)
2020-09-27 16:53:06 <sm[m]> there is a way of getting an interactive shell. I wish I had set it up from the start, it would have been worthwhile
2020-09-27 16:53:17 × tabemann quits (~tabemann@172-13-49-137.lightspeed.milwwi.sbcglobal.net) (Remote host closed the connection)
2020-09-27 16:53:20 <sm[m]> it's in the actions marketplace
2020-09-27 16:54:14 <maerwald> so... it seems zlib fails to build straight on windows
2020-09-27 16:54:40 × Rudd0 quits (~Rudd0@185.189.115.103) (Ping timeout: 246 seconds)
2020-09-27 16:55:09 <p0a> [exa]: you probably can't do that from mmap because mmap is not careful enough
2020-09-27 16:55:16 → tabemann joins (~tabemann@172-13-49-137.lightspeed.milwwi.sbcglobal.net)
2020-09-27 16:56:22 <p0a> [exa]: I'm not entirely sure how the FFI works but the source code seems to use system_io_mmap_file_size, which would report 0 for the pipe I suppose
2020-09-27 16:56:26 <sm[m]> maerwald: you could also compare with mine and see what's different ? https://github.com/simonmichael/hledger/actions?query=workflow%3A%22windows+CI%22
2020-09-27 16:57:01 <p0a> [exa]: in C if you try to mmap a pipe file descriptor you would get EACCES
2020-09-27 16:57:15 <maerwald> sm[m]: you're not using cabal
2020-09-27 16:57:26 <maerwald> so that doesn't help me
2020-09-27 16:57:51 × brewmarche quits (~brewmarch@aftr-62-216-202-61.dynamic.mnet-online.de) (Quit: Leaving)
2020-09-27 16:58:00 <sm[m]> ah. Well it maybe something stack took care of for the windows users. As I've mentioned before, this is one of the things stack is good at
2020-09-27 16:58:22 <sm[m]> it's probably discussed in their issue tracker
2020-09-27 16:58:25 <[exa]> p0a: yeah that's likely... more like, what's the easiest test that the file cannot be mmapped AND doesn't involve rewriting the mmap library to actually force mmap to fail; which is in fact weird that it does not now.
2020-09-27 16:59:51 <maerwald> sm[m]: stack fails in other ways
2020-09-27 17:00:02 <maerwald> but I'm tackling one after another
2020-09-27 17:00:18 <sm[m]> did you google the message yet ? what about https://gitlab.haskell.org/ghc/ghc/-/issues/18550
2020-09-27 17:00:41 <sm[m]> sorry to be annoying
2020-09-27 17:01:55 × revprez_stg quits (~revprez_s@pool-108-49-213-40.bstnma.fios.verizon.net) (Quit: leaving)
2020-09-27 17:02:50 × tabemann quits (~tabemann@172-13-49-137.lightspeed.milwwi.sbcglobal.net) (Ping timeout: 260 seconds)
2020-09-27 17:03:43 <p0a> [exa]: what is mmapable now may not be later :) Why are you using mmap?
2020-09-27 17:03:46 → merijn joins (~merijn@83-160-49-249.ip.xs4all.nl)
2020-09-27 17:04:39 <[exa]> to avoid loading tons of bytestrings to RAM, tbh
2020-09-27 17:04:56 × ryansmccoy quits (~ryansmcco@193.37.254.27) (Ping timeout: 272 seconds)
2020-09-27 17:04:56 × jneira[m] quits (~jneira@125.red-193-152-119.dynamicip.rima-tde.net) (Read error: Connection reset by peer)
2020-09-27 17:05:05 → jneira[m] joins (~jneira@80.30.101.206)
2020-09-27 17:05:33 → ryansmccoy joins (~ryansmcco@156.96.151.132)
2020-09-27 17:06:06 <p0a> how come you're sometimes doing it on pipes?
2020-09-27 17:06:28 <[exa]> it's a shell tool; sometimes you just want to pipe stuff in dynamically
2020-09-27 17:06:53 <[exa]> it's diff, consider `diff <(zcat foo) bar`
2020-09-27 17:06:58 <p0a> the shell tool should know when it's a pipe and when it is not
2020-09-27 17:06:59 × bitmapper quits (~bitmapper@159.2.173.158) (Remote host closed the connection)
2020-09-27 17:07:09 <p0a> well, in that case you show above at least
2020-09-27 17:07:12 → bitmapper joins (~bitmapper@159.2.173.158)
2020-09-27 17:07:12 → revprez_stg joins (~revprez_s@pool-108-49-213-40.bstnma.fios.verizon.net)
2020-09-27 17:07:15 <[exa]> that expands to `diff /tmp/somepipe bar`
2020-09-27 17:07:32 <p0a> Oh oh, not in /that/ case :)
2020-09-27 17:07:41 <p0a> I thought you did `foo | cat bar'
2020-09-27 17:08:24 → d1w joins (585fd1fd@ti0203q160-5312.bb.online.no)
2020-09-27 17:08:31 × d1w quits (585fd1fd@ti0203q160-5312.bb.online.no) (Remote host closed the connection)
2020-09-27 17:08:47 × merijn quits (~merijn@83-160-49-249.ip.xs4all.nl) (Ping timeout: 240 seconds)
2020-09-27 17:09:27 × mirrorbird quits (~psutcliff@h85-8-41-6.cust.a3fiber.se) (Ping timeout: 240 seconds)
2020-09-27 17:10:06 <p0a> instead of the <(zcat foo) trick, just have your shell tool uncompress the file, `tool foo bar'
2020-09-27 17:10:11 × nullheroes quits (~danielvu@168.235.66.22) (Quit: WeeChat 2.9)
2020-09-27 17:12:34 → hyperisco joins (~hyperisco@d192-186-117-226.static.comm.cgocable.net)
2020-09-27 17:13:11 <p0a> I don't know what the solution is, but if mmap doesn't error on a pipe, I'm not sure what to do
2020-09-27 17:14:02 <[exa]> p0a: I'm kinda aiming to simulate the normal diff that survives this without problems, also user experience. :]
2020-09-27 17:14:13 <p0a> [exa]: https://hackage.haskell.org/package/unix-2.7.2.2/docs/System-Posix-Files.html
2020-09-27 17:14:27 <p0a> [exa]: You can use this to determine what kind of file you're dealing with I believe
2020-09-27 17:15:16 <p0a> with isRegularFile, for instance
2020-09-27 17:15:36 <geekosaur> witha race condition
2020-09-27 17:15:52 → Ranhir joins (~Ranhir@157.97.53.139)
2020-09-27 17:16:12 <p0a> I don't know how the regular `diff' works, but it probably doesn't read the whole file into memory. If you avoid doing that, you should be fine? I don't know if mmap is the solution.
←Prev  Next→
Page 1 .. 251 252 253 254 255 256 257 258 259 260 261 .. 5022

All times are in UTC.