Camelia, the Perl 6 bug

IRC log for #darcs, 2013-10-28

| Channels | #darcs index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:02 HugoDaniel when is the next version of darcs expected ?
00:35 sm not long after someone puts some release notes together, I'd guess
00:42 whaletechno joined #darcs
00:46 favonia joined #darcs
00:54 sm and by the way, yay 2.9.8. tags are a sign of progress
02:09 intripoon_ joined #darcs
03:46 gh_ actually I've filled http://darcs.net/Releases/2.10 looking at the changes between 2.8.4 and 2.9.6 . However I'm not sure a release is for anytime soon.
03:46 gh_ we need to discuss priorities/roadmap first
04:45 stepkut joined #darcs
04:57 preflex_ joined #darcs
05:03 stepkut joined #darcs
05:10 mornfall gh_: http://irclog.perlgeek.de/​darcs/2013-04-19#i_6722634 then ;-) ... dunno about the tracker, I vaguely recall that tickets rot there for years on end :-P
05:15 gh_ mornfall, your MIME bug patch has not been reviewed yet, but at least I could screen it
05:16 gh_ I do check the traffic on darcs-devel more often than the IRC channel log
05:19 gh_ the bug tracker enabled us to use that sshfs-related patch ( http://bugs.darcs.net/issue904 ) five years after, so I'm glad we have it :-)
05:19 mornfall oh, I even emailed that patch
05:21 mornfall I must have been hopeless :-P
05:22 mornfall the repos involved in that merge are public, yes, although the merge is done now in the public copies
05:22 mornfall I'll look into making a snapshot and mothballing it in the tracker later this week
05:22 mornfall now I gotta run or I miss a train, which'd be quite unfortunate
05:24 gh_ cool :) (for the future bug ticket)
05:24 gh_ good bye
06:03 arpunk_ joined #darcs
06:09 favonia joined #darcs
06:47 donri joined #darcs
06:56 schlaftier joined #darcs
08:17 raichoo joined #darcs
09:05 kowey joined #darcs
09:13 kowey hmm, ok so tty detection comes from System.IO not from haskeline
09:13 kowey and it seems to work with my trivial tester main = hPutStrLn stderr . show =<< hIsTerminalDevice stdout
09:16 kowey oh hang on…
09:17 kowey ah-hah! I'm running reviewed, and am behind by a patch
09:17 kowey reviewed was just testing isInteractive (which means the flag), screened actually uses the terminal detection
09:18 kowey maybe I should just run screened
09:23 amgarchIn9 joined #darcs
10:55 amgarching joined #darcs
11:02 arpunk joined #darcs
11:31 owst joined #darcs
11:37 mizu_no_oto joined #darcs
12:50 mizu_no_oto joined #darcs
13:18 _ilbot joined #darcs
13:18 Topic for #darcs is now http://darcs.net/ | logs: http://irclog.perlgeek.de/darcs/ | darcs 2.8.4 is out http://darcs.net/Releases/2.8
14:06 mizu_no_oto joined #darcs
14:20 whaletechno joined #darcs
14:35 gh_ joined #darcs
14:41 gh_ kowey, terminal detection has been in reviewed for a while!
14:41 gh_ oh wait
14:41 gh_ no :-)
14:42 gh_ I haven't pushed the interactive changes patches to reviewed on purpose
14:45 stepkut` joined #darcs
14:46 gh_ to me, reviewed => power users, and screened => developers
15:52 mornfall dunno, I just stick to distro packages -- is something I really want missing from 2.8.4? :)
15:52 mornfall (I mean, something that's already in one of those repos...)
15:59 _ilbot joined #darcs
15:59 Topic for #darcs is now http://darcs.net/ | logs: http://irclog.perlgeek.de/darcs/ | darcs 2.8.4 is out http://darcs.net/Releases/2.8
15:59 gh_ mornfall, http://darcs.net/Releases/2.10
16:02 mornfall heh, how old is that diff optimisation patch?
16:05 mornfall also, I thought it was rejected
16:06 amgarching "Remove latex manual from darcs’ repository" ? It was an interesing reading back in 2005, well written. Sad.
16:07 n-dolio mornfall: They finally wrapped some of the output of 'darcs help get' so that it almost fits in an 80 column terminal. :)
16:07 n-dolio I wanted that.
16:07 arpunk joined #darcs
16:09 stepkut joined #darcs
16:11 jyyou joined #darcs
16:13 raichoo joined #darcs
16:21 dolio joined #darcs
16:31 kowey amgarching: it's still around, if I understand correctly
16:31 kowey just not in the same format
16:32 kowey I've tended to lose track of our documentation changes, but the general direction is towards fewer formats for things with different outputs (man pages, help text, manuals, wiki…)
16:33 kowey oh, not there
16:55 gh_ mornfall, I ported it from your adventure branch, I think last year
16:57 gh_ amgarching, we had a duplication problem between the manual, the wiki/website and the command help. I moved some content from the manual to the help strings of the commands. The rest should find its way to the wiki. Anyhow for 2.10 we should find a way to easily generate a manual from that, solving the duplication/maintainance problem.
16:59 * gh_ added that to the to-do list of the 2.10 page
17:00 mizu_no_oto joined #darcs
17:00 gh_ mornfall, the centralization of all path related functions into a single module also found its way into darcs
17:01 jyyou joined #darcs
17:16 gh_ thread-safe filesystem… does darcs lack it?
17:17 gh_ re: http://irclog.perlgeek.de/​darcs/2012-12-30#i_6285737
17:20 Heffalump gh_: in the sense that it uses the current working directory OS global, which isn't thread safe
17:20 gh_ ok
17:22 gh_ how shall we fix it?
17:36 amgarchIn9 joined #darcs
18:03 jyyou joined #darcs
18:23 whaletechno joined #darcs
18:33 Heffalump gh_: a tricky question
18:33 Heffalump one difficulty is that it's a lot slower not to use CWD, so we really need an abstraction to select whether to use it or not
18:37 Heffalump I spent quite a while prototyping a "darcs monad" about a year ago, but the end result seemed a bt too clunky to use
18:47 lelit joined #darcs
18:52 lelix joined #darcs
18:56 sm what's slow Heffalump ?
19:00 Heffalump sm: constructing paths explicitly each time rather than basing them on CWD
19:00 Heffalump at least that's my vague recollection of the situation
19:02 sm hmm
19:02 sm I wonder if that faster file paths lib released since would help
19:05 schlaftier joined #darcs
19:05 sm can't find it
19:05 Heffalump from what I remember, the issue is more to do with the OS interpreting the paths sent to it
19:05 Heffalump i.e. the work it has to do
19:07 sm did that really noticeably slow darcs ?
19:11 sm if so, I guess that might mean our on-disk format uses too many files ?
19:16 amgarchIn9 joined #darcs
19:30 alexei joined #darcs
19:34 Heffalump well, it has to read working, which is unavoidable
19:34 Heffalump but I'm not sure what the total impact was
19:35 pseudolio joined #darcs
19:40 arpunk joined #darcs
20:51 gh_ joined #darcs
21:30 alexei joined #darcs
21:57 sm joined #darcs
23:47 mizu_no_oto joined #darcs
23:55 alexei joined #darcs

| Channels | #darcs index | Today | | Search | Google Search | Plain-Text | summary