Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2014-11-24

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

All times shown according to UTC.

Time Nick Message
00:52 sm lelit: thanks for the report, I noted your issue at http://hub.darcs.net/lelit/
00:52 sm I mean, at http://hub.darcs.net/simon/darcsden/issue/121
00:54 mizu_no_oto joined #darcs
01:01 sm and in my just-rebooted GTD system, so with luck I'll fix it soon
04:08 stepkut joined #darcs
04:48 mizu_no_oto joined #darcs
05:33 Heffalump gh_: I'm fairly sure I know how to make PI use constant memory, just need to make sure I'm preserving behaviour
05:34 Heffalump gh_: I was also having some trouble with darcs-benchmark (http://irclog.perlgeek.de/darcs/2014-11-21#i_9701340)
07:26 stepkut` joined #darcs
07:57 lelit joined #darcs
08:39 gal_bolle joined #darcs
12:15 mizu_no_oto joined #darcs
13:24 c74d joined #darcs
13:25 alexsuraci_ joined #darcs
13:30 c74d joined #darcs
13:33 alexei___ joined #darcs
13:36 amgarchIn9 joined #darcs
14:56 amgarching joined #darcs
15:59 gh_ joined #darcs
16:13 lelit joined #darcs
16:50 Thaalos joined #darcs
16:50 Thaalos joined #darcs
17:50 aristid joined #darcs
17:54 Heffalump hi
17:54 sm hi Heffalump
18:05 lambdabot joined #darcs
18:22 gh__ joined #darcs
18:52 Heffalump gh__: re patch index, I'm hopeful about patch1234, I'll screen it once I manage to benchmark it properly
19:09 Heffalump gh__: what's the problem with darcs log, btw?
19:10 gh__ Heffalump, ok I'll try it maybe later today
19:11 gh__ Heffalump, the problem is with the code that summons a pager. it first generates a temporary file with the whole output of darcs (log) then feeds it to the pager.
19:12 gh__ Heffalump, this was made on purpose at some point in the past, see patch "Restore creation of temporary file when calling a pager"
19:12 gh__ from 2008
19:13 gh__ so it's utterly slow when running darcs log in a big repo on a relatively old machine
19:16 Heffalump hmm, hopefully there's a better solution. I don't see why you'd have to redirect the stdout of the pager.
19:20 Heffalump FWIW darcs log on the screened repo completes in 0.5s on my new laptop
20:38 gh__ joined #darcs
20:47 gh__ Heffalump, is changing the on disk order of [PatchId] a problem? re your "maintain the list of PatchIds inside the PatchIndex structure" patch
21:39 lambdabot joined #darcs
22:06 Heffalump gh__: unless I screwed up what I sent, there's one patch that reverses them in memory only, and then another that reorganises the data structures but doesn't change the ordering in memory either
22:06 Heffalump and the disk structure stays the same in both cases
22:07 Heffalump "maintain the list of PatchIds inside the PatchIndex structure" is the latter one
22:56 amgarching joined #darcs

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