Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2015-03-15

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

All times shown according to UTC.

Time Nick Message
01:18 sm_ hi notdan
02:10 mizu_no_oto joined #darcs
02:47 ilbot3 joined #darcs
02:47 Topic for #darcs is now http://darcs.net/ | logs: http://irclog.perlgeek.de/darcs/ | darcs 2.8.5 is out http://darcs.net/Releases/2.8
03:49 mizu_no_oto joined #darcs
04:03 mizu_no_oto joined #darcs
04:23 kaol joined #darcs
04:37 ankitku joined #darcs
04:37 alexei_ joined #darcs
06:45 alexei_ joined #darcs
07:34 ankitku_ joined #darcs
08:55 alexei_ joined #darcs
14:10 mizu_no_oto joined #darcs
14:27 alexei_ joined #darcs
14:37 aristid_ joined #darcs
16:05 c74d joined #darcs
18:24 mizu_no_oto joined #darcs
19:08 mizu_no_oto joined #darcs
19:33 alexei_ joined #darcs
20:26 dixie_ joined #darcs
21:12 alexei_ joined #darcs
21:20 notdan sm: Heffalump: what do you think of this? https://gist.github.com/co-dan/8067d35c63432424e9e4
21:20 notdan Are you interested in mentoring this year?
21:21 c74d joined #darcs
21:23 sm hey notdan
21:23 sm I'll be able to help some, but not formally mentor
21:25 sm gist wishes: numbering, clarify which bits refer to darcsden generally and which to darcs hub
21:26 notdan Hm, I think they all apply to Darcsden generally
21:28 sm ok, I'll give different feedback if a proposal is intended to be deployed on darcs hub
21:28 notdan Do you run a modified version of darcsden on darcshub?
21:28 sm I think 1 is a great project, Heffalump already did the abstraction, now it needs to be used to add support for easy/no-install dbs
21:29 sm yes slightly modified.. it's darcsden built with -fhub
21:29 dino- joined #darcs
21:29 notdan Ah, yes, indeed.
21:30 notdan Well I think it would be possible to see what changes to include in -fhub and what not to include
21:30 notdan Obviously, -fhub should still use Redis and CouchDB, and not plain files
21:30 dino- I am trying to build the source from http://darcs.net/releases/branch-2.10 but am seeing an error when I `cabal build`: setup: can't find source for Darcs/Patch in src, hashed-storage,
21:31 dino- dist/build/autogen
21:31 notdan But I thought that the the ideas 2,3, and 4 could be implemented on the website
21:31 dino- I am on Linux, GHC 7.8.4, cabal 1.22.0.0, and trying to do this with a cabal sandbox
21:31 sm darcs hub isn't married to redis/couchdb, either
21:32 dino- Any ideas of what I'm doing wrong? Searching for this is coming up with bug reports from 5 years ago!
21:32 sm dino-: strange, I successfully built that branch the other day
21:33 sm maybe starting over with a fresh clone would help
21:34 dino- sm: I'll try that (kicking myself for not sandbox init into a dir other than .)
21:35 sm normally you'd sandbox init inside the darcs source directory
21:36 sm notdan: for darcs hub, 2a (at least one of the formats) would be great, 2b possibly but I haven't heard any demand for that
21:37 dino- sm: I try to re-use some sandbox dirs, especially if I'm doing "branches". To minimize all the building over and over.
21:37 dino- One thing I did before was ctrl-c during the darcs get of sources. Maybe I did that too soon or maybe it was bad that I didn't specify --lazy?
21:37 dino- Not sure how that works exactly
21:38 dino- I thought that was how you interrupt getting the entire history
21:40 sm notdan: 3 is probably only relevant to darcs hub (and any other large repo farms, if they ever exist). It could be nice, but adds operational costs so needs some cost/benefit research
21:42 sm 4 sounds potentially good too, Heffalump knows all about it
21:43 sm I suspect 1, 2+3, and 4 are each a complete project
21:43 sm (GSOC-sized)
21:44 dino- It's building now without that error. This time I didn't interrupt the darcs get at all.
21:44 sm dino-: you might have done it too soon, it should let you know when it's ok to do it
21:45 dino- Yep, I don't think I saw the message and was vaguely uneasy.
21:50 dino- sm: thank you
21:50 sm np
21:53 sm notdan: related to the darcsden/darcs hub distinction, you might want to add the success criteria for each goal, eg: coded, merged in darcs[den] trunk, shipped in a darcs[den] release, deployed on darcs hub. That would help us assess your time estimates, which look a bit unrealistic from here
21:53 c74d joined #darcs
21:53 sm even with your past experience and known contributor skills (happy to be proved wrong..)
21:55 sm thanks for the proposals. afk a little..
22:02 c74d joined #darcs
22:15 dino- I am now seeing `caba test` freeze. I ran it with --show-details=streaming and was able to see incremental output. It seems hung here: issue1923-cache-warning.sh (Darcs2) (PatienceDiff): [Skipped]
22:15 dino- after doing many many tests, most of them resulting in Success
22:15 mizu_no_oto joined #darcs
22:32 c74d joined #darcs
23:03 pointfree-w530 joined #darcs
23:30 pointfree-w530 Are messages like "Creating a patch index, please wait. To stop press Ctrl-C" sent to stderr? If not, could a darcs developer make these messages go to stderr? I noticed it is making ohloh_scm tests like the cat_file_test fail sometimes and currently it doesn't look like I can redirect it to /dev/null.

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