Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2015-09-19

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

All times shown according to UTC.

Time Nick Message
01:47 ilbot3 joined #darcs
01:47 Topic for #darcs is now http://darcs.net/ | logs: http://irclog.perlgeek.de/darcs/ | darcs 2.10.0 is out http://darcs.net/Releases/2.10
03:06 kowey joined #darcs
03:29 peb` joined #darcs
04:22 mizu_no_oto joined #darcs
05:17 mal`` joined #darcs
05:23 kowey joined #darcs
06:56 Heffalump anyone know of any prior art on indicating an "unknown" result in a test script for something like "git bisect" ?
06:57 Heffalump the idea being that in some repository states the test passes, in some states it fails, and in other we don't know because of some other problem with that state like a build failure
06:58 * Heffalump asks and then immediately notices "git bisect skip"
07:01 kowey there's got to be a name for that phenomenon
07:10 Heffalump something to do with a brown paper bag
07:11 Heffalump hmm, except googling that suggests lots of results for something to do with racial discrimination in the US and not many for what I meant, which was "ask a brown paper bag about your problem first before asking anyone else"
08:13 Heffalump notdan: when you come to irill you'll need to call us to get in - gh has emailed his number
08:59 gh_ joined #darcs
09:01 gh_ hi!
09:02 thomie_ joined #darcs
09:03 kowey joined #darcs
09:05 kowey hello!
09:06 vinhdang joined #darcs
09:07 notdan hi @ all
09:07 notdan finally got to the sprint
09:07 vinhdang hello all
09:07 gh_ yeah, we're already 6!
09:12 kowey http://projects.haskell.org/camp/unique
09:13 notdan sm: I updated the patches at http://hub.darcs.net/simon/darcsden/compare/co-dan/darcsden-to-merge
09:13 gh_ discussing marketing, what the darcs website claims (a very different approach, but no explanation)
09:15 kowey http://darcs.net/Using/Model
09:15 kowey very very long
09:16 kowey but having some short and sweet would indeed be nice
09:16 * gh_ dreams of a 20s pretty video / animated gif that would do the job
09:16 vinhdang an infographic is also okay, I think
09:16 vinhdang but at least, there should be somewhere on the website to answer "why darcs"?
09:17 vinhdang in comparison with git and svn
09:18 vinhdang also, the page http://darcs.net/Using/Convert should be updated
09:18 gh_ yes, or maybe no explicit comparison but it should implicitely be undersadable as different from git/svn
09:18 gh_ aiming at an infographic seems a good, realistic objective
09:20 gh_ I think we can separate that task in 2 steps
09:20 vinhdang and QuickStart page is not really QuickStart, I think
09:20 gh_ 1) what we can do: think about what should go into this infographics, draw it on a whiteboard / write some document
09:21 kowey mmm
09:21 kowey documentation is very important indeed
09:21 gh_ 2) what a visual graphist can do: do the actual infographic
09:22 vinhdang there are many infographics maker available online
09:22 vinhdang https://www.canva.com/create/infographics/
09:22 gh_ (graphic designer seems the appropriate name for this job)
09:22 kowey I could have sworn somebody had done a text version of this in the past
09:23 vinhdang http://darcs.net/Features
09:23 vinhdang I think it is here
09:24 gh_ yes, the "some key features" part seems exploitable
09:25 gh_ cherry-picking / interactivity should be understandable as something freeer than what git proposes
09:26 gh_ I think hub.darcs.net should be added to the "some key features"
09:27 kowey http://lists.osuosl.org/pipermail/darcs-users/2009-April/019187.html
09:27 gh_ even if technically it's not darcs itself, it's something people expect
09:28 gh_ vikraman, ever used that infographics building website?
09:28 vinhdang and also, the feature to allows user to modify the code directly on the browser
09:28 gh_ uhh
09:29 gh_ vinhdang, the question was for you!
09:29 vinhdang it is plus in my point of view
09:29 Heffalump freeeeeeeeeeeeeeeeeeeeeeeeeeeer!
09:31 vinhdang the infographic makers seems simple
09:32 vinhdang just a simply online version of Photoshop
09:32 vinhdang especially if you follow the template
09:34 kowey https://gist.github.com/kowey/12be328ed06b869235ae
09:34 kowey <-- that's the 2009 patch that somebody sent for the doc
09:35 vinhdang yeah, it looks good
09:35 vinhdang more clear
09:36 kowey and this is what the front page used to look like: http://darcs.net/FrontPage?revision=2bc43fbbbddaf4126b43b5690383987da450a4b3
09:37 kowey sorry, just slowly picking through my memory of us having a three bullet point front page
10:01 Heffalump if we are making linux binaries, how/where should we build them?
10:23 kowey joined #darcs
10:27 * Heffalump is annoyed by the state of tracking down test failures and is going to do something (TM)
10:54 thomie_ joined #darcs
11:05 gh_ I like gitit's readme https://hackage.haskell.org/package/gitit#readme
11:06 notdan I am glad that readme's are rendered like this on hackage nowadays
11:12 notdan Just a random idea: FL/RL fusion for Darcs.Patch.Witnesses.Ordered, as in <https://wiki.haskell.org/GHC_optimisations#Fusion>
11:21 Heffalump notdan: makes sense - though I'm not sure how much the combinators in Witnesses.Ordered are used in practice
11:21 notdan well we do have some places where mapRL follows reverseRL
11:21 notdan or combinatations of map & filter
11:23 Heffalump fair enough - one challenge might be writing a 'build' operation with witnesses
11:23 Heffalump or there's stream fusion
11:24 Heffalump or just ad-hoc rules like map f . map g = map (f . g)
12:04 thomie_ joined #darcs
12:04 thomie_ left #darcs
12:23 pointfree joined #darcs
12:54 mizu_no_oto joined #darcs
12:55 kowey joined #darcs
13:27 vinhdang joined #darcs
13:56 notdan man, just spent so much time chasing a bug; I had two variables in scope: pset and pset', and of course I was using the wrong one in the wrong place
14:00 Riastradh joined #darcs
14:05 mizu_no_oto joined #darcs
14:47 gh_ :(
14:48 gh_ darcs code, y u no have the same functions for FL's and RL's
20:13 kowey joined #darcs
20:16 gh_ joined #darcs
20:55 gh_ and the sprint continues
21:03 gh_ so "cabal install --enable-tests" fails with ghc 8 because quickcheck does not compile
21:05 gh_ related: another possible refactoring for darcs 2.12 would be to switch to tasty
21:33 kowey joined #darcs
22:43 mizu_no_oto joined #darcs
22:59 Riastradh joined #darcs

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