Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2014-02-09

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

All times shown according to UTC.

Time Nick Message
01:21 intripoon joined #darcs
01:33 dolio joined #darcs
02:20 intripoon_ joined #darcs
02:22 mizu_no_oto joined #darcs
02:50 dolio joined #darcs
04:39 alexsuraci__ joined #darcs
04:44 alexei___ joined #darcs
05:58 mizu_no_oto joined #darcs
06:39 lelit joined #darcs
09:37 alexei___ joined #darcs
09:56 kowey joined #darcs
09:56 kowey morning all
09:59 notdan Howdy
10:01 kowey notdan: thanks for boiling down issue2356
10:07 kowey I wanted to find the 'inexistant' wording in our history, but ran into this issue http://bugs.darcs.net/issue2357
10:09 alexei___ joined #darcs
10:18 kowey err? clang: error: no such file or directory: '170'
10:18 kowey clang: error: no such file or directory: 'patches"'
10:21 kowey seems like this being passed in to GHC is causing something some confusion somewhere
10:21 kowey somewhat
10:31 notdan are you trying to comple darcs on OSX?
10:31 kowey yeah
10:31 notdan you'll need this patch : http://hub.darcs.net/darcs/darcs-screened/patch/20140126163054-e7f45
10:31 kowey oh
10:32 kowey I'd done that by hand
10:32 kowey but good to know I'm not alone then and therefore don't have to do anything about this then
10:32 notdan I ran into the same issue. It's because Apple had decided to replace gcc with clang and all the CPP stuff is broken now
10:32 kowey it's causing lots of confusion
10:33 kowey I tried getting a maths-orientated hacker to play with diagrams
10:33 kowey but she ran into computer-says-no installation woes related to this CPP stuff
10:33 kowey just when things were starting to smooth out
10:34 notdan It's a massive pain in the ass. You need a wrapper script from http://www.haskell.org/platform/mac.html and possible some changes to custom Setup.hs scripts
10:34 kowey ah yes, non-interactive changes is so much nicer
10:34 kowey oh good, it's official now
10:34 kowey no more pointing people to GitHub
10:34 kowey gist
10:36 notdan Btw, I can't reproduce your bug on 2.9.8 (+ 20 patches)
10:36 notdan Which version of regex-posix did you compile darcs against?
10:37 notdan I am using the latest one 0.95.2
10:40 notdan There was a thread on cafe about regex packages and I remember reading something about a better package
10:40 notdan regex-dfa or something
10:41 kowey I'm on a stock Haskell Platform 2013.2 fwiw
10:41 notdan there were two threads about the even!
10:41 notdan kowey: hm
10:41 kowey so, $ ghc-pkg list | grep regex
10:41 kowey regex-base-0.93.2
10:41 kowey regex-compat-0.95.1
10:41 kowey regex-posix-0.95.2
10:42 notdan weird
10:42 notdan same as me
10:43 Heffalump do your terminals have different encodings?
10:44 kowey I've got LANG=en_GB.UTF-8
10:44 notdan hm I dont have LANG defined
10:44 notdan god, encodings are so hard, why can't we all just speak in ASCII :(
10:45 Heffalump we should just abolish all the non-ASCII languages :-)
10:45 notdan it still works if I set lang to en_GB.UTF-8
10:47 mornfall kowey: moving away from gcc is a good thing
10:47 kowey time-travel science fiction based on people going back to persuade computer people to design things for the unreasonably optimistic use case
10:47 notdan mornfall: why? is clang really that better?
10:48 mornfall notdan: basically, yes
10:48 kowey yes, your code will live past year 2000 and be used in places like China
10:48 mornfall kowey: or Europe
10:49 mornfall or Florida, for that matter :P
10:49 kowey I think I tell my Miami police officer story too often
10:49 notdan I haven't heard it ye :)
10:49 kowey first encounter with the American non-anglosphere
10:49 mornfall kowey: I don't remember hearing it
10:50 kowey so there I was fresh out of Uni, about to go to France
10:50 kowey and had to go to the nearest consulate to get my visa
10:51 kowey 8 hours later, I stumble out of the Greyhound bus, bleary eyed
10:51 kowey wandering around Miami while waiting for my appointment
10:52 kowey I find way into a park and kind of loiter around, where I catch a bunch of old men playing dominos
10:52 kowey so I kind of stand there being awkward just looking at them
10:52 kowey until this police officer shoos me away
10:53 kowey in doing so he says to me, I guess to explain he wouldn't be saying much,
10:53 kowey that he was sorry, but
10:53 kowey that he didn't speak “mucho English”
10:53 kowey (end of story… left a strong impression!)
10:53 mornfall :-)
11:00 notdan heh :]
11:04 notdan kowey: going back to that issue, I don't understand where does darcs get the unicode symbol from?
11:04 notdan "hunk inexistent" is not unicode
11:05 kowey one interesting thing about this, notdan
11:05 kowey is that I can search for other strings like "foo"
11:05 kowey or "inex"
11:05 kowey I wonder if the thing that triggers the message is just searching for any string that isn't found in any hunk
11:06 kowey I wonder if we have a nice name for that phenomenon where an error message doesn't really mean what it sounds like
11:06 kowey and is just the (perfectly reasonable but ultimately unrelated) knock-on effect of the real pb
11:07 notdan I wonder if you can just used regex-compat-tdfa as a drop-in replacement: https://github.com/simonmichael/hledger/pull/134
11:08 notdan Oh, here we go!
11:08 notdan % darcs changes -p "inexistant"
11:08 notdan darcs failed:  Text.Regex.Posix.String died: (ReturnCode 17,"illegal byte sequence")
11:09 notdan and I can't search for anything now :S
11:09 notdan oh and it's 100% because of the UTf8 locale
11:09 notdan because
11:09 notdan when I change LANG to an empty string
11:09 notdan it works
11:10 notdan I will attempt to make use of regex-tdfa now
11:10 notdan and see if it works
11:13 kowey I bet searching for foo still works
11:18 notdan nahwait
11:18 notdan I am confused
11:18 notdan it works now...
11:18 notdan wtf
11:19 notdan seriously, I can't reproduce it anymore
11:22 notdan http://paste.lisp.org/display/141187 ugh
11:23 notdan what's the difference between 'pattern' and -p?
11:23 notdan ooh, because -p is not 'pattern' it's 'patches'
11:25 notdan kowey: wait, do you have 'darcs match' as a custom alias?
11:25 notdan s/darcs match/darcs changes match/
11:25 notdan that would explain the difference in our setups
11:40 kowey notdan: I don't think so
11:40 kowey must be a think-o on my part
11:43 notdan darcs changes 'match stuff' will try to list changes to a file/directory named 'match stuff', it does not involve any regex stuff
11:43 notdan anyway, I am going to finish my lunch and try out the regex-compat-tdfa library
11:44 notdan http://hub.darcs.net/shelarcy/regex-compat-tdfa
11:45 kowey darcs changes --match 'hunk inexistant'
11:45 kowey I think is the search I was trying to do
11:45 notdan yeah
11:45 notdan and that produces an error
11:45 kowey I confess that git flags are really starting to infect my brain
11:45 kowey and displace some of the darcs
11:49 notdan fair enough :)
11:49 notdan I already forgot all the git flags and commands
11:49 notdan because they are terribly inconsistent
11:50 mornfall I am actually almost fluent in git these days :-P
11:50 mornfall it still irks me, but I can get work done (most of the time, anyway)
12:11 notdan kowey: I sent a fix for the unicode issue to the patch tracker
12:11 notdan It's also in my darcsden: http://hub.darcs.net/co-dan/darcs
12:15 kowey thanks!
12:15 kowey that was rather fast
12:15 notdan im high on all the mate infusion i have been drinking since morning ;]
12:59 notdan uh-oh, utf8.sh is failin g:(
13:13 notdan it also fails without my regex-compat-tdfa patch though...
13:20 kowey joined #darcs
13:29 raichoo joined #darcs
13:34 Heffalump notdan: what OS, MacOs?
13:44 notdan yep, OS X 10.8
13:44 kowey joined #darcs
13:58 drostie joined #darcs
14:09 Heffalump the utf8 test has quite a lot of stuff in it so is quite sensitive to being broken
14:09 Heffalump and a lot of people don't run it routinely because it just skips silently if the machine running it doesn't have quite the right setup
14:27 marclaporte joined #darcs
14:33 mizu_no_oto joined #darcs
15:16 gh_ joined #darcs
15:20 gh_ hi
15:27 marclaporte joined #darcs
15:42 ccs4ever joined #darcs
16:37 mizu_no_oto joined #darcs
16:40 gelisam joined #darcs
16:40 gelisam hi! I can't figure out how to create a dpatch file out of a record I just made.
16:41 gelisam "darcs send --output=importP.dpatch" complains that I "Can't send to current repository". Didn't it used to ask me which patches I want to include?
16:41 gelisam or maybe it was some other command...
16:43 raichoo joined #darcs
16:46 colDrMcBeardman joined #darcs
16:48 Heffalump gelisam: you probably don't have a remote repository configured
16:48 Heffalump try darcs pull --set-default <upstream-repo> (you can answer 'n' to all the patches if you don't want them)
16:49 Heffalump or darcs send --output... <upstream-repo>
16:49 Heffalump actually darcs send takes --set-default so that's the simplest way
16:49 Heffalump do it once with --set-default and an explicitly specified repo
16:49 gelisam but http://darcsden.com/ is down, so there is no upstream repo at the moment :(
16:50 Heffalump ah, in that case I think you need to make a local clone without the patch you want to "send"
16:50 Heffalump at least that's hwo I normally fake it, not sure if there's any better option
16:50 gelisam ah! okay, I think I should be able to figure out how to do that. thanks!
16:51 gelisam it worked, thanks again!
16:53 notdan hm yeah would be nice to have a simpler option
16:55 notdan I think it is possible (theoretically) to produce a bundle out of selected patches
16:55 notdan without repo comparison
17:25 rdesfo joined #darcs
17:36 marclaporte joined #darcs
17:43 gh_ This year's google summer of code starts quite early. Organization application deadline is Feb 14th, student application deadline is March 21st. Internship happens between May 19th and August 18th.
17:44 gh_ who is a student, here? :-)
17:45 kowey joined #darcs
17:53 gh_ Heffalump, do you think the "possible use of GHC extensions" of http://darcs.net/Releases/2.10  could serve as the basis of a GSoC project? or it's to little work
17:54 gh_ s/to/too
17:56 * mornfall is a perpetual student :-)
17:58 mornfall (and I'm not saying that the money wouldn't come in handy, but probably not worth the stress of working against the darcs inertia... also, I have a thesis to finish)
18:01 gh_ interesting!
18:03 gh_ well the idea of such a project is that the plan is agreed before
18:05 gh_ so the intertia would happen in the "agreement" phase.. if everything goes fine
18:05 gh_ *inertia
18:08 notdan I am a student, but March 21st is a really early deadline :(
18:09 gh_ I think you can change your mind (ie turning down the grant and project) up to April 15th
18:10 gh_ and then the slot gets reassigned
18:12 notdan Oh I am sure if I would be able to apply though D:
18:12 notdan will need to look up the detail;s
18:12 notdan im graduating this year
18:13 mornfall there was a faq entry about that somewhere
18:14 mornfall also, as long as you go to postgrad it doesn't matter :)
18:14 gh_ notdan, https://www.google-melange.com/gsoc/document/show/gsoc_program/google/gsoc2014/help_page
18:14 gh_ notdan, you need to be student on April 21st
18:14 gh_ and >=18 years old :-)
18:15 notdan oh so I am good
18:20 notdan I did the GSoC with diagrams last year, it was sucessful but it's sill in works
18:20 notdan that reminds me that I haven't touched that projects in some time ;(
18:23 Heffalump closed type families aren't available yet
18:23 Heffalump because they're not in 7.4
18:23 notdan They will be in 7.8 though
18:23 Heffalump yes, but we won't have that as our _minimum_ version for a while
18:24 gh_ how about 7.6?
18:24 Heffalump I started looking at the ConstraintKind stuff already (though that reminder of the earlier thoughts is really useful) but I don't think it's really a GSoC projct
18:24 Heffalump no, I think they're 7.8 only
18:24 gh_ ok, too soon then
18:24 Heffalump also, I don't actually know what we'd do with them - it's not clear from the IRC conversation that there's a concrete idea
18:45 Heffalump notdan: yes, definitely possible in theory to do that
18:45 Heffalump gh_: should we organise with haskell.org now to ask for the 2 extra slots?
18:58 gh_ Heffalump, yes
19:25 colDrMcBeardman joined #darcs
19:50 alexei___ joined #darcs
20:05 nomeata joined #darcs
20:57 rdesfo left #darcs
20:58 raichoo joined #darcs
21:21 whaletechno joined #darcs
22:30 Feuerbach joined #darcs
22:30 Feuerbach Hi, how do I edit the last patch's name?
22:32 gh_ Feuerbach, darcs amend --edit, and then select the first patch proposed to you
22:33 Feuerbach cheers gh_
22:39 kowey joined #darcs
22:58 Feuerbach left #darcs
23:32 mizu_no_oto joined #darcs
23:57 marclaporte joined #darcs

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