Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2014-01-30

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

All times shown according to UTC.

Time Nick Message
00:02 stepkut joined #darcs
00:25 mizu_no_oto joined #darcs
02:33 intripoon joined #darcs
03:19 lingxiao joined #darcs
03:47 n-dolio joined #darcs
03:56 mizu_no_oto joined #darcs
05:09 briarrose joined #darcs
05:32 Heffalump sm: ok, I'd better test my patches on Linux..
06:15 Heffalump sm: ?backendPermanent is an implicit parameter. It saves having to add an extra explicit parameter to everything.
06:15 Heffalump means that most of the changes are just in the type signatures
07:08 lelit joined #darcs
07:52 raichoo joined #darcs
08:13 Heffalump sm: I think your analysis above re the repo problems is correct. Definitely ok to cleanup the original repo.
08:13 Heffalump sm: I've pushed an extra patch to hopefully fix the two build problems you reported above - sorry for not testing more carefully myself.
11:22 raichoo joined #darcs
11:39 nomeata joined #darcs
12:42 raichoo joined #darcs
12:49 mizu_no_oto joined #darcs
13:16 amgarchIn9 joined #darcs
13:30 colDrMcBeardman joined #darcs
13:35 ente joined #darcs
13:35 ente joined #darcs
13:46 amgarchIn9 joined #darcs
14:02 mizu_no_oto joined #darcs
14:25 mornfall joined #darcs
14:25 nomeata joined #darcs
14:56 mizu_no_oto joined #darcs
16:18 dleverton joined #darcs
16:58 whaletechno joined #darcs
17:00 f-a joined #darcs
17:02 amgarchIn9 joined #darcs
17:19 byorgey_ joined #darcs
17:21 raichoo joined #darcs
17:30 rdesfo joined #darcs
17:31 sm hi Heffalump
17:32 sm woah, now your darcsden-abstract-backend repo has ballooned to 5.7M.. did you recover it ?
17:33 sm or maybe it's because I ran darcs check on it yesterday
17:33 sm getting rid of it
17:34 * Heffalump appears
17:34 Heffalump sm: no, it's definitely dead
17:34 sm ok it's gone, you should be able to rename to get rid of the _0 now
17:36 Heffalump done
17:37 Heffalump the extenstion-heavy approach is definitely one of the downsides of this patch, btw
17:38 sm yeah I wondered if that makes this harder to build long-term/more GHC version sensitive
17:39 sm though that's not a problem with the current darcsden deployment rate :)
17:39 Heffalump I think that's a relatively small risk as GHC tends not to break things and if it does there's usually a good way out
17:39 Heffalump cod comprehension is perhaps the bigger risk. I've tried to minimise the impact on client code and to document the classes and the likely errors it will see.
17:39 Heffalump s/cod/code/ # typing with a lot of lag
17:39 sm I think these patches are cool, my only concern is do we want to support multiple back ends ? It does make type sigs slightly more hassle
17:40 sm if you see a need, I'll go with it
17:41 Heffalump in a sense I think there's a code design advantage too: even with one backend, it clearly compartmentalises the backend code in one place
17:41 sm more than before ? that does sound good
17:41 Heffalump (biab, tunnels)
17:42 * sm can almost visualise Heffalump's train journey by now
17:42 f-a_ joined #darcs
17:42 sm green fields.. rolling hills.. TUNNEL.. river.. mountains.. TUNNEL ..
17:43 f-a_ the story of my life
17:47 f-a joined #darcs
18:02 colDrMcBeardman at least Heffalump has trains wherever he is.
18:02 colDrMcBeardman in the US, they destroyed our rail network long ago :(
18:03 sm there's still some trains
18:03 colDrMcBeardman sm, some, yes.
18:05 Heffalump I don't actually announce the tunnels on my train journey - the ones I do announce are the part of my journey that's on London Underground and hence is completely cut off for 20 mins or so.
18:05 Heffalump and I live in one of the flattest parts of England so no mountains sadly :-)
18:08 colDrMcBeardman sm, the thing is, though the trains don't go many places anymore and suffer from a lack of "bandwidth". you can literally beat them driving, which is sad since convnetional trains can do 150mph
18:09 sm colDrMcBeardman: they're definitely not for getting there fast. More of a chance to slow down :)
18:09 sm I really enjoyed my last train trips
18:10 sm compared to air travel, especially
18:15 notdan Was it more outrageous is not the speed of the trains, but how expensive they are :(
18:16 notdan I spent $100 and a day going from Oregon to SF
18:16 sm yup almost like air travel there
18:17 colDrMcBeardman yeah, it's sad. the network used to be much bigger and had more throughput.
18:18 colDrMcBeardman eventually, if we get the chance in the US, we'll end up realizing we need to rebuild it.
18:18 sm that's been happening in LA for a while
18:19 colDrMcBeardman well, the economy being broken isn't helping nation wide.
18:19 colDrMcBeardman but if it had never been gutted in the first place, we'd be much better off.
18:29 sm Heffalump: deployed it, working well so far
18:30 Heffalump cool
18:30 Heffalump I'd better do something useful with it soon :-) Got some darcs stuff to progress too though.
18:30 notdan Is there a function that given a list of patches and a patch would drop all the patches that depend on the given patch?
18:31 sm yes, what next ? sqlite back end ?
18:31 sm nice work
18:31 notdan Yay! now we need to abstract over redis too :P
18:31 sm he did that too
18:31 notdan Oh, that's awesome
18:31 sm BackendPermanent and BackendTransient
18:31 notdan that is seriously cool
18:32 notdan maybe I'll be able to run a small darcsden instance on my shell someday then heh
18:32 Heffalump various ideas: an in-memory backend; switch the tests to use a different CouchDB prefix so they don't clobber the real one.
18:32 sm me, I'm fine with moving off couchdb asap
18:32 Heffalump in-memory replacement for Redis would be much easier than Couch with the current code - the BackendPermanent class is pretty CouchDB specific
18:33 Heffalump so the next step on the BackendPermannt side would be to make the abstraction better.
18:33 notdan CouchDB is actually quite nice, but unfortunately not every host can provide it
18:33 sm notdan: yes we want to be able to just cabal install darcsden; cd repo; darcsden with no further setup
18:33 sm for local use, I mean
18:35 f-a left #darcs
18:37 sm even better would be: darcs ui
18:38 Heffalump indeed
18:50 lelit joined #darcs
18:53 mizu_no_oto joined #darcs
18:59 vikraman joined #darcs
19:04 vikraman joined #darcs
19:35 mornfall joined #darcs
19:38 vikraman joined #darcs
19:47 amgarchIn9 joined #darcs
20:11 byorgey_ joined #darcs
21:58 xymox joined #darcs
22:02 mizu_no_oto joined #darcs
22:32 rdesfo left #darcs
22:34 raichoo joined #darcs
22:49 mizu_no__ joined #darcs
22:58 schlaftier joined #darcs

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