Perl 6 - the future is here, just unevenly distributed

IRC log for #darcs, 2015-08-10

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

All times shown according to UTC.

Time Nick Message
00:00 fr33domlover S is specified explicitly, so this time the endpoint is available for free
00:22 pointfree NDN is not an overlay. A mapping between content and location does not need to be maintained (unlike, say, a DHT). Content-centric routing can can happen at the link layer (or over any layer for extra connectivity).
00:22 pointfree I think NDN would be cool for load-balancing/mirroring repositories/patches/spontaneous branches
00:22 pointfree It has been done with git: http://janakj.org/papers/grand.pdf
00:22 pointfree But yeah, manually adding endpoints would be fine to start.
00:32 mizu_no_oto joined #darcs
00:58 lf94 joined #darcs
07:29 Heffalump notdan: how's it going?
10:15 mizu_no_oto joined #darcs
11:52 z3r0 joined #darcs
13:45 fr33domlover pointfree, oh I see. A content-based internel would be cool :)
13:46 fr33domlover I'm slowly working on a semantic desktop in which identifiers are not URIs like in RDF, but are random garbage. I was thinking how to locate information in a distributed manner
13:46 fr33domlover perhaps NDN can fill this spot
14:07 karlpinc MasseR: Thanks.
16:20 notdan hi @ all
16:25 Riastradh joined #darcs
17:00 notdan Heffalump: I agree that darcs diff might be fundamentally wrong, but isn't it the most "practical" in a sense? most of the recorded patches do make sense as a unified diff
17:00 notdan of course patches-in-context approach is much better because it can be applied to primitive patches, lists of patches, lists of lists of patches etc
17:47 * Heffalump appears
17:48 Heffalump notdan: right, that's what I meant about it being most convenient at present.
18:11 notdan Hm looking at the code it seems that there is no easy way to get the context around a patch -- there is a way to get a context around a filehunk specifically, but that works only for changes on pristine/working tree
18:11 notdan because it relies on 'mReadFilePS' from ApplyState
18:28 Heffalump what things other than FileHunks would you want context from?
18:36 gh_ joined #darcs
18:42 pointfree Looks like http://darcsden.com/ is now a website about recycling and e-waste.
18:44 notdan Heffalump: ah, I meant to say that I was expected something that wraps around the Prim datatype -- like PatchInfoAnd, not a function that takes a FileHunk in
19:06 Heffalump ah, I see, it immediately gets turned into the Doc, we never really pass the context lines around in a useful way
23:33 mizu_no_oto joined #darcs

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