|
weekly Rakudo status meetings with phase transitions: Tue 19:00 UTC | IR clogs at irclog.perlgeek.de/phasers/today Set by moderator on 25 May 2010. |
|||
|
12:44
masak joined,
mberends joined
12:45
colomon joined
|
|||
| colomon | pre-report, as I intend to be touring a lighthouse during today's report. | 12:46 | |
| mostly vacationed this week | |||
| merged in Prakash Kailasa's nested hyper patch, and went on to implement hash hypers. | 12:48 | ||
| in the process, we broke 33,000 passing tests. | |||
| just conferred with the #perl6 brain trust and decided to extend nested hypers to handle hashes as well. | 12:49 | ||
| .eor | |||
| jnthn | There's more than one way to read, "in the process, we broke 33,000 passing tests" ;-) | ||
| masak | with high probability, I won't make it to tonight's meeting, so here's a pre-report of sorts: | ||
| * I've gotten rolling with my GSoC work. here's the report from week 1: use.perl.org/~masak/journal/40369 -- weekly blog reports are probably what I'll end up doing. | |||
| * correct Buf.decode behaviour is currently blocking on Parrot implementing some bytes->string mechanism. | |||
| * but it struck me that in the meantime, one possible workaround is to implement bytes->chars decoding in the PIR code of the Buf.decode method. may or may not be worth the trouble. | |||
| * if someone wants inspiration for what to hack on before Rakudo Star, the "Regressions" list in use.perl.org/~masak/journal/40371 contains good candidates. | 12:50 | ||
| .eor | |||
| moritz_ | I'll miss the meeting too | 12:53 | |
| * with a kick-start from pmichaud++ I managed to get mostly sane Match object behaviour back | 12:54 | ||
| * got two good submissions for the MAIN routine supporting thing... now I still don't know how to plug it into Rakudo. Plans and ideas are very welcome. | 12:55 | ||
| * for the current challenge there's already one good-enough submission for Str.trans - will put that into rakudo in the next few days (or next week) | |||
| * plan to get JSON::Tiny working again (already running all the test files again), then backport into the book | 12:56 | ||
| $report.eof | |||
| masak had his $report in a given block :) | 12:57 | ||
| moritz_ | .oreally? | 12:58 | |
| masak | colomon, too :) | ||
|
13:52
masak joined
17:51
PerlJam joined
18:46
[Coke] joined
18:47
Util joined
18:48
cognominal joined
18:59
bkeeler joined
|
|||
| jnthn | oh hai | 19:03 | |
| Anyone about? | |||
| cognominal | howdy | 19:04 | |
| just a lurker here... | |||
| jnthn | :-) | ||
| Anyone about with anything to report/ask/say? :-) | |||
| If not, this is probably going to be very short. | |||
| moritz_++, masak++ and colomon++ pasted reports earlier. | 19:05 | ||
| mberends has a few items | |||
| TimToady | I'm just going through a phase. | ||
| diakopter | I, urm, made progress on a potential new backend for rakudo. dunno how relevant that is. | ||
| jnthn | On my part this week, I got backtraces working, the bracketed infixes working, and a few minor bits. | 19:06 | |
| mberends: Go ahead. :-) | |||
| mberends | you're probably aware of FakeDBI becoming reality, the most positive thing to me is recruiting Su-Shee++ as an eager assistant | 19:07 | |
| jnthn | \\o/ | ||
| jnthn is very happy to see this. | |||
| mberends | the other matter is that proto will need more testing and polishing before R *, please test it occasionally and give feedback | 19:08 | |
| .eor | |||
| bkeeler | My $job load is trending up, so I'm back in lurker mode for a little while | ||
| TimToady | maybe proto should be renamte P* | 19:09 | |
| *renamed | |||
| or Comet :) | |||
| Snowman would be the PAUSE equiv... | 19:10 | ||
.oO(where you put your modules on ice) |
|||
| mberends | :-) | ||
| jnthn | mberends: Will try and give proto and FakeDBI a spin this week. :-) | ||
| TimToady | just seems to me we're coming up on the last chance to rename proto... | 19:11 | |
| jnthn | Aye, and it's...a bit less of a prototype now than it once was. | ||
| mberends | it's a pity masak++ cannot attend, it's his babby | ||
| diakopter | "masak's babby" | 19:12 | |
| [particle] | perhaps an email to p6c can sort it out? | ||
| or wait for user backscroll | |||
| jnthn | [particle]: No, that'd probably just get more people involved than need to be. | ||
| IMO, it's up to those who are working on it at the end of the day. | |||
| I think it's worth giving a less "this is just a throwaway" sounding name though. | 19:13 | ||
| TimToady | or move from proto to plasm | ||
| mberends | TimToady: nice :) | 19:14 | |
| jnthn | :-) | 19:15 | |
| TimToady | free associations "ŠÆ" us | ||
| diakopter | if anyone has time (I'm intentionally not looking at you jnthn since you're on a critical path or 12) I'm glad to discuss planning post-R* ideas for Rakudo targetting a different backend | 19:17 | |
| bkeeler | Such as DCLR? | 19:18 | |
| diakopter | DCLR? | ||
| bkeeler | or whatever the dynamic CLR is called | ||
| diakopter | there's the DLR, but no, that's not what I mean | ||
| TimToady | DLR | ||
| bkeeler | Sorry | ||
| diakopter | bkeeler: i've been blogging about it on diakopter.blogspot.com | 19:19 | |
| bkeeler | I'll add that to my reader... | ||
| diakopter | it's on the planetsix :) | ||
| bkeeler | Ah yes, then I've been reading it | 19:20 | |
| Riiight, I remember now | |||
| diakopter | I guess I forgot to announce it, but I fixed contextuals so they work | ||
| jnthn | mberends: I guess proto naming is something to discuss when masak++ is around, anyways. | ||
| bkeeler | I thought perlesque was a completely seperate thing? | ||
| [particle] | let's call it shirley. | 19:21 | |
| TimToady | in any case, such a retargeting is for the post R* phase | ||
| mberends | diakopter: I personally would prefer to postpone extra backend discussion until July, assuming R* is completed in June | ||
| diakopter | Damian said he would be using it in his oscon talks | 19:22 | |
| oh, that's july | |||
| diakopter pauses perlesque discussion until jnthn is done with the meeting; I thought a 2min pause was it.. | 19:23 | ||
| sorry for interrupting the meeting; I thought I waited long enough | |||
| mberends | no problem, I'd like to ask if this page is still useful to plan the "Distribution" ? wiki.github.com/rakudo/rakudo/whats...nto-rakudo | 19:24 | |
| jnthn | mberends: It's still the only place we really have that I know of for doing so. | 19:25 | |
| mberends: So yes, please do ammend/update it. | |||
| (that goes for anyone) | |||
| diakopter: No worries, it *is* interesting stuff for post-R*. :-) | |||
| mberends | I'd like to ask what form the "Distribution" will take. An installable archive full of utilities? | 19:26 | |
| The reason for the question is that I would like to soon begin producing alpha versions of what this deliverable is going to be. | 19:29 | ||
| jnthn | mberends: Good question. It would be good to hear pmichaud's thoughts exactly, but I'm thinking essentially it'll be something that you can install easily, and afterwards you've got a working perl6 executable, a configured proto so you can install update modules, and the bundled modules installed (if proto can handle installing off a location on disk rather than a git checkout, I guess that we can delegate all those parts of the installation to proto). | 19:30 | |
| diakopter | note, the windows distribution on sourceforge does much of that... | ||
| mberends | let's also put random thoughts about the packaging into that wiki page, we can always refine it as we learn more. | 19:31 | |
| jnthn: proto almost gets there, but only if you give the right commands. It could get a meta-command to put all the pieces together. | 19:33 | ||
| jnthn | mberends: I was thinking we could have some "install script" of some kind that would just call proto with the right commands to do the installation of each of those. | 19:34 | |
| diakopter: Yes, that'd make a good base. | |||
| diakopter: Better than re-inventing the wheel. :-) | |||
| mberends | jnthn: I'll work on a file: source as alternative to git: and svn: | ||
| jnthn | mberends: +1, that sounds sane. | 19:35 | |
| mberends: Though...I wonder if we want that? | 19:36 | ||
| mberends: I mean, if somebody installs FakeDBI, but then we fix some bug, wouldn't it be nice if we then had it pointed to the repo, so we can say "oh yes, we fixed that, you can get latest by doing XXX". | |||
| mberends | jnthn: I'm also thinking about http: sources instead of git: and svn: for non core developers | ||
| jnthn: there is a proto.pl upgrade <ProjectName> command | 19:37 | ||
| (proto already uses Perl 5 LWP for http: downloads to self-bootstrap) | 19:39 | ||
| jnthn | mberends: http: sounds nice not just for that, but also because then people have the ability to make "releases" of their module seperate to their git master. | ||
| mberends | jnthn: indeed, end users of modules do not need all the developers' files. | ||
| jnthn | Quite. | 19:41 | |
| mberends | ok, I'll update the proto roadmap, discuss a rename with masak++, and invite all interested parties to join in the development work | 19:42 | |
|
19:42
masak joined
|
|||
| jnthn | mberends: My point was more though that, if we for example ship FakeDBI in Rakudo *. We'd maybe like to install it but then have the source set to point at a place one can get updates. | 19:42 | |
| lolitsmasak! | |||
| masak is backlogging | 19:43 | ||
| I'm all for renaming proto, but only if the new name is at least twice as good. | 19:44 | ||
| diakopter | protootorp | ||
| mberends | plasm | 19:45 | |
| masak | I'd still much rather see proto sink back into the shadows because something much better takes over. in that case, it doesn't need a rename. | ||
| that said, I'm not going to be dogmatic about the fate of proto. :) | |||
| Tene | masak: shouldn't that then be "only if the new name is at least twice as bad"? | 19:46 | |
| masak | Tene: that amounts to the same thing :P | ||
| 'plasm' sounds like an assembler of some kind. 'shirley', you're not serious? | 19:47 | ||
| [particle] | re: shirley... en.kendincos.net/video-dpntlnff-ses...-gonk.html | 19:48 | |
| masak | ah. botched a paraphrase. it's supposed to be "'shirley', you're joking?" | ||
| jnthn | .oO( Perl6 Library System. We take the acronym. pls install FakeDBI ... pls update Web ;-) ) |
||
| mberends | the best outcome would be where CPAN assimilates Perl 6 and makes a separate installer redundant. | ||
| [particle] | rename proto to pls | 19:49 | |
| masak | jnthn: :) | ||
| mberends | jnthn: not bad! :-) | ||
| [particle] | heh | ||
| mberends | what docs will ship with R*? | 19:51 | |
| anything written in or for Pod6 ? | |||
| masak | jnthn: hm. I find I cannot rule out 'pls' at once. will think about it and try it on different people. :) | 19:52 | |
| of course, it would be pronounced 'pliis' :) | 19:53 | ||
| jnthn | That'd be pliising. :-) | ||
| masak | :) | ||
| jnthn | mberends: I'm not sure we have much in the way of docs written in Pod6. | 19:55 | |
| u4x is maybe the closest to the kind of thing you'd want to ship (in that it provides the kinda docs perldoc does). | |||
| But last I looked it's only got so much content. | |||
| mberends | jnthn: the only Pod permissible in P6 source files is Pod6. I'm writing some, just to be able to work with it. | 19:56 | |
| almost every Perl 6 user is going to try the command 'perl6doc whatever', and we should provide something better than a NYI message for that. | 19:58 | ||
| jnthn | mberends: Do you have any ideas of what we could do that (a) can be done on a by R* timescale and (b) would be an improvement on that? | 19:59 | |
| mberends | jnthn: perhaps bundle Damian's P5 based Pod6 tools. grok did that. | ||
| jnthn | mberends: Are they up to date with the latest Pod6 spec? | 20:00 | |
| mberends | jnthn: no, but close enough. | 20:01 | |
| jnthn | OK. | 20:03 | |
| Anyways, I guess we can wrap up this quiet little #phasers and discuss further bits on #perl6. ;-) | 20:04 | ||
| *:-) | |||
| mberends | ok | 20:06 | |
|
20:55
masak left
21:07
bkeeler left
|
|||
| sorear | jnthn: the Synopses are written in POD6 | 23:22 | |