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.
05:38 eternaleye joined 08:04 sorear_ joined 17:15 takadonet joined 18:44 mberends joined 18:47 pyrimidine joined 18:57 samlh joined
colomon o/ 19:01
moritz_ \\o
PerlJam greetings 19:02
jnthn o/
We're a few more than last week, it seems. :-) 19:03
moritz_ is more than last time
jnthn :-)
PerlJam is here and not-distracted for a change
jnthn Let's go alphabetically rather than randomly. :-)
TimToady lurks harder
colomon I am sitting in my office rather than running around a lighthouse hundreds of miles from here. 19:04
PerlJam TimToady: lurk smarter, not harder.
colomon Is that alphabetically counting initial symbols?
jnthn colomon: I guess the lighthouse sounded like a bright idea at the time.
19:04 masak joined
moritz_ colomon: I guess that's "you first" 19:05
jnthn I guess so too. :-)
colomon okay.
I did a batch of Numeric work the last few days.
particularly making sure the Cool versions of things were reasonably consistent. 19:06
And added infix:<mod>, and significantly changed infix:<div>.
Made the single-arg closure Orderings work for min, max, and minmax. 19:07
and did a bunch of work on hash and nested hash hypers.
I've been planning on doing my best to finish up the Numeric work this week and trying to clean up Range. 19:08
19:09 _ilbot2 joined
moderator weekly Rakudo status meetings with phase transitions: Tue 19:00 UTC | IR clogs at irclog.perlgeek.de/phasers/today
colomon .eor 19:10
moritz_ colomon++
jnthn colomon++
PerlJam colomon++ indeed
masak colomon++
moritz_ I think jnthn is next in alphabetic order
jnthn I fear so.
colomon admittedly, getting sidetracked is pretty common for me. ;)
jnthn Mostly dealt with RT tickets with my Rakudo tuits.
More backtrace improvements. 19:11
Warnings get line numbers now.
(though not a backtrace)
Got attribute introspection mostly fixed up.
Discovered a $*MULTINES leak (STD has the same issue - or had...maybe fixed now)
Also finally got around to sorting out chomp on Windows 19:12
That cleared up a couple of test fails for me (and probably other Windows folk)
Started preparing for Kiev; giving a talk on signatures and a talk on R*.
Want to try and make some vaguely useful-ish thingy in Perl 6 as an example to show off and discuss. 19:13
There's a hackathon in Kiev on the Saturday too.
colomon that's this Saturday?
jnthn Yes.
General one, not just Rakudo
But "Rakudo with Jonathanom" is mentioned in the schedule. :-) 19:14
I'll help any interested folks with whatever they like - either getting a first Rakudo build, writing Perl 6 programs or hacking on Rakudo itself.
If nobody wants that, I guess I get a day of hacking time. ;-)
PerlJam "jonathanom"? sounds like you're the main course.
jnthn I'm open to suggestions of what to focus on. 19:15
PerlJam: heh, I'd never seen the instrumental case in such a tasty light. 19:16
.eor
masak next, I think?
masak not much to report. GSoC grant progressing as planned. wrapping up the Web.pm grant. submitting bugs, one a day on average.
.eor 19:17
jnthn mberends?
PerlJam masak: :( I was semi-hoping for something more exciting
jnthn Yeah, like 10 bugs a day
;-)
moritz_ like, what does "progressing as planned" mean? 19:18
masak that there's a plan, and I'm not deviating.
much.
PerlJam masak: "I'm now so far ahead of schedule that I've decided to implement something else too" :) 19:19
jnthn masak: Planning "decode" for this week?
masak jnthn: yes, among other things.
jnthn OK, sounds good.
pmichaud <-- here
jnthn \\o/ 19:20
masak will make sure Bufs generally behave as arrays and as strings.
jnthn mberends: feel free to jump in at any point you're here :-)
moritz_: you next :-)
moritz_ worked on a few challenges since last meeting, and now tries to incorporate MAIN sub into rakudo that patricaks++ has been preparing
masak my main focus will be on Web.pm this week, though. it seems to me there are certain advantages to have that ready for R*.
jnthn masak: Agree.
moritz_ I also got JSON::Tiny up-and-running 19:21
there are a few test failures, but they are due to type mismatches, not to real JSON bugs (afaict)
enough to update the book chapter (which I also did)
there are two thiings for which JSON::Tiny needs workarounds right now 19:22
1) backslash escapes in character classes
2) multis are not exported (I work around this by having an only-sub exported which re-dispatches to the mulit... ugly. Not want that in the book) 19:23
apart from that, most time is consumed by $real-world-hassle-with-new-car
.eor
jnthn moritz_: Haz we a ticket on the multis not exported?
moritz_ jnthn: don't think so... can do that if it helps you
jnthn Please do, so it's there to jog my memory 19:24
masak I don't recall a ticket for that.
moritz_ another good solution would be
jnthn That's exactly the kind of nasty that I'd rather didn't make it into R*.
moritz_ proto to-json($x) is export { nextsame }
but that NullPMCA'ed last time I tried
jnthn :/
OK
Ticket it, I'll investigate. 19:25
moritz_ will ticket it, and assign to you
jnthn proto semantics are going to have to change hugely after R* anyway.
moritz_ who's next?
pmichaud?
jnthn I think we exhausted the mmmmms.
pmichaud PerlJam?
PerlJam Um ... I've done nothing but try to understand stuff.
I'm going to be looking over the book this week though; making sure the examples really work, looking for typoes, grammaros etc, 19:26
jnthn That'd be great.
PerlJam oh, and I'm about to update the regex chapter to use "my" on them :) 19:27
moritz_ PerlJam: also feel free to leave # TODO markers if there's somethiing you find unclear, incomplete or so
PerlJam (except that I got sidetracked)
I think that's about it.
END OF LINE.
jnthn pmichaud? 19:28
pmichaud Mainly I worked on details of list implementation. I'm currently drafting a document of my observations/conclusions for people to react to as I'm implementing the darn thing.
I fixed exit() handling in Parrot, this means that "exit" in the REPL now works properly.
I worked out how to use inversion lists for character classes in regexes. 19:29
(mainly how to build them and manpulate them in the regex compiler)
fixed about a half-dozen tickets in RT 19:30
refactored the code for curlies as hash constructor -- {; a => $_ } now dtrt
(as does {;} and the other variations I could come up with) 19:31
found an intermittent bug with parsing of colonpair signatures, fixed that
refactored PAST/NQP so that Parrot's pasm constants (e.g., .CONTROL_LOOP_NEXT) remain symbolic in the generated PIR, instead of becoming integers 19:32
applied bacek++'s patch to NQP so that it can now do multies
*multis
fixed the ** range quantifier to better handle <.ws>. It's still not perfect, but it'll trip up fewer people than it did before. 19:33
moritz_ (it certainly helped me)
pmichaud fixed underlying issues with .match throwing "Can't substr beyond of string..." exceptions
that's about it for me this week.
masak \\o/
jnthn pmichaud++ 19:34
masak pmichaud++
pmichaud my plan for today is to work-work-work on lists until I'm done.
that's critical for R*.
After that, I'll fix closures, also critical for R*.
both of these should be in place by the weekend.
colomon \\o/
pmichaud after that, I plan to just run through RT and try to close more tickets.
jnthn \\o/ 19:35
Excellent.
masak wow.
pmichaud (and publish some R* notes)
jnthn pmichaud: It's *so* nice to have you back hacking on things.
pmichaud eor for me
colomon +1
spinclad \\o/ 19:36
masak indeed.
Tene what's the current goal for R*?
pmichaud Tene: you mean "goal" as in "date"?
jnthn This month.
Tene Yeah, that's what I meant.
colomon Week of the 21st? (if humanly possible?) 19:37
pmichaud Ideally, sometime in June. That way we meet our 2Q 2010 goal we set last year.
If we can do it by the 22nd, then we meet the "Spring 2010" goal, at least for the northern hemisphere.
we'd also be able to announce it at YAPC::NA
-however-
colomon +1
pmichaud I personally would prefer to release something we're all comfortable with later than to release something just to make a deadline 19:38
PerlJam What non-rakudo things will be in R* ? Is there an official list somewhere?
jnthn wiki.github.com/rakudo/rakudo/whats...nto-rakudo
spinclad we have a definite goal, but release-critical bugs take precedence ... 'when it's ready'
jnthn Is the most "official" thing so far.
PerlJam okay
jnthn Though I've no idea if pmichaud++ has reviewed and agrees with any of it. :-) 19:39
pmichaud I will make a point of reviewing it.
PerlJam what's zavolaj again?
pmichaud if I haven't reviewed by next #phasers, then I will drop whatever I'm doing to review it. :)
moritz_ PerlJam: NCI wrapper
[particle] releasing r* during spring 2010 would help dispel the myth that rakudo depends on one man 19:40
jnthn PerlJam: Maps Perl 6 signatures to Parrot NCI.
PerlJam ah. okay.
excellent
pmichaud [particle]: is that myth really prevalent?
masak never heard that myth
[particle] it exists outside the p6 circles, from my experience
prevalent? i don't think so 19:41
pmichaud [particle]: okay. I'm not too worried about that myth then.
jnthn If we ship a turd, we'll have more to worry about.
pmichaud Certainly one can look at the release history and say "pmichaud who?!"
[particle] jnthn: agreed
jnthn OK. Any other things we should discuss? 19:42
pmichaud anyone else on the reporting list?
am I last?
jnthn Ah, good point.
colomon sorear? 19:43
moritz_ what should be the criterions for modules to be included in R*?
19:43 dukeleto joined
moritz_ I mean I have two nice modules on github 19:43
but there's no reason to add a JSON parser to R*, is there? 19:44
pmichaud R* is a distribution. It can include whatever modules we think are likely to be useful to others.
PerlJam "useful" in a multi-dimensional sense? 19:45
[particle] R* needs a web presence
colomon ufo, maybe?
moritz_ +1 to ufo
jnthn Aye
[particle] faq to separate it from rakudo
jnthn +1 to FakeDBI
mberends sorry, at dinner atm...
pmichaud we can put together our initial guidelines for what we think should and should not be included, but I think it (like Perl 6 itself) will likely end up being a case of experience being the best teacher. And we should make that clear up front.
PerlJam I mean, a JSON parser is useful to parse JSON, but it's also useful as an example of how to write Perl 6.
masak ufo is most useful when installed. so something'd need to run 'make install' on it. 19:46
jnthn masak: I think all the modules we include with R* should be installed in some sense.
masak oh, good.
jnthn As in, installing R* gives you the modules "ready to run" too.
masak aye.
PerlJam As long as we don't inadvertently recreate the dual-life mess that Perl 5 had :) 19:47
pmichaud agreed, that's my thinking and intent.
masak PerlJam: that's the idea of a distribution, IIUC.
pmichaud PerlJam: we're not only recreating it, we're embracing it.
masak or *an* idea.
jnthn Mess is awesome!
;-)
pmichaud just like linux distributions have a "dual life" from the packages they choose to include. 19:48
there is no notion of "official distribution".
PerlJam ah, but something needs to wear the moniker of officialdom (in some sense) I think
pmichaud that's to be chosen by the marketplace. 19:49
jnthn IMHO, ideally though the modules that come with the distribution should be installed much like they would have been if you'd installed then with p[roto|ls] and thus updatable in the same way. But don't know if that's realistic for R*.
pmichaud jnthn: +1
masak PerlJam: Yapsi is official.
PerlJam wfm
spinclad and complete.
pmichaud PerlJam: when you can tell me which distribution of Linux is the official one, we can do something similar with Perl 6. :)
masak :)
PerlJam pmichaud: They are all official but only by virtue of sufficient care and feeding. 19:50
pmichaud PerlJam: and that becomes your answer then.
masak pmichaud: could someone from outside theoretically take Rakudo and create a non-Star distribution? 19:51
pmichaud masak: I'm hoping it's not just theoretical at some point.
jnthn masak: Someone *in practice* could. :-)
Tene I could do it right now!
masak *gasp* :)
pmichaud masak: I'm hoping that does indeed happen. Indeed, it will be a mark of success for R*. 19:52
Tene masak: Fedora is currently shipping rakudo.
masak .oO( Rakudo Steal-Our-Thunder... )
PerlJam someone should even be able to take R* and swap out Rakudo for another Perl 6 implementation
Tene Has been for quite a while.
jnthn masak: The idea of "Bio-Rakudo" for example - the Rakudo compiler + a bunch of modules for doing bio stuff - is quite a cool idea, for example.
Tene So, in at least some sense they're doing it already.
pmichaud well, fedora is shipping the rakudo compiler, yes.
masak Tene: that's not really what I meant, but sure.
pmichaud but we're talking about creating alternate distributions based on Rakudo *.
pyrimidine :)
Tene They're shipping rakudo along with an OS to run it with! ;) 19:53
Tene nods.
masak ah! a bio-lurker!
pmichaud Just like Debian has become Ubuntu and others... hopefully someday R* will be the foundation for other Perl 6 distributions.
pyrimidine looks away, whistling
PerlJam Is there anything we need to do to get R* in the various linux dists? Should we email whoever are maintaining the rakudo compiler packages for each dist? 19:54
pmichaud
.oO( I'm glad this is being logged so I'll have notes for creating my R* presentation in two weeks )
pyrimidine would be nice to see a .deb package
jnthn
.oO( I'm glad this is being logged so I'll ahve notes for creating my $* presentation in 5 days :-) )
*have, R*
:-)
TimToady
.oO( I'm glad this is being )
 19:55
PerlJam jnthn: a $* dist has some interesting possibilities :)
spinclad $*: anonymous contextual
Tene The price of R* is Whatever dollars? 19:56
pmichaud apparently, blood be the price of R* :)
PerlJam ----+ <- rapier 19:57
[particle] . <- wit
pmichaud () <- Nil 19:58
masak o.O <- masak
PerlJam so ... how did rakudo get in the linux dists? Are there soem champions among us? 19:59
pmichaud some early packagers, yes.
pyrimidine how recent is the fedora package?
pmichaud I will have to depart briefly... post further questions for me either here or in #perl6 20:01
PerlJam sends a pile of parentheses to help pmichaud with lists ;> 20:02
pyrimidine just found the answer, looks somewhat up-to-date 20:03
mberends backlogs
jnthn OK, anything else for #phasers this week? Otherwise, we can adjourn to the pub^W^W#perl6. 20:04
PerlJam not me.
colomon other than pmichaud's big projects,
anything we really feel we need for next week's release? 20:05
jnthn Not that I can immediately think of. 20:06
pmichaud I'm not opposed to holding next week's release a few days if we think it's important for R*
jnthn pmichaud: In what sense does next week's Rakudo release relate to R*? 20:07
pmichaud: That is, are you envisioning that next week's Rakudo release would be the Rakudo we put in the distribution?
(I was expecting more that we'd just pick a git revision that works well with everything.)
pmichaud Ideally, I'd like R* to use a released rakudo, yes. 20:08
jnthn Ah.
pmichaud if it doesn't work out, then okay.
if we need an extra release, that's okay too.
spinclad for R*, iwbn to get some practice packing and unpacking the distro (nice independent task from release-blockers)
jnthn spinclad: Yes - we've to date never tried to actually put a distro together. 20:10
spinclad (will help)
jnthn \\o/
Maybe some of us could start playing with that a little bit in the next week or so. 20:11
pmichaud that would be great
maybe we should put together a R* repo
mberends could we base a distribution on a rakudo-star target in Rakudo's Makefile? 20:12
spinclad look it over with many eyes
jnthn pmichaud: +1
pmichaud I'll do that when I get back from errands
(waiting on others to leave)
jnthn OK, I think we're done here.
moritz_ \\o/ 20:13
good fight, good night (or so :)
jnthn Thanks, danke, tack, dakujem, etc. :-)
20:13 pyrimidine left
mberends o/ 20:13
jnthn o/
20:13 mberends left
masak \\o 20:14
谢谢
spinclad o/ 20:15
20:18 samlh left 20:37 takadonet left
colomon \\o 20:38
22:20 eternaleye joined 23:31 eternaleye joined

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!