Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
00:00 reportable6 left 00:01 reportable6 joined, ChanServ sets mode: +v reportable6 02:13 Kaiepi left, Kaiepi joined 06:00 reportable6 left 06:04 reportable6 joined, ChanServ sets mode: +v reportable6
AlexDaniel shareable6: uptime 07:36
shareable6 AlexDaniel, 1 week, 4 days, 19 hours, 36 minutes, and 14 seconds, 525.339844MiB maxrss. This is Rakudo version 2019.07.1-94-gd1f9d2848 built on MoarVM version 2019.07.1-50-gb614a7b4d implementing Perl 6.d.
AlexDaniel jnthn: ↑ that little bot is just a Cro::HTTP::Server 07:37
so, yeah
jnthn: and that's how my htop looks like imgur.com/a/QXDu5hZ 07:39
taking more memory than gitlab is quite an achievement :) 07:41
but yeah, it's *much* better than it was before 07:42
and that's more than a week of uptime
08:49 AlexDaniel left 09:13 robertle joined 10:19 anagura joined 10:39 anagura left 10:42 anagura joined 10:47 anagura left 11:27 ufobat_ left 11:31 chloekek joined 11:55 AlexDaniel joined 12:00 reportable6 left 12:04 reportable6 joined, ChanServ sets mode: +v reportable6
AlexDaniel wifi ✓ socket ✓ 12:10
that's gonna feel quick :)
lizmat: re .LAVE, I don't think we'll be adding cute names any time soon :) 12:14
lizmat hehe... yeah... I think .code will do :-)
I think only "dd" is specifically calling .perl on what it gets... and that's not even spec :-) 12:15
AlexDaniel there's also $*PERL
call it $*LANGUAGE or whatever, I don't care 12:16
lizmat yeah, we can deprecate that for $*CAMELIA or whatever
indeed
AlexDaniel lizmat: something to think about, we probably don't want the rename to be tied to the language version (like the alias thing that happened to 6.d), but at the same time additions like .code and $*LANGUAGE will need to be added to some particular version 12:19
lizmat we could tie them to 6.e 12:20
AlexDaniel right, and the language release can happen much later, not a big deal
just please not at the same time :)
lizmat indeed...
AlexDaniel the noise about the alias was so loud that people kinda didn't even notice that we had an awesome language release 12:21
lizmat yup
sadly :-(
AlexDaniel the language release can also happen before the full rename, “Deprecated $*PERL in favor of $*LANGUAGE” doesn't even necessarily mean that rename will happen 12:22
so that kinda works too, depending on how the timeline will go 12:23
lizmat we'll see... atm I'm keeping out of the discussion as much as possible, recovering from a cold, not having slept enough and an emotional rollercoaster 12:26
12:30 chloekek left
AlexDaniel lizmat: get well :) 12:31
lizmat AlexDaniel: thank you, I will
AlexDaniel lizmat: people are getting derailed a bit but I think the discussion is mostly healthy 12:33
though yeah, I also don't want to say anything before jnthn tells if we go this way 12:34
jnthn: I wish you strength ;)
lizmat goes out for a bit& 12:38
12:55 chloekek joined 13:36 Kaiepi left 13:37 anagura joined 13:40 AlexDani` joined 13:41 anagura left, AlexDaniel left 13:56 AlexDani` left 14:04 lucasb joined 14:08 AlexDani` joined 14:10 AlexDani` is now known as AlexDaniel, AlexDaniel left, AlexDaniel joined 14:45 cygx joined
cygx AlexDaniel: on the contrary, I would argue a proper announcement of a new language name should conincide with a new spec and compiler release as well as new infrastructure (website, repositories, ...) going live 14:47
*coincide 14:48
AlexDaniel cygx: been there done that
not again
cygx what do you mean? were you involved with some other programming language changing its name? 14:49
because the 'Raku' thing wasn't that
AlexDaniel yes, but Raku announcement completely shadowed the language release 14:50
14:55 chloekek left
AlexDaniel also I'm not sure how fast we'll be able to do 6.e 14:56
vrurg: wanna be the release manager for the language?
vrurg: to clarify, that's unrelated to compiler releases. Last time Zoffix did the language release while I was doing rakudo releases
cygx I would argue the trigger on the name change should only officially bepulled *after* all other pieces are in place 14:57
I would argue *that* was the mistake last time
Perl6 can now also be called Raku - so what? 14:58
AlexDaniel cygx: that's correct, it was basically “ok here's the name, we haven't figured out what it means in practice”
cygx: a full rename is more straightforward, and I think the plan is to include more details in the PR
cygx: but I'd guess a TL;DR of it will be “if you see Perl 6 anywhere, change it to Camelia (or whatever), possibly adding “previously known as Perl 6” if there's a need for that” 14:59
lizmat but please, don't start doing that until we have a decision 15:01
cygx not only a decision, but also a proper plan how the transition should happen 15:03
I don't want it to happen at all, but if it does, please don't drop the ball
lizmat cygx: understood, we want to make new mistakes 15:04
cygx exatcly ;) 15:07
I slightly worried when I see the claim that a rename would be 'straightforward' 15:09
timotimo lots of things get renamed, not many things get an alias
cygx there are lots of things that need to happen for that (changes to the setting, a new website, fixing the documentation, new marketing material, renaming the binary, ....) 15:10
if you do it piece by piece, you probably lose all the momentum the change might bring...
.o( Raider heißt jetzt Twix – sonst ändert sich nix! ) 15:12
AlexDaniel cygx: when I say straightforward I mean straightforwarder than an alias :) 15:14
timotimo all we need is a straight forwarder
AlexDaniel timotimo: should've used that instead of pumpking :) 15:15
too late anyway
timotimo i wouldn't want it to sound like we only want straights to take that position
cygx we really should have bound that name instead of assigned it
reference semantics would have made things trivial... 15:16
timotimo if we had something like pypy where you can compile the interpreter with a different objectspace and then objects get a "become" method ... 15:17
cygx btw, have people given some thought to stuff like the books, the TPF grants, etc? 15:43
bytes are easy to push, hardcopies and money less so? 15:44
timotimo it'll still be a perl-family language, so i'd imagine TPF would still be there for it 15:45
16:02 chloekek joined
AlexDaniel cygx: people have thought, people haven't decided on a precise plan 16:13
yet
currently the only question is whether we really want to go in that direction, which jnthn will decide 16:15
then there will be a PR, which we'll be ironing out for some time 16:16
16:21 AlexDaniel left 17:12 chloekek left
cygx the proof-of-concept for freezing precompiled modules seems to work: 17:20
github.com/cygx/p6-freezer
copies the precompiled moarvm files to cwd and may load them again
17:37 AlexDaniel joined
timotimo now put everything in a single file :) :) :) 17:53
cygx that's the easy part 17:54
timotimo base64 for the bytecode files?
cygx why? perl6 can open binary files just fine 17:55
just write a header that encodes the offsets/lengths
timotimo AFAIK perl6 will currently refuse to run a script that has invalid utf8
cygx oh, you mean embed it into the script itself? 17:56
timotimo yep
so that it's all literally just a single file
cygx thanks to pamplemoussse, we can now precompile scripts, you know 17:57
timotimo right
18:00 reportable6 left
timotimo so basically write a script that'll at compile time read the files into buf8 objects and executes it, then turn that into a precompiled script 18:02
18:02 reportable6 joined 18:31 lucasb left
cygx timotimo: I now use a precompiled module to store the bytecode blobs 18:56
seems to work fine
timotimo damn 18:57
cygx next on the agenda, parsing a META6.json file and generating a single .moarvm file from that 19:00
that might actually be useful for something
but not today
o/ 19:19
19:19 cygx left 20:00 japhb left
vrurg AlexDaniel: I saw your proposal and it's interesting. Need to consider it carefuly though. Do we have a list of duties for language RM somewhere? 21:31
AlexDaniel vrurg: just this: github.com/perl6/roast/blob/master...e-guide.md
vrurg Aha, the one I once edited. ;) 21:33
I'm honored! Need to consider if I can handle it well enough. 21:34
AlexDaniel vrurg: but actually, we don't even have a list of actual duties for rakudo releases. Sorta discussed that at TPC a bit. Sure there's a release guide, but some things you just decide to do because you think it's right to do them 21:35
vrurg I don't clearly understand yet would is included into the range of duties. But considering that most of my latest activies are one way or another related to 6.e, I really like it. 21:40
AlexDaniel vrurg: well, the end goal is to declare 6.e released and start a 6.e-errata branch. You decide how you get us there :) 21:44
we also have tags? 21:45
why?
I don't know
vrurg AlexDaniel: Aha. In either case, 6.e won't be ready before the next year. So, plenty of time. 21:47
AlexDaniel vrurg: why so?
vrurg: I'm not objecting, just want to know why :)
21:48 MasterDuke left
vrurg AlexDaniel: how many features planned for 6.e we have? Unless I miss something, but just few so far. 21:48
AlexDaniel vrurg: yeah, we have to figure out how we want to create and keep that list 21:50
vrurg: we had that repo for 6.d: github.com/perl6/6.d-prep
vrurg: but I'm not happy about it, I don't want a new repo for every language release
vrurg Exactly. problem-solving + 6.e lablel would suffice. 21:51
AlexDaniel sounds about right 21:52
vrurg Maybe a wiki page on problem-solving for the current status of implemented changes, but this is optional. 21:53
AlexDaniel vrurg: or a ticket like github.com/perl6/problem-solving/issues/68 or github.com/perl6/problem-solving/issues/64 21:54
vrurg: then people also can leave comments
vrurg: or github Projects feature
vrurg BTW, if name change gets approved, I think the '6' in language versions would remain. 21:55
I never used Projects before. But the name suggests that it's about the right thing for the task.
AlexDaniel like “camelia v6.d”? Looks alright to me
although, to be completely honest, letters are a bit too unusual… 21:56
like why not semver? but whatever… 21:57
then we'd be able to do proper versioning of language versions instead of forever-changing errata branches
I don't know… Not trying to make any point just yet, just my thoughts out loud 21:59
vrurg Just lets have some traditions be there. Somehow I like 6.x versioning. Maybe because it's distinctive from other languages.
AlexDaniel did that a lot at TPC too :)
vrurg: yeah, it probably doesn't hurt. Something about this snowflake way of doing things just annoys me lately, it's just me I guess 22:00
vrurg Hope your TPC was better than this year in Pittsburg.
AlexDaniel: besides, we have enough hustle with name change. Doing the same to the versioning would be an overkill... ;) 22:01
AlexDaniel vrurg: TPC was amazing. Really positive atmosphere regarding perl 6 and its future. 22:02
vrurg: what happened in Pittsburg?
vrurg Less people than last year, not so many interesting presentations. Two years ago in Washington it was tremendously good, I was breaking myself apart, trying to visit everything I conisdered interesting. 22:03
This time it was like 2-3 really interesting talks to me. 22:04
AlexDaniel I missed a bunch of talks… 22:05
vrurg If things go well, I plan to start own tutorial on Perl 6 though with focus in how things are implemented, not how they work. 22:06
AlexDaniel so can't comment on that, but the ones I've seen were pretty great
vrurg People do great job on their presentations, that's no doubt. But the subjects... 22:07
AlexDaniel vrurg: jnthn's keynote was really uplifting
vrurg Ha, I bet it was! Perhaps I'm disappointed because there was Damian and Liz previously. Not this year. :) 22:09
On the plus side, my spoken English got better and I had more contacts with people. :) 22:10
AlexDaniel vrurg: yeah but liz did a keynote too
vrurg Hope she gets her US visa problems resolved for the next year. 22:11
Anyway, "ok ramblers, let's get rambling!" I promised Xliff to resolve a problem with `require`. CompUnit is likely to return a pre-cached Stash without the module being currently imported. Took all my day yesterday just to get to this point. 22:13
And I take the job. I like it, the rest is not that important. :) The only thing currently worries me is that I might start looking for a second job around. But that is not decided yet. 22:14
AlexDaniel good! 22:15
thank you :)
vrurg Thank you too! As I said, it's honor too. I didn't expect to go that far just a bit over a year ago when first installed rakudo on my macbook. ;) 22:16
23:36 robertle left 23:37 robertle joined
vrurg Is anybody still working on precompilation fixes? 23:51

Powered by Cro and the Raku® Programming Language.

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

Thank you!