Parrot 2.8.0 released | parrot.org Log: irclog.perlgeek.de/parrot/today | Nopaste: nopaste.snit.ch:8001 | close 25 tickets; remove deprecated items (especially CodeString);
Set by moderator on 24 September 2010.
dalek kudo: d4b72da | KodiB++ | src/builtins/Signature.pir:
[builtins/Signature] Use ObjectRef instead of Perl6Scalar.
 00:09
kudo: fb57ef1 | KodiB++ | src/Perl6/Grammar.pm:
[Grammar] Permit ">>" after infix:<->.

This fixes RT #77800.
kudo: 3221cad | KodiB++ | src/ (6 files):
Merge branch 'perl6scalar-cleanup'
ast: 71fdf1d | KodiB++ | S03-metaops/hyper.t:
[hyper.t] Added tests for RT #77800.
 00:10
00:11 theory left 00:54 davidfetter left
dukeleto whiteknight++ # dealing with the IRS 01:10
01:11 Coke joined
whiteknight blah 01:11
I didn't get a chance to play with that git filter-branch thing today 01:12
there's always tomorrow 01:16
cotto or is there?
whiteknight maybe not 01:20
dukeleto whiteknight: i am adding you to the GitMigration wiki page 01:28
whiteknight nice 01:30
01:31 whiteknight left
dukeleto whiteknight: "Now you's can't leave" 01:31
cotto Item 4: whiteknight 01:40
Speaking of slackers, that guy who is me needs to finish up the github plugin tests. I'll get after him to do that. 01:42
Crud. There's work involved.
dukeleto cotto: how is that stuff going? 01:53
01:53 Andy joined
dukeleto msg whiteknight let me know if you have questions about languages.git+submodules, I am still thinking about the details 01:53
aloha OK. I'll deliver the message.
01:58 theory joined
cotto dukeleto, well. I suspect the last bit of coding will be to dtrt with ambiguous commit ids. 01:59
dalek tracwiki: v36 | dukeleto++ | GitMigration 02:01
tracwiki: trac.parrot.org/parrot/wiki/GitMigr...ction=diff
02:35 janus left, janus joined
dalek tracwiki: v11 | cotto++ | GitHubTracPluginTests 03:09
tracwiki: trac.parrot.org/parrot/wiki/GitHubT...ction=diff
03:12 ash_ joined 04:57 Andy left 05:33 theory left 06:26 allison left
dalek thub-trac: e74f64d | cotto++ | TODO:
knock off some todone itmes
 06:42
thub-trac: 2e69984 | cotto++ | github/github.py:
make code slightly smarter about dealing with ambiguous commit identifiers
thub-trac: 62a6b3d | cotto++ | github/github.py:
rename a variable for consistency
thub-trac: 3d10019 | cotto++ | github/github.py:
fetch multiple ambiguous commits when applicable
thub-trac: ca23f16 | cotto++ | github/github.py:
typo fix
 06:43
thub-trac: caaa5eb | cotto++ | github/github.py:
tighten a regex to avoid catching r0
thub-trac: 5186eab | cotto++ | github/github.py:
redirect to github when the hook works
thub-trac: b30d31d | cotto++ | github/github.py:
don't link to an ambiguous commit id (not optimal, patches accepted)
thub-trac: 9fd2a1e | cotto++ | github/github.py:
be a little more paranoid about db corruption
06:44 jsut_ joined 06:49 jsut left
dalek tracwiki: v12 | cotto++ | GitHubTracPluginTests 06:52
tracwiki: add a slot for another batch of test results
tracwiki: trac.parrot.org/parrot/wiki/GitHubT...ction=diff
cotto I'm going to call that "ready". 07:05
contradictions welcome 07:06
as are additional test cases
bacek_at_work ship it! 07:07
cotto in contrast to the external client that is inexplicably hitting my test svn server on localhost at port 8080 asking for google.com
dalek thub-trac: 3f755db | cotto++ | github/github.py:
tooltip improvement and unhelpful debug statement removal
 07:08
tracwiki: v13 | cotto++ | GitHubTracPluginTests 07:09
tracwiki: I love the smell of passing tests in the morning.
tracwiki: trac.parrot.org/parrot/wiki/GitHubT...ction=diff
cotto is aloha hosted in a reliable place? 07:10
07:16 fperrad joined
cotto bacek_at_work, ^ 07:16
bacek_at_work cotto, much better place now. 07:26
It's on my devbox @work
cotto sounds like a perfectly stable place that'll never run anything unstable ;) 07:27
bacek_at_work cotto, of course not! I'm not developer anymore. At least officially. 07:31
cotto really? You not writing code sounds like a waste.
bacek_at_work heh. 07:33
I wrote 15 lines of code recently.
In whole 2 weeks time.
otoh, someone have to design systems end-to-end. 07:34
cotto indeed
you know what you're doing in that area
bacek_at_work Yeah. But sometimes I'm way too tired of Visio and PowerPoint. 07:35
But they are my main tools at $dayjob now.
cotto You're only rising to your level of incompetence.
bacek_at_work I didn't reach it yet.
But World Domination is just around the corner :) 07:36
cotto assuming it exists
there's always Parrot's GC and PIRATE
(and a dozen other parroty things)
bacek_at_work yeah 07:38
moritz it seems a recent parrot change broke line numbers in rakudo
bacek_at_work And I'm going to make GenGC ready for 2.9 :)
cotto bacek++
07:50 tadzik joined 08:01 ruoso left 08:14 ruoso joined
dalek TT #1807 created by moritz++: Parrot changes between r49277 and r49386 break rakudo's line number ... 08:26
TT #1807: trac.parrot.org/parrot/ticket/1807
bacek ~~ 08:55
10:10 contingencyplan left 10:29 ruoso left 11:01 tadzik left, tadzik joined 11:08 sjn left 11:09 sjn joined 11:29 bacek left
Coke ~~ 11:31
11:37 tadzik left, tadzik joined 11:52 whiteknight joined
whiteknight good morning, #parrot 11:55
moritz good localtime, birdfolks 11:56
whiteknight msg dukeleto Thanks, I'll definitely ask if I need anything. This is a test, and one that I think I can run several times before I have to get it right
aloha OK. I'll deliver the message.
whiteknight has imported the /languages repo into git 12:56
moritz whiteknight++
whiteknight irclog? 12:58
blah
moritz irclog.perlgeek.de/parrot/today
whiteknight github.com/parrot/c99 13:06
irclog is irclog.perlgeek.de/parrot/today
irclog?
aloha, irclog is irclog.perlgeek.de/parrot/today
aloha whiteknight: ... but irclog is irclog.perlgeek.de/parrot/today ...
whiteknight >:(
moritz aloha: irclog?
aloha moritz: irclog is irclog.perlgeek.de/parrot/today
moritz whiteknight: I think aloha tries to be less intrusive by mostly speaking only when spoken to 13:07
whiteknight: I for one welcome our new, silent overlord bot
whiteknight github.com/parrot/bf 13:12
moritz: blah. I demand the bot must do what I want, as soon as I want it!
13:19 Patterner left 13:29 Psyche^ joined, Psyche^ is now known as Patterner
atrodo bf++ 13:31
tadzik is lua for Parrot visible for C, as liblua in the official implementation? 13:42
whiteknight tadzik: in a manner of speaking, yes. You could load libparrot and call methods on the lua compiler 13:44
we don't have a friendly wrapper that would act like a liblua
13:48 patspam joined, luben_work joined
tadzik wklej.org/id/395502/ -- is it known? 13:56
13:58 smash joined
smash hello everyone 13:58
tadzik hello
14:14 luben left
whiteknight tadkiz: probably not. fperrad usually keeps lua pretty well maintained 14:14
send him a message about it 14:15
(unless you can fix it yourself, then send him a friendly patch!)
tadzik I doubt it :) I have very little experience in Parrot
but I can file that on Github, that's not a problem 14:16
whiteknight that's cool too. Bug reports are always valuable
github.com/parrot/forth 14:46
14:58 allison joined
Coke allison: ~~ 15:04
allison Coke: hi 15:05
15:06 theory joined
whiteknight hello allison 15:09
allison hi whiteknight 15:10
15:17 kid51 joined
whiteknight allison: I sent you an email a while back about the state of the 1023 form. Did you get a chance to look at that? 15:20
allison whiteknight: will look for it now 15:21
15:21 ruoso joined
allison whiteknight: I know I replied to one on the directors list, but this may be more detailed questions 15:22
whiteknight probably much more detailed. Not a huge rush. The single biggest question I have is about the NLNet grant, which I'm sure you can answer pretty quickly
allison whiteknight: I'm not seeing it, do you have an approximate date and subject? 15:23
whiteknight: the NLNet grant is complete, and I don't expect to be paid for the rest of my work (consider it a donation)
whiteknight The email was titled "Parrot Directors" for some reason, and was sent on Sept 6 15:24
The question I have about NLNet is who was it paid to, the Parrot or Perl foundations?
I only see an invoice for about half of the money in the PaFo documents 15:25
exactly who took receipt of that money, and exactly when it was received is going to dramatically affect our average receipts, and may bump us up to a higher IRS filing fee
Coke the grant was initially processed through the TPF. 15:26
whiteknight okay, but was it completed through PaFo? I see an invoice in the documents folder, but no indication whether that invoice was paid to PaFo
Coke (yes, I know that's the the. I blame the TPF. ;) 15:27
allison whiteknight: which invoice? (path in Dropbox folder)
Coke I believe the TPF cut us a check for the non-disbursed amounts so we could handle the remainder of disbursement.
allison whiteknight: the initial money was received by TPF in two payments of $35k each from NLNet 15:28
Coke (and that's the limit of my understanding, I never saw any actual docs.)
whiteknight allison: funding/nlnet/nlnet_invoice.pdf. 15:29
okay. I guess I'm trying to figure out why that invoice doc is in there at all. It is on TPF letterhead, so I'll assume it was paid and handled to TPF
atrodo (Coke> same gripe i have with PDF files) 15:30
allison whiteknight: okay, that is a tpf invoice, and was paid to tpf (is way before ParF existed)
whiteknight: let me dig through email archives for the exact amount transferred to Par 15:31
F
whiteknight okay. That clears me up. Thanks. so the money that you earned, you've donated back to PaFo. The rest of the unpaid money was given to TPF for final dispersal?
er, given by TPF to PaFo?
allison whiteknight: yes, almost. Up to a certain date, TPF handled all payments. After that date, TPF transferred all remaining funds to ParF, and ParF handled all payments. 15:32
whiteknight gotcha. That's probably very good news for us 15:33
allison whiteknight: Parf has made almost all due payments to grant recipients.
whiteknight: some remaining payments are due to me, and I'm donating them back. Some remaining payments are due to others, and we need to double-check with them.
whiteknight: we are tracking the status of all NLNet grant funds and payments in a google doc, I'll share it with you 15:34
whiteknight Jerry already shared it with me
allison whiteknight: great
whiteknight So I'm good on those details. I'm primarily interested in putting together our financial records so I can get form 1023 completed ASAP 15:35
allison whiteknight: the ones marked in green are the ones I'm still not sure of (that is, I don't know if you, patrick, jonathan, and simon were paid) 15:36
whiteknight: jhorner will have those details (will probably involve looking over our accounts for the past 2 years to be sure) 15:37
whiteknight I was not paid, I'm choosing to keep that money there for now to keep some padding in the account
so that's not an issue
allison whiteknight: makes sense 15:39
whiteknight: I found your message, will reply in detail
whiteknight thanks, I would really appreciate it
again, no huge rush. I doubt I will be in a position to complete the paperwork within the next week or two
allison whiteknight: but to answer the quickest and most important question: Garvey Schubert Barer were the original lawyers in our incorporation 15:40
whiteknight ok
allison whiteknight: we were planning to use them for the 501(c)(3), but they initially quoted us $1,000 for the cost, and later quoted us $20,000 for the cost
whiteknight: so, we asked SFLC if they could help 15:41
whiteknight oi
allison whiteknight: Karen Sandler is the lawyer handling us at SFLC
whiteknight: they will only charge us the cost of the registration fee with the IRS, which is about $750
whiteknight: what's needed now is to get Karen the final pieces of information she needs to complete the forms 15:42
whiteknight: mainly, a summary of our financials since incorporation, and bio info and job descriptions for all board members
whiteknight: on the other details, she'll have the best answers 15:43
whiteknight: it may make the most sense for you, me, and her to have a quick catchup phone call sometime in the next couple of weeks.
kid51 is very glad to see these issues being discussed :-) 15:45
15:51 cogno joined
whiteknight allison: yes, that likely would make good sense. I may try to contact her when I get more of the details put together myself 15:55
allison whiteknight: I'll CC her on my reply to your message. 15:56
whiteknight thanks, that would be most helpful 16:04
what would be really cool, if we can do it, is find somebody to make us an "inaugural donation" in time for the acceptance of our paperwork 16:06
wishful thinking, and all that
dalek rrot: r49387 | pmichaud++ | trunk/compilers/pct/src/POST/Compiler.pir:
[pct] Revert another retrogression introduced by r49325.
 16:15
16:15 cogno_ joined 16:16 cogno left
fperrad ping tadzik 16:18
dukeleto whiteknight: we won't receive any donations unless we ask people 16:21
whiteknight++ # working on PaFo stuff
16:21 cogno_ left
whiteknight dukeleto: of course not. I'm just musing out loud. I can't give an ETA on the completion of the paperwork anyway 16:22
dukeleto whiteknight: i am wondering if we can ask TPF for some of the money that Google will give them for GSoC 16:23
16:24 ingy left, nwellnhof joined
nwellnhof kid51: ping 16:25
kid51 nwellnhof pong
nwellnhof i can't reproduce the examples/shootout failure
do you have mnore details?
tadzik fperrad: pong 16:26
smash whiteknight: ping 16:27
nopaste "kid51" at 192.168.1.3 pasted "string_macros branch: t/examples/shootout.t test 16 fails" (43 lines) at nopaste.snit.ch/23876 16:28
dukeleto whiteknight++ # converting various languages to git
16:29 davidfetter joined
fperrad tadzik, the interactive mode of luap seems broken, but if you put your Lua source in a file, it works fine 16:29
allison whiteknight: reply sent, with karen cc'd 16:31
tadzik fperrad: I see
dalek rrot: r49388 | nwellnhof++ | branches/string_macros/src/string/encoding/utf16.c:
[str] Fix non-ICU build
kid51 nwellnhof: To be fair, I'm getting this failure in trunk as well. 16:32
nwellnhof kid51: yes, it looks like an OOM condition caused by the new GC code.
kid51 Yes, ISTR discussion of that possibility several days ago. 16:33
nwellnhof do you know if gc_ms2_tuning passes this test?
kid51 Don't know; will have to checkout and test. 16:34
16:35 contingencyplan joined
kid51 But I suspect any branch forked from trunk subsequent to gc_massacre merge will display this failure. 16:36
nwellnhof kid51: with how much memory are you testing? 16:37
kid51 From 'top': 16:38
Mem: 524512k total, 245280k used, 279232k free, 51828k buffers
Swap: 263160k total, 21448k used, 241712k free, 133104k cached
nwellnhof ok, then gc_ms2_tuning probably doesn't help. 16:39
sorear kid51: /usr/bin/free with linux 2.6.32 will report peak RSS without polling, btw
kid51 total used free shared buffers cached
Mem: 524512 312624 211888 0 56344 186444
-/+ buffers/cache: 69836 454676
Swap: 263160 21448 241712
sorear erp, /usr/bin/time 16:41
kid51 sorear: I don't appear to have /usr/bin/time
'which time' -> no output 16:42
16:42 cogno joined
kid51 nwellnhof: Confirm that in gc_ms_tuning branch, t/examples/shootout.t test 16 FAIL 16:42
... with output similar to what was pasted
16:43 dukeleto left, PerlJam left, pmichaud left, dalek left, Util left
nwellnhof hmm, examples/shootout/regexdna.pir uses about 1GB memory. 16:44
kid51 has installed time from debian 16:46
nwellnhof: It was passing in trunk before GC merge.
sorear: What format/options to /usr/bin/time would be helpful here? 16:47
sorear kid51: the default should show (24954maxresident) or so 16:48
kid51 time running which program? prove?
Running : time ./parrot "/home/jimk/work/gc_ms2_tuning/t/examples/shootout_16.pir" < examples/shootout/regexdna.pir_input 16:50
... I get:
Killed
real 0m17.757s
user 0m3.701s
sys 0m1.852s
sorear probably the parrot process itself 16:51
kid51 such as what I just posted?
sorear that kind of output comes from the bash builtin 'time'
use a full path to use the 'time' you just installed'
kid51 Command terminated by signal 9 16:52
3.65user 1.39system 0:14.70elapsed 34%CPU (0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (289major+184151minor)pagefaults 0swaps
16:54 cogno left
sorear 0maxresident -> bad OS 16:54
kid51 similar output in string_macros branch
What do you mean "bad OS"? This is Debian Lenny 16:55
I've been building/testing Parrot on this box since Nov 2006.
sorear Yes, but I didn't get nonzero maxresident values until a kernel upgrade+reboot yesterday 16:56
getrusage(2) says supported since 2.6.32 16:57
kid51 uname -a
Linux li11-226 2.6.18.8-linode22 #1 SMP Tue Nov 10 16:12:12 UTC 2009 i686 GNU/Linux
Well, we simply can't be designing Parrot such that its tests no longer work except on the latest and greatest Linux kernels. 16:59
Any more than we can be designing Parrot such that it compiles only with 4G physical memory 17:00
sorear Oh, no 17:01
I meant for personal use
As in, you can use this to better quantify "Parrot uses $A_LOT of memory in you $ACTION"
moritz example: compiling rakudo :-) 17:02
kid51 Is the amount of pagefaults reported by /usr/bin/time meaningful?
Rakudo is in the unfortunate position of no longer being compilable on the kinds of laptops Rakudo developers used for most of its existence. 17:03
moritz it was significantly faster to compile on the last release than on HEAD 17:04
kid51 moritz: By HEAD, do you mean "Rakudo HEAD on Parrot HEAD"? 17:06
whiteknight allison++ 17:09
moritz kid51: yes 17:11
kid51 moritz: That would be consistent with fact that gc_massacre merge took place after Sept Parrot release. 17:12
moritz (rakudo changed not much since the release, so I'm sure it's a parrot change) 17:15
17:16 mikehh left
Coke kid51: you might want to start stumping for this ticket: 17:22
trac.parrot.org/parrot/ticket/827 17:23
That should make it somewhat easier to have a smoke test somewhere that verifies we're under some threshold.
whiteknight github.com/parrot/lisp 17:24
kid51 whiteknight: you're the owner on that one (TT #827) 17:25
Coke: Sounds like the one I created 4 days ago: trac.parrot.org/parrot/ticket/1805 17:26
Coke 2 different options, but yah. 17:29
17:33 kid51 left
smash whiteknight: can i adopt punie ? 17:33
Coke that's allison's. 17:34
was, at least. 17:35
smash i'm mostly sure that she'll be okay with that 17:36
but i can start a new one, no problem
moritz smash: at YAPC::EU, it seemed her plan was to start a new one too :-) 17:37
no, sorry, confused with pynie
smash: just write a mail to parrot-dev; I'm nearly sure nobody will speak up against it
whiteknight smash: sure thing. On github it's easy enough to fork it if you want, or I can create the git repo and push it under your username if you prefer that 17:40
github.com/parrot/lolcode
smash no problem, i can create a new one
moritz: pynie is too much python for me :P 17:41
17:41 cogno joined
moritz smash: :-) 17:41
whiteknight smash: I might be interested in helping with a punie language, but I don't know anything about Perl 1
17:41 cogno left
whiteknight so if somebody can find a reference to follow, I will help out 17:42
smash whiteknight: i want to revamp it to use nqp-rx, i already have the building it place, but still can't parse anything usefull yet, net to fix somethings yet 17:43
s/net/need/
whiteknight okay, that's cool. Do you think I should port the one from the languages repo to github, or just wait for yours to come to fruition? 17:44
smash err, i'm not sure when i will have time to get something usefull, it can take a while now.. i'm completly overwhelmed with work 17:45
whiteknight to be fair, the old punie isn't useful either 17:46
smash true, let me take a better look at my current version over the weekend and i'll let you know the state of it on monday 17:48
but feel free to port the current version anyway, even for history history i guess 17:49
i'm pretty sure allison is okay with it, i already talked to her about this 17:53
nwellnhof kid51: r49389 should fix the examples/shootout test 17:56
whiteknight github.com/parrot/ook 17:59
tadzik I like the TODO section
allison smash, whiteknight: yes, I'm cool with smash driving punie (and the direction he's taking)
whiteknight yay!
smash allison: thank you 18:00
allison whiteknight: and the only real reference on Perl 1 is the original implementation 18:01
whiteknight: plus a few patches by schwern to make it run on modern hardware/compilers
whiteknight allison: fun. Where can I get a copy of such a beast?
allison whiteknight: they're available publicly, or I can tar up my already patched copy and send it to you 18:02
smash whiteknight: you have it on git
*github
there's a branch with perl1 18:03
github.com/mirrors/perl/tree/perl-1-maint 18:04
nwellnhof tadzik: can you try if r49389 helps with TT #1804? 18:06
tadzik nwellnhof: well, the server it was affecting is currently down, not to say burned (literally)
nwellnhof: I'll check it out once it's back 18:07
nwellnhof how much memory did that machine have?
tadzik 2.5 GB 18:08
nwellnhof do you an idea how much of that was free? 18:09
18:10 mikehh joined
nwellnhof s/you/you have/ 18:10
whiteknight wow. Perl 1 is not nearly as small and archaic as I would have guessed 18:11
tadzik nwellnhof: juerd.nl/i/d58dc3794529cdd4337f43e2fd044204.png I was given such answer :)
seems like around 1.5 GBs 18:12
nwellnhof hmm, that should have been enough even before r49389. 18:14
18:18 davidfetter left 18:20 smash left 18:23 perlpilot joined
whiteknight github.com/parrot/pheme 18:27
moritz whiteknight++
tadzik Error validating server certificate for 'svn.parrot.org:443': -- has anything changed recently?
18:31 nwellnhof left
whiteknight github.com/parrot/punie 18:31
18:36 bacek joined 18:45 whiteknight left
Coke tadzik: works for me here on IE8. 18:47
what browser are you using?
tadzik Coke: that was rakudo build with --gen-parrot switch, so I guess it was just an svn client. Works ok from browser too, and synced when I allowed it 18:48
Coke works in chrome...
tadzik weird
19:01 luben_work left
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#360) fulltest) at r49389 - Ubuntu 10.10 beta amd64 (g++-4.5 with --optimize) 19:13
19:22 nwellnhof joined
GeJ Bonjour everyone. 19:37
cotto hio GeJ 19:45
GeJ ?Heya cotto. 19:46
19:53 jsut joined 19:58 jsut_ left
GeJ nwellnhof: ping 20:01
nwellnhof GeJ: pong
20:02 perlpilot left, luben joined
GeJ nwellnhof: in the string_macros branch, src/string/encoding/shared.h on line 288, you declared a ASSERT_ARGS_encoding_ord_error macro but didn't use it anywhere. 20:04
(don't blame me, it's kid51 who asked me to make the codingstd test pass)
nwellnhof GeJ: The ASSERT_ARGS_* macros are generated by headerizer 20:05
Maybe I forgot to run make headerizer
GeJ That I didn't know.
nwellnhof Is that the only codingstd failure?
GeJ my C-fu is nonexistent. 20:06
except for a file that lacks some svn metadata, yes that is all.
nwellnhof OK, I will look into it. 20:07
GeJ I can handle the metadata part if you want. 20:10
nwellnhof I have no idea what's going on with that SVN metadata stuff. I think it's because I use git-svn. So if you can fix SVN things, that would be nice. 20:13
GeJ On it.
20:14 patspam left 20:19 kid51 joined 20:37 dukeleto joined
nwellnhof kid51: r49389 should fix the examples/shootout failure. 20:46
kid51 Yes. It does.
At least, I've confirmed that on my Linux/i386.
nwellnhof good to hear.
kid51 Now trying it on Darwin/PPC. Keep fingers crossed
20:47 fperrad left
kid51 The shootout's memory demands are still sufficiently intense that on this 256M phys mem box, I could not simultaneously run 'make test' and 'prove t/examples/shootout.t'. 20:48
Are you going to port the changes that made shootout.t pass in trunk to string_macros branch? 20:49
And/or gc_ms2_tuning branch?
nwellnhof I will definitely port them to the gc_ms2_tuning branch. 20:50
I can port them to the string_macros branch, too. It's no big deal.
kid51 If you or anyone could suggest some "memory management for dummies" thing to read, I would be grateful 20:51
Or (what may be the same thing) "garbage collection for dummies"
nwellnhof Wikipedia should be a good start. 20:52
Parrot's GC isn't very sophisticated (for now). It's basically what the Wikipedia GC page calls "naïve mark-and-sweep". 20:55
kid51 nwellnhof I've closed TT #1803 -- but for future reference, you may want to add something there about what was wrong and why your code revision fixed (or ameliorated) it. 21:00
nwellnhof kid51: yes, we should add a test for that. something like the gc-leaky tests we already have. but they're still disabled for the new GC. 21:01
kid51 goes hunting for new parroteers in northern climes. 21:06
21:06 kid51 left 21:20 whiteknight joined
GeJ nwellnhof++ # fulltest passing on branches/string_macros 21:22
whiteknight nice 21:23
21:39 nwellnhof left
GeJ is dalek dead? 21:43
Ahhh, now it comes back to me. A long time ago I had this crazy idea about messing with Data::Dumper. 21:45
cotto It was him or us.
jnthn GeJ: Yes, server explosion. 21:46
Well, apparently there was smoke, anyway. 21:47
GeJ ouch.
jnthn Juerd++ is working on it. :-)
oops
GeJ oh, feather blew up?
jnthn oh, no, that was the right window
GeJ: Yeah. 21:48
21:52 sjn left 21:53 sjn joined 21:54 dukeleto left 21:57 ruoso left 21:58 davidfetter joined 22:34 allison left
cotto seen dukeleto 23:00
aloha dukeleto was last seen in #parrot 2 hours 22 mins ago joining the channel.
23:00 gottreu left 23:11 whiteknight left
GeJ clock 23:15
clock?
aloha GeJ: LAX: Thu, 16:15 PDT / CHI: Thu, 18:15 CDT / NYC: Thu, 19:15 EDT / UTC: Thu, 23:15 UTC / LON: Fri, 00:15 BST / BER: Fri, 01:15 CEST / TOK: Fri, 08:15 JST / SYD: Fri, 09:15 EST
23:21 whiteknight joined
bacek_at_work ~~ 23:25
cotto ~~ 23:27
whiteknight ?? 23:28
bacek_at_work !! 23:34
whiteknight $$\\ 23:38
23:39 tadzik left
GeJ G'Day bacek, Hi whiteknight. 23:39
whiteknight hello GeJ 23:40
bacek_at_work G'Day GeJ
GeJ At the PIR level, what's fastest way to create strings (potentially very long strings built via a lot of concatenations) ? 23:43
whiteknight StringBuilder PMC
cotto Note the description: TODO: Add description here. 23:45
23:48 ruoso joined 23:58 tcurtis joined
tcurtis Good evening, #parrot. 23:58
bacek_at_work seen dalek
aloha dalek was last seen in #parrot 7 hours 26 mins ago saying "parrot: review: trac.parrot.org/parrot/changeset/49388/".

Powered by Cro and the Raku® Programming Language.

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

Thank you!