www.parrotcode.org | Last Release: 0.7.0 "Severe Macaw"
Set by moderator on 3 September 2008.
chromatic Work keeps me busy; have to keep the news hopper full. 00:01
00:02 fengc joined, particle1 joined 00:09 AndyA joined
cognominal Steve has shrinked chromatic : nano-chromatic 00:11
00:26 kid51 joined 00:28 xiaoyafeng_ joined
dalek r30945 | chromatic++ | trunk: 01:02
: [STM] Fixed a typo.
diff: www.parrotvm.org/svn/parrot/revision?rev=30945
01:22 xiaoyafeng_ joined
dalek r30946 | jkeenan++ | trunk: 01:23
: Applying additional patch supplied by Andy Dougherty in RT 57920. Implement use of SIGUSR1.
diff: www.parrotvm.org/svn/parrot/revision?rev=30946
wknight8111 I was going to make some comment like "if i had a dollar for every time I built parrot..." 01:27
01:27 xiaoyafeng__ joined
wknight8111 but then i realized that it takes so long to build parrot on my system that it's not really a tenable financial option 01:27
01:29 Ademan joined
kid51 wknight8111: Which slow system would that be (like my 5 y.o. iBook)? 01:34
01:38 bacek joined
wknight8111 I bought this IBM laptop used, back in undergrad 01:38
it's probably 5 or 6 years old
kid51, you said you had some perl work needed to be done?
I always like to practice my perl
01:41 Whiteknight joined
kid51 Whiteknight: You can plunge in along with everybody else; we had a great response. I just posted a follow-up to the list and created two new RTs: 58740 and 58742. Everybody into the pool! 01:47
Whiteknight adult swim!
purl well, adult swim is putting fake mustaches on everyone.
01:48 xiaoyafeng__ joined 01:49 dmknopp left 01:52 xiaoyafeng___ joined 01:53 particle joined 03:28 Andy joined
Andy anyone wanna help a cage cleaner get started? 03:29
Just got mail from someone I don't have tuits for. 03:30
Tene kid51, I think, was soliciting help from new cage cleaners. 03:50
There's some mail on the list about it.
Andy yeah, I know.
I forwarded the guy's email to the list. 03:51
04:12 tetragon joined 04:19 particle1 joined 04:55 Theory joined 06:07 masak joined 06:24 bacek joined
bacek g'day 06:24
alloha purl
moritz is speaking to bots a first sign of insanity? ;-) 06:25
goood localtime() ;-)
bacek moritz: ;)
06:26 uniejo joined
bacek any waiting bugs, that can be fixed by casual developer? 06:26
moritz bacek: yes, pmichaud reported a bug on Str.rindex 06:27
shouldn't be too hard, it doesn't work right now on non-Str's
cotto_home this one might be interesting: #42853: [CAGE] API Doc Signature Checker 06:28
bacek moritz: I already added patch to ticket. Few days ago 06:29
moritz bacek: I usually miss those unless they go to the list
bacek moritz: sorry, my fault. But "add parrot-porters@ to cc" not always displayed in RT. 06:30
cotto_home: good ticket. We probably can use languages/c99 for it 06:31
moritz bacek: did you choose "comment" or "reply" when adding the patch?
I think "comment" doesn't send any mails
bacek moritz: reply. 06:32
moritz bacek: weird.
RT has its subtleties, I think 06:33
bacek moritz: #58644 with patch for rindex
moritz bacek: I'm testing that right now
bacek moritz: I'm not sure that explicit casting to String is "good idea" 06:35
masak rakudo: .perl.say for {a => ["b", "c"]}.kv # I have a fun bug for you, too! :)
polyglotbot OUTPUT["a"ā¤"b c"ā¤]
moritz bacek: it's the usual way we du it in any-str.pir, I think
masak shouldn't the second line be ["b", "c"] ? :)
moritz bacek: or by assigning to a $S1 register 06:36
bacek moritz: ok.
moritz masak: any kind of hash dereferencation flattens lists over-eagerly
bacek masak: ouch... What this command means?
masak bacek: :)
moritz: known issue?
moritz masak: at least to me ;) 06:37
masak I didn't find any RT ticket about it
moritz masak: I think I didn't open one because I know that list assignment isn't done yet, and jonathan is working on the lazy list branch
masak: feel free to open one, though
masak asked ihrd to report this one 06:38
he'd the one who found it
moritz rakudo: my %a = (a => [[3], [4,5]]); say %a<a>.perl
polyglotbot OUTPUT[[3, 4, 5]ā¤]
NotFound I'm not sure that all C functions must have pod doc. On short static functions for internal usages, and with descriptive names of the functions and his parameters, it's just a vaste of spacetime.
moritz same thing, different disguise
masak aye. 06:39
NotFound In addition: We use headerizer to avoid the need to re-check the declarations, and now we have to re-check them for the pod?
dalek r30947 | moritz++ | trunk: 06:43
: [rakudo] make .rindex work on non-strings, closes RT #58644
: Patch courtesy by Vasily Chekalkin, bacek++
diff: www.parrotvm.org/svn/parrot/revision?rev=30947
06:44 Zaba joined
dalek r30948 | cotto++ | trunk: 06:44
: [dynpmc] make subproxy.pmc's test filename match its pmc filename for #44457
diff: www.parrotvm.org/svn/parrot/revision?rev=30948
bacek karma 06:45
purl?
purl yes, bacek?
bacek krama bacek
karma bacek
purl bacek has karma of 57
bacek oh... 06:46
moritz karma pmichaud 06:47
purl pmichaud has karma of 1604
cotto_home karma cotto 06:48
purl cotto has karma of 110
cotto_home w00r. triple digits
moritz karma cotto_work
purl cotto_work has karma of 2
moritz karma cotto_home
purl cotto_home has karma of 3
Tene karma tene
purl tene has karma of 175
moritz we need to sum them all up ;)
Tene karma everyone
purl everyone has karma of 13
cotto_home everyone++ 06:49
moritz karma Auzon
purl auzon has karma of 19
Tene karma moritz
purl moritz has karma of 199
bacek moritz++ # new anniversary!
karma moritz
purl moritz has karma of 200
bacek ;)
moritz w00t 06:50
cotto_home msg kid51 #44457 would be a good ticket for one of the recent Perl volunteers
purl Message for kid51 stored.
Tene karma dmknopp
purl dmknopp has karma of 7
Tene new committer coming up
karma pmichaud
purl pmichaud has karma of 1604
Tene karma jonathan
purl jonathan has karma of 804
cotto_home karma c 06:51
purl c has karma of 7129
Tene karma c++
purl c++ has karma of -81
cotto_home ((c++)--)++
bacek wow... It's bad...
Tene karma perl
purl perl has karma of 536
Tene karma perl6
purl perl6 has karma of 21
Tene karma perl 6
purl perl 6 has karma of 1
bacek (c++)++ # language of choice for many years ;)
cotto_home karma parrot 06:53
purl parrot has karma of 23
Tene parrot++ # keeping me entertained 06:54
bacek karma gc
purl gc has karma of -3
cotto_home blame Whiteknight
Tene karma malloc
purl malloc has neutral karma
bacek gc++ # for keeping whiteknight busy ;)
Tene karma free
purl free has karma of -1
cotto_home karma-- #for keeping Tene busy
bacek karma free beer
purl free beer has karma of 10
Tene karma karma 06:55
purl karma has karma of 59
bacek karma beer
purl beer has karma of 232
bacek wow... "free" should have karma ~ -200
Hinrik karma amrak sah
purl amrak sah has neutral karma
Hinrik karma amrak lartuen sah 06:56
purl amrak lartuen sah has neutral karma
Hinrik success
moritz rakudo: my %a = (a => [[3], [4,5]]); say %a.perl
polyglotbot OUTPUT[{"a" => [3, 4, 5]}ā¤]
Tene Ouch. 07:00
karma polyglotbot
purl polyglotbot has karma of 1
Tene polyglotbot++ # fun to write
masak Tene++ # for writing it
Tene karma tene 07:01
purl tene has karma of 176
Tene Hm, 176 = 2*2*2*2*11 07:02
That's a very nice number.
moritz Tene: I filed RT #58744 for this one (list flattening). If you have some spare time, this might be a good thing to investigate 07:04
Tene gimme rt link
moritz it makes a real mess of some of the tests
rt.perl.org/rt3/Ticket/Display.html?id=58744
Tene rakudo: say [[3], [4,5]].perl 07:07
polyglotbot OUTPUT[[[3], [4, 5]]ā¤]
Tene rakudo: say [[3], [4,5]].item.perl
polyglotbot OUTPUT[[[3], [4, 5]]ā¤]
bacek rakudo: say (a => [[3,4]]).perl 07:08
polyglotbot OUTPUT[("a" => [[3, 4]])ā¤]
bacek moritz: they didn't stored flat...
moritz rakudo: say (a => [[3], [4]]).perl
polyglotbot OUTPUT[("a" => [[3], [4]])ā¤]
bacek rakudo: say (a => [[3,4], [5,6]]).perl
polyglotbot OUTPUT[("a" => [[3, 4], [5, 6]])ā¤]
Tene So it's the assignment...
bacek rakudo: my $a= (a => [[3,4], [5,6]]); say $a.perl 07:09
polyglotbot OUTPUT[("a" => [[3, 4], [5, 6]])ā¤]
bacek Tene: no.
moritz bacek: yes. Assignment to hash containers
bacek rakudo: my %a= (a => [[3,4], [5,6]]); say %a.perl
polyglotbot OUTPUT[{"a" => [3, 4, 5, 6]}ā¤]
bacek moritz: yes. gotcha...
Tene rakudo: say [[3], [4,5]].hash.perl 07:10
polyglotbot OUTPUT[{"3" => [4, 5]}ā¤]
Tene rakudo: say (a => [[3], [4,5]]).hash.perl
polyglotbot OUTPUT[{"a" => [[3], [4, 5]]}ā¤]
bacek moritz: try comment out line 25 in src/calsses/Hash.pir
Tene No luck. Doesn't help. 07:12
rakudo: say {a => [[3], [4,5]]}.hash.perl 07:14
polyglotbot OUTPUT[{"a" => [[3], [4, 5]]}ā¤]
bacek rakudo: my %a= hash(a => [[3,4], [5,6]]); say %a.perl
polyglotbot OUTPUT[{"a" => [3, 4, 5, 6]}ā¤]
bacek rakudo: my $a= hash(a => [[3,4], [5,6]]); my %b=$a; say %b.perl 07:15
polyglotbot OUTPUT[{"a" => [3, 4, 5, 6]}ā¤]
bacek rakudo: my $a= hash(a => [[3,4], [5,6]]); my %b=$a; say %b.perl; say $a.perl
polyglotbot OUTPUT[{"a" => [3, 4, 5, 6]}ā¤{"a" => [[3, 4], [5, 6]]}ā¤]
bacek rakudo: my $a= hash(a => [[3,4], [5,6]]); my %b; %b<a>=$a; say %b.perl; say $a.perl 07:16
polyglotbot OUTPUT[{"a" => {"a" => [3, 4, 5, 6]}}ā¤{"a" => [[3, 4], [5, 6]]}ā¤]
Tene rakudo: my $a= hash(a => [[3,4], [5,6]]); say $a.hash.perl
bacek rakudo: my $a= hash(a => [[3,4], [5,6]]); my %b; %b<a>=$a<a>; say %b.perl;
polyglotbot OUTPUT[{"a" => [[3, 4], [5, 6]]}ā¤]
OUTPUT[{"a" => [3, 4, 5, 6]}ā¤]
bacek rakudo: my $a= hash(a => [[3,4], [5,6]]); my %b; %b<a>=$a<a>; say %b.perl; 07:17
polyglotbot OUTPUT[{"a" => [3, 4, 5, 6]}ā¤]
bacek Tene: can you add output of "requestor" name in polyglotbot?
Tene Sure.
Give me example output of what you want 07:18
bacek Tene: "bacek: OUTPUT[...]"
polyglotbot: my %a; say %a.WHAT
Tene Hm. Not so easy, perhaps. 07:27
I'll look again tomorrow.
Need to sleep now.
bacek Tene: g'night 07:28
dalek r30949 | allison++ | pdd27mmd: 07:31
: [pdd27mmd] Eliminate deprecated 'n_*' opcode in PGE test.
diff: www.parrotvm.org/svn/parrot/revision?rev=30949
07:40 iblechbot joined
dalek r30950 | allison++ | pdd27mmd: 07:42
: [pdd27mmd] Eliminating deprecated 'n_*' ops from PCT and PCT tests.
diff: www.parrotvm.org/svn/parrot/revision?rev=30950
07:43 apannu_ joined 07:50 Zaba_ joined
dalek r30951 | allison++ | pdd27mmd: 08:06
: [pdd27mmd] Remove no-longer needed underscores from multi name.
diff: www.parrotvm.org/svn/parrot/revision?rev=30951
08:18 Sargun joined
Sargun heya 08:18
is there a parrot backend for gas?
moritz what's gas?
purl rumour has it gas is taco bell's fault
Sargun gas = gnu assembler
moritz I haven't heard of one, but that doesn't mean much 08:19
but I don't think that parrot has similar ambitions as llvm
(in terms of being a low-level VM) 08:20
Sargun LLVM sucks 08:23
moritz I can build parrot with llvm-gcc, so it can't suck too much ;)
08:24 mberends joined
Sargun Well, my first pass of LLVM was getting random "crashes" 08:24
(seg faults and the like)
Tene greets sargun
Sargun hiya Tene
moritz immature != sucking, IMHO 08:25
Sargun moritz, its 4 years old.
might as well be GNU hurd
moritz age != maturity 08:26
Tene LHC just completed its first full loop, it looks like. 08:27
moritz OMG WE R ALL GOING 2 DYE 08:28
Sargun Tene, no collision.
Tene Sargun: I know.
Sargun: what brings you to #parrot?
moritz see today's xkcd ;)
Sargun hehe 08:29
08:29 Zaba joined
Sargun Tene, Looking for a virtual machine that allows: A) One to code in assembler, more intel or gas stylel assembler rather than anything too high level 08:29
B) runs on most platforms maturely. 08:30
its either gas in cygwin
pasm, or llvm
llvm's assembly is just...weird.
PASM is a bit too high level IMHO.
Tene What's this for?
Sargun My own personal enjoyment, and school. 08:31
Trying to convince my professor to get away from MASM
on MS Windows
Tene You probably want PIR, not PASM.
moritz that's even more high level ;)
Sargun PIR is even higher level than PASM
masak PIR++
Sargun I'd prefer if I could just take AT&T syntax and dump it to the bytecode. 08:33
So many VMs are just limited.
moritz that's because developer's time is limited :/ 08:34
dalek r30952 | allison++ | pdd27mmd:
: [pdd27mmd] Enable custom mark vtable function in CallSignature PMC.
diff: www.parrotvm.org/svn/parrot/revision?rev=30952
08:34 Ademan joined 08:48 Whiteknight joined 08:51 barney joined
dalek r30953 | chromatic++ | pdd27mmd: 08:52
: [PMC] Fixed in-place promotion during shifts from Integer to BigInt.
diff: www.parrotvm.org/svn/parrot/revision?rev=30953
r30954 | bernhard++ | trunk: 09:11
: #55196: [BUG] print/say opcodes have different float precision
: add a testcase
diff: www.parrotvm.org/svn/parrot/revision?rev=30954
09:58 Debolaz joined
dalek r30955 | allison++ | pdd27mmd: 10:01
: [pdd27mmd] Convert 'is_equal', 'is_equal_num', and 'is_equal_str' over to real
: vtable functions, that do multiple dispatch by default. Also, provide default
: 'is_equal*' multis for scalars.
diff: www.parrotvm.org/svn/parrot/revision?rev=30955
10:16 xiaoyafeng joined 10:18 xiaoyafeng_ joined 10:21 xiaoyafeng__ joined 10:27 xiaoyafeng__ joined 10:29 xiaoyafeng___ joined
dalek r30956 | allison++ | pdd27mmd: 10:38
: [pdd27mmd] Move debug prints for successful multiple dispatch after check for
: null multi.
diff: www.parrotvm.org/svn/parrot/revision?rev=30956
r30957 | allison++ | pdd27mmd: 10:41
: [pdd27mmd] Change direct multi call to vtable call, to allow for PMCs
: that override the default multi dispatched vtable function.
diff: www.parrotvm.org/svn/parrot/revision?rev=30957
moritz t/op/gc.t seems to loop in the pdd27mmd branch 10:51
dalek r30958 | allison++ | pdd27mmd: 10:52
: [pdd27mmd] Changing scalar PMC's 'cmp_num' and 'cmp_str' to multis so they don't
: mask child class multis.
diff: www.parrotvm.org/svn/parrot/revision?rev=30958
r30959 | allison++ | pdd27mmd: 10:55
: [pdd27mmd] Fix reversed logic in 'isne' opcode.
diff: www.parrotvm.org/svn/parrot/revision?rev=30959
10:57 Whiteknight joined 11:28 barney joined 11:36 bacek joined 11:42 xiaoyafeng___ joined 11:46 xiaoyafeng___ joined
dalek r30960 | allison++ | pdd27mmd: 11:47
: [pdd27mmd] Remove deprecated 'n_*' opcodes from Integer PMC tests.
diff: www.parrotvm.org/svn/parrot/revision?rev=30960
11:48 xiaoyafeng____ joined, bacek joined
dalek r30961 | allison++ | pdd27mmd: 11:54
: [pdd27mmd] Updating bad failing test: 20 + 30 should never equal 30, even if
: the two PMCs added are strings. The 'add' opcode must follow the numification
: behavior of the PMC.
diff: www.parrotvm.org/svn/parrot/revision?rev=30961
11:57 xiaoyafeng____ joined
dalek r30962 | allison++ | pdd27mmd: 11:58
: [pdd27mmd] Remove deprecated 'n_*' opcodes from String PMC tests.
diff: www.parrotvm.org/svn/parrot/revision?rev=30962
11:59 xiaoyafeng_____ joined 12:07 particle joined
dalek r30963 | allison++ | pdd27mmd: 12:08
: [pdd27mmd] Changing declarations of 'is_equal_string' to match the vtable
: function name in vtable.tbl.
diff: www.parrotvm.org/svn/parrot/revision?rev=30963
12:10 Zaba joined
dalek r30964 | allison++ | pdd27mmd: 12:32
: [pdd27mmd] Enable dump syntax for MultiSubs that does work in the new system.
diff: www.parrotvm.org/svn/parrot/revision?rev=30964
r30965 | allison++ | pdd27mmd: 12:33
: [pdd27mmd] Change test to look for new virtual namespace for multis.
diff: www.parrotvm.org/svn/parrot/revision?rev=30965
12:38 UltraDM joined 12:41 iblechbot joined 12:54 Andy joined
jonathan hi all 13:06
I've run into various issues today (surrounding obtaining Russian visa for going to Far East Perl - don't ask, the red tape is incredible). 13:07
So postponing Rakudo day until tomorrow.
13:09 particle joined
bacek jonathan: they just warns you ;) 13:09
13:19 xiaoyafeng_____ joined 13:22 xiaoyafeng______ joined
pmichaud jonathan: no problem, tomorrow works for me also (probably better for me too, given the way this morning has gone) 13:23
jonathan pmichaud: Cool. Think that I can almost stop running around madly soon. 13:25
Printer failure just after you ordered a courier and need to print off the docs you want couriering, is rather annoying. 13:26
Anyway, will do Rakudo stuff tomorrow, when I've got a day that I can fully concentrate on it.
13:29 xiaoyafeng______ joined 13:33 xiaoyafeng______ joined 13:35 xiaoyafeng_______ joined, gryphon joined 13:38 paco joined
mberends all: I would like to contribute a Rakudo based perl6doc for POD6 13:59
moritz cool 14:00
14:06 mberends joined
pmichaud indeed, very cool 14:08
particle started an effort in the src/utils directory -- I don't know how far along it is or where it ended up 14:09
particle it's a start, anyway 14:10
pmichaud perl6: my %a = (a => [1,2,3]); say %a<a>.WHAT; 14:11
polyglotbot OUTPUT[Arrayā¤]
pmichaud perl6: my %a = (a => [1,2,3]); say %a.perl;
polyglotbot OUTPUT[{"a" => [1, 2, 3]}ā¤]
pmichaud okay, it's not the hash itself that's flattening the list. 14:12
perl6: my %a = (a => [1,2,3]); say %a.kv.perl;
polyglotbot OUTPUT[["a", "1 2 3"]ā¤]
pmichaud I suspect it's the .kv method
perl6: my %a = (a => [1,2,3]); say %a.kv[1].perl;
polyglotbot OUTPUT["1 2 3"ā¤]
pmichaud that looks wrong to me. :-)
particle indeed
pmichaud oh yes, definitely. 14:13
$S1 = iter[$S1]
looks like it only gets strings out of the iterator :-|
(src/classes/Mapping.pir:107)
mberends particle's perl6doc.pir outputs pod6 verbatim and discards the ambient source code 14:16
particle mberends: like i said, it's a start
it parses some pod6, generates ast, and emits pod6 14:17
this is basically so the parser can be made more complete
then i figured i'd move on to proper emitters
you're very welcome to help. i suggest working on the parser first. 14:18
mberends my Rakudo implemention seems slower to compile/execute but easier for me to code. the parser is doing a decent subset and there are emitters for xhtml and man coming along. 14:22
pmichaud mberends: fantastic!
moritz mberends: do you have a repository URL that you want to share with us? 14:23
mberends not yet. i want to clean the source a bit (1-2 days) and then svn it up or pass it to you via nopaste 14:24
particle so, this is a parser for pod6 written in perl 6? 14:25
mberends yes
particle ah, i see. do you use a perl 6 grammar for parsing the pod? 14:26
dalek r30966 | pmichaud++ | trunk:
: [rakudo]: spectest-progress.csv update: 158 files, 3288 passing tests
diff: www.parrotvm.org/svn/parrot/revision?rev=30966
mberends no, it's line scanning event based, like perl5's Pod::Parser 14:28
dalek r30967 | pmichaud++ | trunk: 14:32
: [rakudo]: Fix .kv method on Mapping to not stringify values (RT #58744)
diff: www.parrotvm.org/svn/parrot/revision?rev=30967
14:33 xiaoyafeng_______ joined 14:35 rdice joined
moritz ah, that's the first half of the ticket 14:36
rakudo: say :a.WHAT 14:41
polyglotbot OUTPUT[Pairā¤]
moritz rakudo: say :a.value
polyglotbot OUTPUT[1ā¤]
14:41 xiaoyafeng_______ joined 14:43 xiaoyafeng________ joined 14:55 jq joined
dalek r30968 | moritz++ | trunk: 14:56
: [rakudo] add tests for pairs to spectest_regression
diff: www.parrotvm.org/svn/parrot/revision?rev=30968
r30969 | kjs++ | trunk:
: [pirc] update README file.
diff: www.parrotvm.org/svn/parrot/revision?rev=30969
moritz svn-- 14:57
dalek knows about my commit, and here locally it still hangs
particle hangs, or is throttled? 14:59
moritz hang
not even control+c interrupted it
NotFound moritz: same has happened to me a lot of times. 15:00
15:00 Andy joined
NotFound Usually, waiting some time it returns. When tcp timeout, I suppose. 15:00
moritz blames the svn server, and the client for being non-interruptable
15:01 xiaoyafeng________ joined 15:03 slavorgn joined, xiaoyafeng_________ joined 15:07 xiaoyafeng_________ joined 15:14 kj joined 15:29 xiaoyafeng_________ joined, rurban joined
rurban Another license question: May I copy C code from a BSD project? kaffe (jvm) is LGPL and GPL but the license owner went bankrupt and the new owner is ... well somewhere 15:33
pmichaud rurban: (a random thought) maybe it would be better to put this stuff in a separate repo? (more) 15:36
15:36 Theory joined
pmichaud you wouldn't have to worry about licensing issues, and since you're the one working on Parrot install issues it might be good for you to blaze-the-trail for how libraries-outside-of-the-repo integrate with Parrot 15:36
rurban With a seperate licence, you mean? A aspeerate license wil only raise the speed of the first initial prototype. Then after some months I assume the initially borrowed reader code will completely go away. 15:37
15:37 xiaoyafeng_________ joined
rurban I found now a BSD source of aJVM classfile reader. 15:37
pmichaud okay, it was just a thought. 15:38
as far as what's permitted and what isn't, I'm not the person to answer those :-)
rurban I understand Allison because she wants to keep the repo clean. 15:39
15:40 Zaba joined 15:55 xiaoyafeng_________ joined 16:01 davidfetter joined
rurban Just testing schwern's parallel test #57530 on cygwin with 2 cores: Lots of "Use of uninitialized value" but not bad. Just spec/S04-statements/for.rakudo 31 failed 16:01
moritz rurban: did you update trunk to r30967 or earlier? 16:02
s/trunk/your working copy/ 16:03
rurban r30943
moritz the warnings are normal
that explains the failure in for.rakudo
is it faster than serial test?
16:03 xiaoyafeng_________ joined
rurban ha! Warning: trailing whitespace in line 29 of languages/perl6/t/harness 16:03
moritz rurban: that can easily be fixed 16:05
rurban I'll update the ticket then with my results. A msvc test would be nice.
particle yes, in vim i type ",ts" and all trailing spaces in the current buffer are removed 16:06
" remove trailing whitespace
map ,ts :%s/\\s\\+$//<Enter>
imap ,ts <Esc>:%s/\\s\\+$//<Enter>a
moritz adds that to his ~/.vimrc 16:07
rurban In Xemacs it's simplier :)
16:07 peepsalot joined
rurban Shift-F5 <space>+$<enter><enter> 16:08
nopaste "particle" at 98.232.28.49 pasted "my .vimrc" (128 lines) at nopaste.snit.ch/14009
moritz I don't know if 5 key strokes are really easier than 3 ;)
rurban You had a macro which is fooling. With a macro I'm at one keystroke 16:09
moritz ok, without a macro it's :%s/\\s+$// 16:10
rurban This could have gone to chromatics hack book
moritz anyone on macosx or windows (not cygwin) here? 16:11
rurban I'm still getting the "Use of uninitialized value" with latest revision
16:12 xiaoyafeng_________ joined
moritz yes, that's expected 16:12
NotFound That remembers me a disucssion about if it was more efficient to set ctrl-alt-supr to halt or to reboot X-) 16:13
particle i'm on windows, but without a working compiler
moritz meta-x ctrl-x butterfly ;)
NotFound Counting keypresees, and making a statistic of wheter who may times each result was wanted. 16:14
rurban I can verify moritz' cygwin test: no noticable speed difference at all. even a bit slower 16:18
moritz probably because windows doesn't do forking 16:20
maybe we should just set the default number of parallel tests to 1 16:21
so cygwin users won't get harmed, and everybody else benefits 16:22
rurban 3 sounds good because it does no harm at all and may help on better platforms
the speed diff is in microsecs
moritz ok 16:24
so neglictible
rurban I'll test now my branch on all my virtual machines.
and MSVC6
Ha, I only got Test::Harness 3.12. I'll test with 3.13 now 16:26
That was IT! Two times faster on cygwin now with 3.13 16:28
137 secs against 78 now
moritz cool 16:29
dalek r30970 | allison++ | pdd27mmd: 16:33
: [pdd27mmd] Rework the Ref PMC tests because 'add' no longer modifies the
: destination by default.
diff: www.parrotvm.org/svn/parrot/revision?rev=30970
particle umm... moritz? 2jobs: 4m, 1job: 2m? 16:34
moritz particle: oh, I confused the two
particle *phew*
moritz particle: it's the other way round, of course
particle hasthelhcdestroyedtheearth.com/ 16:36
moritz wonders if there's any IRC channel on which he won't see that link today 16:37
probably #idle ;)
Infinoid [09:37] -!- Infinoid [infinoid@quack.glines.org] has joined #idle
[09:37] <@Infinoid> hasthelhcdestroyedtheearth.com/ 16:38
moritz lol
rurban BTW: Brookhaven's LHC wasn't allowed to start, because US politicians were not that scientific 16:39
moritz so it's good that the earth can be destroyed from europe as well ;-)
rurban I remember that almost all evil started here. 16:40
NotFound rurban: did you talk about Pascal?
16:41 AndyA joined
rurban I visited CERN some time ago but was not allowed to see it. 16:41
Haskell also
16:51 AndyA joined
rurban hmm, mingw is not stable. vanillaperl with a working gcc says: inter::progs - Determine what C compiler and linker to use...Compilation failed with 'gcc' 16:54
But MSVC works okay.
moritz you can pass the --verbose option to configure to see what's going on 16:55
(but I guess you know that already)
rurban I see. --cc=cl --link=link should be it 16:57
17:00 kjs_ joined
rurban Now it fails with mingw "ar" as lib. Bad. The librarian has no Configure option 17:02
17:03 Zaba joined 17:13 rurban joined 17:17 xiaoyafeng_________ joined
dalek r30971 | rurban++ | cygwin070patches: 17:19
: fix MANIFEST, win32 readme for msvc-gcc mix
diff: www.parrotvm.org/svn/parrot/revision?rev=30971
rurban Where are the MSVC dcs for the libparrot.lib importlib quirks?
dcs=>docs
dalek r30972 | allison++ | pdd27mmd: 17:20
: [pdd27mmd] Remove deprecated 'n_*' opcode from read-only PMC test.
diff: www.parrotvm.org/svn/parrot/revision?rev=30972
r30973 | allison++ | pdd27mmd: 17:22
: [pdd27mmd] Remove unnecessary '__' in multi name.
diff: www.parrotvm.org/svn/parrot/revision?rev=30973
moritz can sombody help me with Configure.pl? 17:23
I want it to stop passing the -Wl,-E option to the linker
(because llvm-ld doesn't understand it)
particle you probably want to adjust your hints file to deal with llvm 17:24
moritz oh wait, just solved my problem myself ;)
rurban MSVC does not like the explicit "msvcrt.lib" in C_LIBS. Witout it links fine. 17:25
17:25 xiaoyafeng_________ joined 17:32 xiaoyafeng_________ joined 17:40 xiaoyafeng_________ joined 17:46 peepsalot joined 17:48 xiaoyafeng_________ joined 17:52 AndyA joined 18:06 sjn joined
NotFound moritz: ping 18:16
moritz NotFound: pong 18:17
NotFound moritz: your message in the list about 'interp' undeclared identifier... Have you tried to add an #ifdef PARROT_INTERP to see what says? 18:18
moritz NotFound: I haven't because I have clue of the internals 18:20
NotFound Just try: #ifndef PARROT_INTERP \\n #error Doh! \\n #endif 18:21
18:22 peepsalot joined
NotFound Somewhere at the beginning of default.pmc 18:22
moritz NotFound: doesn't have any effect 18:25
NotFound Must be a macro expansion problem, then. 18:27
18:31 muixirt joined
NotFound The problem is clearly at CONST_STRING, not the interp thing... I think I got it: expanded CONT_STRING refers to his line number... but the line number of the PMC, and looks like llvmc does not honor the #line directives. 18:32
The .pmc, I mean. 18:33
moritz I'll check later 18:34
NotFound I'll put on the mailing list. 18:35
18:36 purl joined 18:43 purl joined 18:46 Theory joined 18:48 teknomunk joined 18:58 sjansen joined 18:59 purl joined
szbalint is there an alternative online repository browser that is not parrotvm.org? It's timing out all over the place for me and I wouldn't want to mirror the whole repo just to track down a change... 19:06
PerlJam szbalint: if you used git you'd already have a mirror of the whole repository :) 19:08
moritz szbalint: perlcabal.org/svn/parrot/revision/?rev=30945 but I guess it's the same server as parrotvm.org 19:09
szbalint thanks, the perlcabal link is much faster :) 19:10
pmichaud PerlJam: got back to crp just in time, eh? 19:12
PerlJam pm: yep. I don't even have a house yet, but I'm getting things ready at my sister-in-law's 19:13
in a few minutes I'll be trudging back out into the heat to anchor down two play-house thingies so that they don't blow away. 19:18
19:29 {-} joined 19:39 ewilhelm joined
moritz pmichaud: do you have any objections to testing rakudo in parallel by default? 19:51
pmichaud as long as spectest_regression continues to run and we don't break things on well-known platforms, no objection
moritz ok 19:52
ewilhelm moritz, are you currently working on applying that patch? 19:54
moritz ewilhelm: blocking on more test reports from other platforms
muixirt moritz, are there known issues with x86-64 w.r.t. rakudo? 19:58
ewilhelm moritz, would defaulting to jobs=1 speed-up the application?
just apply it, then test?
moritz ewilhelm: no, that's not a blocker
muixirt: I know of none 19:59
ewilhelm: appy it, test it, report to me.
muixirt moritz, just a sec
TimToady phone 20:00
NotFound Some objection to commit now the addition of -I and -L options to parrot command line? 20:03
ewilhelm moritz, you need tests on amd64 linux?
moritz ewilhelm: not really, parrot+perl should abstract that away 20:04
20:05 obra joined
ewilhelm moritz, so what reports do you need? 20:06
muixirt moritz, t/spec/S16-io/basic-open definitly hangs on my x86-64 linux machine
NotFound Objectionability timed out
moritz ewilhelm: windows/msvc, macos,*BSD
dalek r30974 | julianalbo++ | trunk: 20:07
: add -I and -L options to parrot command line
diff: www.parrotvm.org/svn/parrot/revision?rev=30974
moritz muixirt: please file a bug report if you haven't already
muixirt moritz, i was afraid you would say this ... 20:08
moritz muixirt: and to enable the rest of the testing, temporarily remove that test from t/spectest_regression.data 20:09
20:11 Theory joined
cotto_work seen kid51 20:12
purl kid51 was last seen on #parrot 18 hours, 24 minutes and 59 seconds ago, saying: Whiteknight: You can plunge in along with everybody else; we had a great response. I just posted a follow-up to the list and created two new RTs: 58740 and 58742. Everybody into the pool!
NotFound Now you can do: ./parrot -I examples/nci -L examples/nci compilers/nqp/nqp.pbc examples/nci/xlibtest.nqp 20:13
Well, the .nqp does not use the include, but it does not harm.
20:23 ewilhelm left
dalek r30975 | julianalbo++ | trunk: 20:30
: fix segfault in example Xlib when failed to open display
diff: www.parrotvm.org/svn/parrot/revision?rev=30975
20:32 AndyA joined
dalek r30976 | allison++ | pdd27mmd: 20:37
: [pdd27mmd] Remove deprecated 'n_*' opcodes from BigInt tests.
diff: www.parrotvm.org/svn/parrot/revision?rev=30976
moritz NotFound: did I get this right, that parrot build depens on the C compiler reporting the correct line number? 20:45
particle yes, it does 20:46
NotFound moritz: the CONST_STRING machinery depends on that, right.
moritz funny thing to do ;) 20:47
NotFound We have the Large Line Number Collider.
moritz lol 20:48
20:57 muixirt2 joined 21:04 Andy joined 21:08 {-} joined 21:22 Theory_ joined
cotto_work {-}++ 21:29
NotFound cotto_work: Is that a perl6 operator? 21:31
moritz it's the "plonk" operator ;) 21:32
jonathan It's a plausible looking sequence of non-alphanumerics, so probably. ;)
moritz jonathan: may I ask you to test the latest parallel-testing-patch on windows? 21:34
cotto_work NotFound, no idea. I just noticed that someone joined with that name. 21:35
moritz hey, you can even use \\ as a nick
cotto_work Although I guess it could be if you wanted it to. ;) 21:36
NotFound I know a guy nicknamed \\dev\\null 21:37
jonathan moritz: Bit tired now, will do tomorrow at start of Rakudo day.
Badger me if I forget.
moritz jonathan: ok, no problem
rurban NotFound: Your r30975 fix to "fix segfault in example Xlib when failed to open display" does not work for me. 21:48
NotFound rurban: Maybe you forget to rebuild the pbc?
cotto_work NotFound, why does the code load a pbc instead of the pir? 21:49
21:49 s1n joined
rurban There's no pbc 21:49
NotFound cotto_work: why not? 21:50
rurban DisplayName: "" \\n Hello \\n segfault
cotto_work because you have to compile the pir
rurban I should say that the X server is also not running. 21:51
NotFound cotto_work: good point.
rurban: can you -t it? 21:52
rurban ../../perl6 xlibtest.p6 works correct: Cannot open display + trace 21:53
dalek r30977 | allison++ | pdd27mmd:
: [pdd27mmd] Fix calls to overflow in Integer PMC, so redispatch to MMD can use
: full original signature.
diff: www.parrotvm.org/svn/parrot/revision?rev=30977
nopaste "rurban" at 212.183.51.65 pasted "cygwin segfault with empty DISPLAY" (39 lines) at nopaste.snit.ch/14011 21:55
"rurban" at 212.183.51.65 pasted "ok with p6 with empty DISPLAY" (12 lines) at nopaste.snit.ch/14012 21:56
21:58 rurban joined
cotto_work when's the mmd merge planned for? 22:01
NotFound rurban: that trace lacks the part where the display is opened. 22:02
moritz cotto_work: I don't think before the release on tuesday (but I don't really know)
cotto_work I wouldn't expect it before Tuesday, but it sounded like it was getting pretty close. 22:03
nopaste "rurban" at 212.183.51.65 pasted "-t with empty DISPLAY (first part)" (69 lines) at nopaste.snit.ch/14013
NotFound rurban: some more opcodes 22:07
rurban I know why 22:09
It must be: load_bytecode 'Xlib' not .pbc (extension automatically) 22:10
so it found an old Xlib.pbc in runtime library
now it works correct
NotFound Ah, good. 22:11
nopaste "rurban" at 212.183.51.65 pasted "fix examples/nci/xlibtest.pir" (14 lines) at nopaste.snit.ch/14014
rurban I'll also sleep now, bb 22:12
moritz good idea, good night
cotto_work looks like at least 38 failing tests, so it might be a little while
'night 22:13
moritz (also rakudo build segfaults)
NotFound I tested today, and it blocks in t/ops/gc.t
cotto_work NotFound, it doesn't block but it does take a long time.
NotFound cotto_work: long that 5 minutes? 22:14
Will retry. 22:15
cotto_work I don't know exactly how long. I just noticed that it was taking a long time when I checked if the test suite was done.
NotFound My machine is not very fast, BTW
nopaste "cotto_work" at 131.107.0.101 pasted "test results for r30977 in mmd branch" (39 lines) at nopaste.snit.ch/14015 22:16
cotto_work That's x86/Debian. 22:17
NotFound 1 TODO passed. 22:18
cotto_work: at first look many of them are related to bigint 22:30
cotto_work I was just checking out of idle curiousity. I'm sure chromatic and allison are very aware of which tests are failing and how to fix them. 22:33
(and anyone else hacking that branch) 22:34
22:34 kid51 joined
NotFound An additional eye always can be helpful 22:34
szbalint good $localtime
cotto_work very true 22:35
wow. It's $localtime already! 22:37
szbalint kid51: I was thinking about taking RT #46783 , but when I looked at smartlinks.t it turned out the TODO-ed tests do not actually create files temporary or otherwise. I'm just a little confused now, is there some policy to create special test files to _read_ from, instead of using 'docs/pdds/pdd03_calling_conventions.pod' for example, or the todo is a false alarm? 22:39
kid51 szbalint: Good to hear from you. 22:41
I actually know little about those RTs. They were created en masse about a year ago by another 'cage-cleaner'. 22:42
And he no longer has the same time available for the project.
I stumbled upon them just a few nights ago.
So I'm as new to them as you are.
I suspect we'll have to look at the whole bunch of them and see where to begin. 22:43
szbalint ah I see
NotFound I killed several cage tickets of that sort when discovered that were related to unused and already deleted code. 22:45
So no wonder if some of them are outdated.
They are useful anyway, if they help to detect unused code. 22:46
kid51 szbalint: cotto_home suggests RT 44457 might be good to look at
22:47 josh joined
cotto_work yup. That should be a fairly easy one to tackle for anyone who knows Perl. 22:47
josh hi ya'll
kid51 Hi, Josh. BTW, we have many Josh-es here. Which one are you? 22:48
szbalint someone's already tackling the problem apparently
at least according to RT
josh i'm very new to parrot - i'm playing around with it, and i wanted to try running a test using the c99 compiler, e.g. ./parrot languages/c99/c99.pir examples/c/test_main.c
uh....i'm the *unkown* josh :) 22:49
kid51 szbalint: Don't let that stop you from taking a look at it. The person who posted today is just as much a noob as you are -- or as I would be if I looked into that ticket.
cotto_work szbalint, are you referring to 44457?
szbalint cotto_work: yeah
cotto_work go ahead. That's me.
nm
I thought you were referring to the one who took it, not Igor.
I'm definitely not him. 22:50
josh ....i'm getting: error:imcc:No such file or directory in file 'languages/c99/c99.pir' line 56 - which when i check c99.pir it looks like the files referenced there exist
kid51 szbalint: Most of the Perl 5-ish tickets we either have now or are likely to have in next several months are likely to require some study of obscure corners of the distro. 22:51
We pick the low hanging fruit quite quickly.
I will be looking to generate new RTs for people to work on, but don't wait on me! 22:52
NotFound josh: are you executing it from the languages/c99/ directory?
josh notfound: i tried that also with the same result
my obvious conclusion is that i don't understand the basics of running parrot 22:53
perhaps i'm looking in the wrong places, however i was unable to find any helpful documentation on this
hence, looking for pity-help on irc
kid51 josh: So did you run: 'perl Configure.pl; make; make test' ?
josh kid51: yes, otherwise it's kinda hard to run a ./parrot 22:54
szbalint It's very useful to have a list of cage cleaning tasks as it provides a less steep learning curve than just trying to hack on parrot internals right away. A good way to get involved.
NotFound josh: given that c99 has no README and no doc, is hardly your fault. 22:55
josh notfound: i see 22:57
NotFound josh: for me it says "Failed to parse source" 22:58
$ ../../parrot c99.pir ../../examples/c/test_main.c
Failed to parse source
josh notfound: $ ../../parrot c99.pir ../../examples/c/test_main.c still gives me the "no such file or directory" issue - i'm using the 0.7.0 build fwiw 23:00
NotFound josh: did you execute make on languages/c99 ?
josh aha! no, i didn't
now i get the "Failed to parse source" error - fun 23:01
NotFound After building parrot, you must build each language you want to test.
Well, at least now it not your fault, is c99's fault ;)
kid51 szbalint: True, but for most of this year there has been exactly one cage-cleaner: me. So the list existed in my head. 23:02
Or, at least the Perl 5 part of that list was in my head.
Here is a search query you can use at rt.perl.org to locate [CAGE] RTs:
NotFound There is a link at the "Usefull RT queries" pages. 23:03
kid51 tinyurl.com/4wwnct 23:04
And here's one for [TODO] items: tinyurl.com/5efpbr 23:05
And here's one for [BUG] RTs: tinyurl.com/59dmc4 23:07
And here's one for [PATCH] RTs: tinyurl.com/5jfj43
szbalint nice 23:08
kid51 Bookmark 'em ... and life will be sweet!
kid51 focuses on doughera's latest patch. 23:09
23:29 AndyA joined
szbalint kid51: anyhow, I've submitted a patch for #46783 - forgot to cc the list though so I'm telling you now :) 23:31
23:31 Limbic_Region joined
kid51 szbalint: Thanks. I'll take a look at it, hopefully tonight. 23:34
23:43 bacek joined
dalek r30978 | jkeenan++ | trunk: 23:53
: Applying 3rd patch submitted by Andy Dougherty in RT 57920. Logic of how step conducts probe for AIO is improved, but no better results on Darwin.
diff: www.parrotvm.org/svn/parrot/revision?rev=30978
23:59 Zaba_ joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!