Parrot 0.9.1 Released | parrot.org/ | 451 RTs left!
Set by moderator on 24 February 2009.
wayland76 Patch on ticket trac.parrot.org/parrot/ticket/426 should hopefully eliminate some duplicate code :) 00:08
00:09 AndyA joined 00:19 tetragon joined 00:28 japhb joined 00:29 ron joined 00:49 eternaleye joined
dalek rrot: r37251 | allison++ | trunk/examples/streams (11 files):
[cage] Updating titles of Stream examples for better navigation.
 02:01
rrot: r37252 | allison++ | trunk/docs/parrotbyte.pod:
[cage] Removing extra sections at the end, according to the Parrot
 02:05
purl i heard coding standards was that a joke? :P
wayland76 No, it's a standard code for talking like a Parrot 02:20
Hey, I've had an idea. Lets declare an "International 'Talk like a Parrot' day" :)
dalek rrot: r37253 | allison++ | trunk/docs/pdds/pdd07_codingstd.pod:
[doc] Updating coding standards on Pod documentation to match current
 02:28
02:43 Woody4286 joined
dalek rrot: r37254 | allison++ | trunk/t/native_pbc (4 files):
[cage] Skipping all native PBC tests, until a more robust testing
 02:59
rrot: r37255 | allison++ | trunk/docs/project/release_manager_guide.pod:
[cage] Removing the native PBC step from the release manager's guide,
 03:03
rg uhm ... coke? cotto? anyone? since cores are now selected with --runcore, I've overlooked that -r tests will match that aswell. as a result, any test that is skipped for -r is also skipped for every core. should i prepare a patch to match /-r\\b/ or should that rather be decoded as --run-pbc and matched as such? 03:07
hmm i'm reopening TT#403 and call it a night. 03:11
dalek rrot: r37256 | Util++ | trunk/lib/Parrot/Docs/Section (2 files):
[cage] Added standard SVN properties
 03:12
03:31 Tene_ joined 03:42 janus joined 03:44 tetragon joined
dalek rrot: r37257 | Util++ | trunk (13 files):
[codingstd] Fixed copyrights that were only off by single characters.
 04:22
rrot: r37258 | Util++ | trunk/lib/Parrot/Docs/Item.pm:
[codingstd] Uncuddled an else.
 04:30
rrot: r37259 | Util++ | trunk/docs/submissions.pod:
[codingstd] Re-sequenced =item numbers to pass pod_syntax.t
 05:31
06:41 tetragon joined 06:58 TiMBuS joined 07:12 uniejo joined 07:17 Psyche^ joined 07:34 wayland76 joined
dalek rrot: r37260 | jquelin++ | trunk/examples/pir/befunge:
moved befunge language to pir example to clean up languages shipped with
 07:41
rrot: r37261 | jquelin++ | trunk/examples/pir/befunge (5 files):
befunge no more part of languages/ - update build & infrastructure stuff
 08:03
08:06 korshak joined
dalek rrot: r37262 | jquelin++ | trunk/config/gen/makefiles/languages.in:
befunge moved outside of languages/
 08:07
08:10 korshak left, korshak joined
dalek rrot: r37263 | jquelin++ | trunk/config/gen/makefiles/languages.in:
befunge moved outside of languages/
 08:11
08:11 clunker3 joined 08:13 ujwalic joined
dalek rrot: r37264 | jquelin++ | trunk (4 files):
yet other befunge references squashed
 08:15
rrot: r37265 | jquelin++ | trunk (4 files):
removing last remnant bits of befunge language
 08:23
tracwiki: v57 | wiml++ | WikiStart 08:34
tracwiki: broken smoke.parrotcode.org link
tracwiki: trac.parrot.org/parrot/wiki/WikiSt...ction=diff
shorten dalek's url is at xrl.us/beizzq
08:45 masak joined 09:15 mikehh joined 09:22 korshak left, korshak joined 09:33 korshak left, korshak joined
dalek rrot: r37266 | cotto++ | trunk/src/pmc/fixedpmcarray.pmc:
[PMC] Remove special case code from FixedPMCArray's set_pmc.

and encapsulation trump optimization.		 09:36
09:44 clunker3 joined
dalek rrot: r37267 | cotto++ | trunk/src/pmc (2 files):
[PMC] switch Fixed- and ResizablePMCArray to ATTRs and remove a long-neglected Pod section
 10:03
wayland76 @karma bacek
moritz purl: karma bacek 10:04
purl bacek has karma of 97
10:05 krunen joined
bacek oh... too close. 10:06
cotto to windows 98? 10:07
bacek good evening, btw
cotto: no :)
cotto: too close too 100.
wayland76 Office 97! Oh noes :) 10:13
bacek wayland76: :) 10:14
wayland76: No. I just promised myself that I will not try to apply for commitbit in Parrot/Rakudo before I'll have 100 points of karma 10:16
cotto rakudo: say <++kecab>.reverse 10:17
polyglotbot OUTPUT[bacek++ā¤]
cotto (it still seems like a chicken/egg problem to me, though) 10:18
moritz
.oO( kebab? /me hungry... )
bacek :) 10:25
wayland76 bacek: I'd suggest basing it on successful patches instead :) 10:26
purl: karma c/c 10:27
purl c/c has no karma. It's a wee bit tricky to explain, but you'll understand in time, young Perl Jedi.
bacek wayland76: I hope at lease 2/3 of my karma is from dalek
10:27 korshak left
bacek at least... 10:27
ujwalic what are the rules for to use invoke in PASM 10:28
error:imcc:The opcode 'invoke' (invoke<0>) was not found. 10:29
moritz in pir it's just ()
don't know if that also applies for PASM
ie $P0() calls invoke on $P0 10:30
ujwalic examples\\nci\\QtHelloWorld.pasm in examples is not working for me
so i created QtHelloWorld.pir its working fine 10:31
I am on windows XP
moritz submit patch :-) 10:33
ujwalic ok 10:34
moritz: need to rename PQt.C to PQt.cpp 10:47
dalek rrot: r37268 | cotto++ | trunk/src/dynpmc/gdbmhash.pmc:
[PMC] switch GDBMHash to ATTRs
 11:21
12:04 skv joined 12:33 rg1 joined 13:06 gryphon joined 13:11 gryphon joined
ujwalic moritz: trac.parrot.org/parrot/ticket/427 13:17
moritz ujwalic: thanks 13:18
ujwalic :)
used invokecc, set_args .... in PASM 13:19
Coke Infinoid: ping 13:39
msg Infinoid www.parrot.org/smoke should probably 307 to the smolder address. 13:41
purl Message for infinoid stored.
Infinoid Coke: can do.
I've been mostly working on adding bounces and aliases for things that parrotcode.org had; is this one of those?
Coke smoke.parrotcode.org/smoke 13:42
purl smoke.parrotcode.org/smoke is as close as you'll get.
Coke but since smoke is dead...
Infinoid ah, I didn't know there was a smoke.
(I wish I had a master list of what parrotcode does, I've been working off of a list I scraped from google's cache)
should that bounce to the smolder project address, or the smolder list of recent reports? 13:43
we already had one for /smolder, I've added a similar one for /smoke. 13:47
13:58 mikehh joined
Coke Infinoid: you can checkout the svn repo that houses parrotcode 13:59
Infinoid that would be awesome. URL? 14:01
Coke svn.perl.org/perl.org/docs/live/parrotcode
Infinoid thanks
Coke (took me that long to find my checkout)
Infinoid Coke++
Coke be sure to check the .htredirects file.
Infinoid good stuff, that'll help a lot 14:02
Coke And I can commit to that repo, so feel free to send patches to me or webmaster@parrotcode.org 14:03
Infinoid I suppose that all depends on how incremental we want this changeover to be. 14:06
I sorta figured I'd get parrot.org handling all of its links, and then we'd just change DNS over one day
but I'm flexible.
Coke fair enough. 14:07
easier to just shut it down than to update it as we go.
just saying.
Infinoid cool. I've always wondered who to ask about the content over there :)
14:10 particle1 joined 14:12 AndyA joined 14:29 kj joined
Coke allison, chromatic and I all have commit bits, as do ask & robrt. (but they are admins, not content folks) 14:58
15:04 ron joined
Coke tries to check out parrot via an http proxy and is getting a TON of 15:07
svn: REPORT of '/parrot/!svn/vcc/default': Could not read response body: SSL err
or: decryption failed or bad record mac (svn.parrot.org)
updates a chunk of files, dies. 'svn up .' does more. at this rate I'll have a full checkout in 2 days. 15:08
Infinoid uh. there's gotta be a better way 15:09
Coke yah, an rsync mirror would be awesome. =-)
NotFound I've done it several times with squid without problems. 15:10
Infinoid Well, I've already got a git mirror, and an rsync server, so that shouldn't be too tough.
Last time I tried squid, it needed some additional HTTP command codes configured to get svn to work 15:11
Coke no doubt my proxy sucks. 15:12
it's HTTP::Proxy
Infinoid in my squid.conf: extension_methods REPORT MERGE MKACTIVITY CHECKOUT
Coke great for simple stuff.
NotFound Sucks or socks?
Coke sucks.
Infinoid sucks is still better than socks
15:13 diakopter joined
Coke yay, finally finished a checkout I started yesterday. =-) 15:13
(it's only taking a while because I get distracted by actual work.)
diakopter !seen chromatic 15:15
clunker3 chromatic was last seen on #parrot 5 months, 5 days, 6 hours, 38 minutes and 52 seconds ago, saying: "is good"? Me fail English!
Fayland_logger Saw chromatic Quit: Quit: Leaving on Tue Mar 3 17:04:40 2009
Coke on the bright side, I finally have a platform where Devel::NYTProf works. 15:16
(strawberry perl. go fig.)
Infinoid Coke: rsync://squawk.glines.org/parrot-trunk 15:17
will be auto updated as soon as I can remember the crontab syntax for "once every 10 minutes" 15:18
Coke Infinoid: a day late!
moritz Infinoid: */10 * * *
Coke (but thanks)
Infinoid eh.
ron Is there a way to get a sub pmc for a pmc method declared in a pmclass without looking for the method name in a namespace? 15:21
Coke for a generic PMC or an object PMC? 15:22
ron generic PMC
Coke looks like there's a "find_method" vtable entry.
15:23 diakopter joined
Coke which is exposed, unsurprisingly, by the find_method opcode. 15:23
Infinoid surprise-- 15:24
NotFound That's boring. We need a program that generates random names for opcodes ;)
ron Should that work for pmclass METHOD declarations as well as VTABLE declarations - I tried it for a declared METHOD and didn't seem to have luck with it.
Coke Given the descriptive text, yes, I'd expect it to work.
given that it works via the PMCs own vtable, it's possible the PMC override find_method. 15:25
... or that the version on default.pmc sucks.
15:26 diakopter joined
Coke but yes, my expections don't match reality. 15:26
nopaste "coke" at 65.91.151.195 pasted "I'd expect this to work, but it's not found" (7 lines) at nopaste.snit.ch/15848 15:27
Coke ah, nytprof fails on strawberry also when I run the same code I run on macosx. wtf. 15:29
15:31 Tene joined, Theory joined
NotFound =item C<PMC * Parrot_find_method_with_cache> 15:37
Find a method PMC for a named method, given the class PMC,
"the class PMC"
ron I didn't find a tt or rt yet. Should I double check and file a tt for the generic pmc case if nothing else turns up? 15:42
Coke ah... 15:49
pmichaud file a tt
NotFound "Looks up method $3 in $2's vtable" This description in the opcode is confusing 15:50
nopaste "coke" at 65.91.151.195 pasted "also fails." (7 lines) at nopaste.snit.ch/15849 15:51
15:57 rdice joined 15:59 ron left
dalek kudo: 531ca2c | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 317 files, 7015 passing, 140 failing

   S32-num/complex.rakudo aborted 140 test(s)		 16:33
shorten dalek's url is at xrl.us/bei24w
16:35 Whiteknight joined, riffraff joined
dalek kudo: d8f59ef | (Moritz Lenz)++ | src/classes/Complex.pir:
implement .re and .im in Complex
 16:38
shorten dalek's url is at xrl.us/bei25p
16:49 wayland76 joined 17:02 korshak joined 17:04 davidfetter joined
pmichaud #parrotsketch in 86 17:04
Infinoid joins and lurks now as he'll probably forget. 17:07
davidfetter HAI UTIL 17:30
pmichaud #parrot sketch in 60 17:31
Whiteknight ha, daylight savings time strikes again!
davidfetter heh 17:32
NotFound # Failed (TODO) test 'read on null PMC throws exception' # Expected: # /some crazy exception/ What? 17:41
And of course, no ticket. 17:42
17:43 korshak left
dalek rrot: r37269 | Util++ | trunk/config/gen/makefiles/root.in:
Change make fulltest to unconditionally run all its tests (TT#432)
 17:44
purl dalek: that doesn't look right
17:46 PacoLinux joined
Coke Util-- 17:49
Util Issue, Coke? 17:50
Coke yup. adding to the ticket. 17:51
17:51 particle1 joined
Util Shall I revert for now? 17:51
Coke In general, if you're changing something big, you might want to wait more than 11m for feedback. =-)
feedback shortly.
dalek rrot: r37270 | NotFound++ | trunk/t/pmc/io.t:
[cage] set ticket for TODO in t/pmc/io.t
 17:52
17:53 alvar_ joined 17:54 alvar_ joined
Util Hi, davidfetter! 17:55
FYI, your report of RPM problems led me to a hole in the release procedures; I am writing up the ticket now. (Well, *now* I am dealing with my first bad Whoops)
This problem caused the parrot.spec file to have DOS line-endings.
Coke Util: comment added (so yes, If it were up to me, I'd revert that.)
Util It is obvious to you that the old codepath is needed. 17:58
It is obvious to me that the new codepath is needed.
(More discussion on *why*, I will add to the ticket).
I will revert the change to target "fulltest", and re-apply it as a new target.
Sound good for now?
Coke I don't disagree that we need to get all the tests passing. 17:59
(new target) yes, that's fine, of course.
danke. 18:01
util++
Util Allison (and at least one Release Manager before her) have issued pre-release calls for people to `make fulltest`, the results of which are *rarely* complete. It was just this use-case I was trying to address. I can see your point of wanting a target that either passes everything, or shows failure at the end. 18:02
Coke (rarely complete) those tests need fixing, stat. =-) 18:03
if we're not fixing the tests that we know are failing now, running more tests that are failing isn't going to help us, I think. 18:04
(but I can see how having someone motivated to run the tests and get all the failures at once would be helpful) 18:05
dalek rrot: r37271 | Util++ | trunk/config/gen/makefiles/root.in:
Change make fulltest - revert r37269 (TT#432)
purl dalek: that doesn't look right
Util Reverted. Suggestions for new target name? `alltest`? `prerelease_tests`? 18:06
Coke purl, ignore dalek.
purl, owner?
purl?
purl, you in?
huh.
purl Coke: huh?
owner is Masque
Coke?
i haven't a clue, coke
Util purl is on downers 18:07
18:08 barney joined
Coke Util: fulltest_all 18:10
Util I like it. 18:11
Coke There should be a ticket to convert both of those targets into a single 'fulltest' that runs under a single harness to address both of our concerns. 18:18
dalek rrot: r37272 | Util++ | trunk/config/gen/makefiles/root.in:
Added target make fulltest_all (TT#432)
 18:22
Coke kicks off a fulltest run to celebrate. 18:23
Infinoid will miss #ps due to meetings, but didn't have much to report anyway. 18:24
At the rate I'm currently going at, the website roadmap item probably won't be done before 1.0 18:25
dalek rrot: r37273 | allison++ | trunk/src/pmc/integer.pmc:
[cage] Removing TODO that's already been done (generated MMD header

TODO was done. Resolves RT #46619.		 18:27
18:30 Andy joined
Coke clock? 18:30
purl Coke: LAX: Tue 11:30am PDT / CHI: Tue 1:30pm CDT / NYC: Tue 2:30pm EDT / LON: Tue 6:30pm GMT / BER: Tue 7:30pm CET / IND: Wed 12:00am IST / TOK: Wed 3:30am JST / SYD: Wed 5:30am EST /
18:31 Psyche^ joined 18:33 chromatic joined
allison Util: you'll have to put a skip for the main license file, as the Artistic License is owned by the Perl Foundation 18:39
Util Will do!
chromatic Hm, we already have docs/project/ticket_wrangler.pod 18:50
cotto Whiteknight, having help with the ATTR conversion would be nice.
allison chromatic: which shows it was a terrible name 18:51
cotto Could you take the ManagedStruct family?
allison chromatic: anything worth merging in from the old doc?
chromatic I'll review. Do you think a rename and an update is appropriate? 18:52
Coke Might be worth doing an svn rename to the new name and then replacing it with your version.
chromatic: so, yes.
cotto I can nopaste the places where its macros are hit during make test.
chromatic That was my thought as well. I'll do that.
Whiteknight I can do ManagedStruct, yes 18:53
nopaste would be good, email would be better for the list
nopaste "cotto" at 96.26.202.243 pasted "ManagedStruct UnionVal macro usages" (52 lines) at nopaste.snit.ch/15851 18:54
Whiteknight thanks cotto, I'll get to work on that asap 18:55
cotto See the wiki page for any gotchas (trac.parrot.org/parrot/wiki/PMCUni...nTasklist) 18:56
shorten cotto's url is at xrl.us/bei3qx
cotto Whiteknight++
Whiteknight don't ++ me, I've been too lazy this week
cotto It's only Tuesday. 18:57
19:00 barney joined, skv joined
Tene pmichaud: if you want to create a cardinal account, that would be great. 19:01
pmichaud okay, I'll start there.
19:05 baest joined
barney Some git instructions are also in trac.parrot.org/parrot/wiki/LeaveTheNest 19:06
chromatic Q: how in the world do I move Pheme, or at least complain until someone does it for me, preserving its history?
Tene I've got the start of a PCT-based scheme compiler. 19:07
I'll post it to github probably tonight.
chromatic Note that Pheme follows "The Little Schemer"'s Scheme more than anything else at the moment.
Definitely not traditional R5RS. 19:08
barney I need to check up on eclectus on occasion 19:09
Coke sings kumbayah. 19:10
barney www.myvideo.de/watch/484939/Kumba_Ya 19:11
Infinoid NotFound: the C api or the PMC api? 19:12
NotFound Infinoid: C api for anow
Don't checked if the pmc does the same.
Infinoid Ok. Improving the C API is on my pdd13 list but *way* far in the future 19:13
I eventually want to move most things over to using the pmc api, and then move most of the guts into the pmcs directly
Util envisions Yippies surrounding the SVN repo server, willing it to rise into the air.
NotFound Is just a minor problem: the function does not look for the code position specfified, but for the previous.
Infinoid But I've got a few major milestones to go before that
hmm. that sounds a bit weird 19:14
NotFound Yes, but is what all callers expect.
Except the ones I introduced for trace and backtrace. I added 1 before the call to account for that. 19:15
Infinoid jonathan is the one who knows everything about this code (at least, I hope he does, because I sure don't)
NotFound The main problem is that the exception object does not contain the throwing point, but the continuation to resume. 19:16
And thus the look for the previous position makes sense. 19:17
Infinoid could catch blocks break that assumption entirely? (I'm not very familiar with this stuff)
pmichaud I thought the point was that an exception could get the annotation from its caller chain
NotFound And the annotations opcode pass the next position, I suppose that is done that way for consistence.
pmichaud Tene: do you have an email address I can use for the cardinal account creation? 19:19
Tene pmichaud: tene@allalone.org
NotFound pmichaud: but the functions called by the throw opcode does not pass the current position, the resume continuation points to the next opcode. 19:20
Infinoid NotFound: is the continuation point *always* the throwing point plus one?
if so, maybe a macro would help. if not, maybe there should be separate fields for each
NotFound Infinoid: that depends of the size of the opcode.
Infinoid well, plus sizeof(opcode_t) 19:21
I'm just wondering if the continuation could point to a completely different function than the thrower
NotFound I tryied a patch modifying the throw_from_op functions. Is doable, but I'm not sure is the desired way. 19:22
Infinoid like I said, I don't know much about this stuff.
19:22 Whiteknight joined
NotFound Othe problem is that some opcodes pass a NULL continuation. 19:22
19:22 Patterner joined
Infinoid for fatal exceptions, I'm guessing? 19:23
maybe you do want a separate field or method, then
NotFound No, the continuation is builded inside the opcode, there is no way to pass a differente location.
purl okay, NotFound.
Infinoid shiny 19:24
19:24 contingencyplan joined
pmichaud NotFound: my point is that when we throw an exception, we invoke a new context. 19:24
so we know the position of the throw because that's the caller of the exception handler. 19:25
NotFound And for exceptions throw from C there is the general problem that not all runcores update pc
pmichaud: we know the context, but not the position
pmichaud but we know what annotations were in effect at the time of the throw 19:26
because that's hold in the context, yes?
*held
NotFound Mmm... Not sure. 19:27
pmichaud I'm pretty sure that's what jonathan++ said as to why we didn't need to know the exact position of the throw.
the annotations keep track of whatever position information is important.
NotFound Well, the current implementation works fine, we can just accurately describe what the lookup function does. 19:29
19:33 slavorgn joined
NotFound Regarding pbc segment headers: the 'size' field is used in bytecode segment and pir debug segment, ignored in the others. 19:35
<NotFound> Infinoid: that depends of the size of the opcode. --> Wrong, must be 'size of the instruction' 19:40
There are variants of die and throw with different number of operands. 19:41
Infinoid hmm. ignored meaning, "not checked when reading values from the segment buffer", or "not written to the struct when reading", or "not written to the pbc when writing", or what?
ah, ok
NotFound The size is zeroed in the other segments, and other field is used for the size of the first table. 19:42
Infinoid ... the size field in pbc or in c structs? 19:43
NotFound In the pbc.
Infinoid sounds lovely. I bet a lot of this stuff was getting in the way when writing pirc 19:44
NotFound These fields are described i packfile.c line 1296 but not in the pdd
itype, id, size 19:45
And the bytecode segment use size, not the segment size, for the number of opcodes. 19:46
19:46 slavorgn joined
Infinoid uh. I thought that segment header code was generic (there's a sort of pseudo-subclassing going on in the segment code) 19:46
NotFound pbc_checker.cpp works and shows correct results following this description. 19:47
Infinoid sounds like something worth fixing, anyway
NotFound In the current release, it accurately shows 32 and 63, be and le files. 19:48
64
Infinoid nice
NotFound And is a lot easier to read than pacfile.c... at least to me ;) 19:49
Infinoid I've been mostly treating packfile.c like a black box, for this stage of the pdd13 work. for stage 1 I'm just wrapping pmcs around the C structures, which is why I got blocked on the annotations reality/pdd13 mismatch 19:50
NotFound Initially I was thinking that there were alignment problems, but looks that all where due to the missing fields in the pdd description and the assumptions I was doing to compensate. 19:51
Infinoid its good that you're finding this stuff. any mismatch is worthy of a ticket, regardless of whether it ends up being a doc fix or a code fix 19:52
NotFound The initial question is: itype and id fields must go?
Infinoid that poor bot 19:53
19:53 slavorgn joined
dalek rrot: r37274 | allison++ | trunk/lib/Parrot/Docs/Section/Parrot.pm:
[doc] Add support policy to front page of generated HTML docs.
 19:54
Infinoid NotFound: I'm not sure anyone but jonathan can answer those questions :) 19:55
NotFound Let's write a ticket, then. 19:56
dalek rrot: r37275 | chromatic++ | trunk (1 files):
[docs] Renamed ticket wrangler guide to ticket triaging guide; will edit
 19:58
20:00 particle1 joined
Infinoid wants a kdiff3-like tool which compares documentation on the left with reality on the right, and produces a merged result 20:03
Coke allison: I can't remember, are these changes keeping history? 20:04
s/changes/migrations/
(old url in the commit message is wrong, I think.) 20:05
allison Coke: no history, just a relocation
NotFound Infinoid: that is the reason I wrote a program completely independent of parrot code and configuration information,
Just looks at the reality of the bytecode file.
allison Coke: sigh, yes the url is wrong 20:06
Coke: and that mailer commit hook *really* wants to keep [svn:parrot] at the front. Even after I change it everywhere, it just prepends it to the subject in addition to the configured subject prefix 20:08
20:08 donaldh joined
Coke you sure that's not the mailing list? 20:08
20:10 ron joined
ron find_method came up earlier. it only seems to be used for getting object pmc methods. could the pmc_ops documentation about searching the vtable just be off? 20:14
allison Coke: it *is* the mailing list (just confirmed by sending a commit message directly to myself from the script). Fixed. 20:18
NotFound There is some HLL that already generates code annotations? 20:23
dalek rrot: r37276 | allison++ | trunk (3 files):
[pheme] Relocating to svn.parrot.org/languages/pheme/trunk.
Infinoid NotFound: from use.perl.org/~JonathanWorthington/journal/38233 I think the plan was to make PCT generate annotations for all the PCT languages, but I don't know if that's been done yet 20:27
moritz it hasn't. 20:28
dalek tracwiki: v68 | allison++ | Languages 20:31
tracwiki: trac.parrot.org/parrot/wiki/Langua...ction=diff
shorten dalek's url is at xrl.us/bei4ag
dalek rrot: r37277 | chromatic++ | trunk/docs/project/ticket_triaging.pod:
[docs] Revised ticket triaging and wrangling documentation. Review welcome.
Coke chromatic: a lot of what's in there is still rt-specific language. 20:33
I can try to clean it up to be more trac specific and relegate rt to a footnote. 20:34
chromatic I revised a lot of it. Feel free.
Coke (e.g. there is no parrot queue, there is no stalled ....) 20:35
Sure, no problem.
dalek tracwiki: v131 | chromatic++ | ParrotRoadmap
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/bei4bb
dalek tracwiki: v132 | chromatic++ | ParrotRoadmap
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/bei4bd
Infinoid chromatic: If the ticket wrangling guide can't explain why the word "issues" is better than "problems", what hope do we mere mortals have? 20:36
pmichaud getting PCT to generate annotations is indeed in the plan. I'm not sure when it will happen -- certainly before the end of March.
chromatic Infinoid, that's a good question. I wasn't sure what to do with that section.
Coke chromatic: delete it. 20:37
moritz is there a way to unsubscribe from the [svn:languages] mails?
(while keeping the parrot commits, that is)
Coke chromatic: hurm. this devolves into "commit messages" at some point.
I think that belongs in a separate doc. 20:38
NotFound For example the annotations lookup function behaviour I was discussing is not causing any problem, but can be an issue
Infinoid chromatic: Also, the file says nothing about adding tests for regressions, though you've indicated in the past (and I agree) that these are important.
chromatic Good point.
Coke, are you actively editing this now?
Coke chromatic: just reading it.
Infinoid Though maybe that should go in the patch submitting guide, rather than the ticket triaging guide... 20:39
chromatic Okay, one moment then.
Coke was going to edit in a separate pass with more time.
ah, excellent.
purl well, ah, excellent. is my log publicly available?
Coke (getting beaten to it)++
purl: no one here wants to see your... log.
purl Coke: huh?
Infinoid Maybe by "publicly available", he meant trading shares of it on nasdaq. 20:40
NotFound If purl is a broker the crisis can have some hidden causes 20:41
Infinoid Brings whole new meaning to the phrase "market deflation".
20:43 pjcj_ joined
dalek rrot: r37278 | chromatic++ | trunk/docs/project/ticket_triaging.pod:
[docs] Removed a confusing discussion of "issue" versus "problem" and added

ticket triaging guidelines.		 20:44
Infinoid chromatic: If the ticket triaging guide shouldn't have a section on commit messages (as Coke mentioned), a simple example would probably be sufficient. 20:45
Mentioning the ticket number and the submitter in the commit message is also important I think, though I don't see those mentioned anywhere. 20:46
But if that stuff is moving to a different document, I'm getting dangerously close to bikeshedding.
chromatic I think the commit message stuff is in there. 20:47
I don't mind a little bit of overlap in the book.
NotFound So the toy pir preprocessor I wrote yesterday is the first "language" that annotates? Can I claim some prize? 20:48
Infinoid Ah, there it is. I was just searching for the wrong thing
NotFound: You win the game and have unlocked secret characters!
NotFound Ooooooooohhh
Let's play again!
Infinoid Here's some annotations you can play with. :) 20:49
chromatic: Are you still fiddling with the document? If so, I've got a couple of niggles. If not, I can commit them myself. 20:52
chromatic Go ahead. 20:55
Infinoid Line 337, "=item If related to an RT ticket, use the ticket title" applies to trac as well, and it'd be nice if the number in the commit message was preceded by "RT" or "TT"
line 387, they should create trac tickets, not RT ones. 20:56
20:57 korshak joined 20:59 korshak left, korshak joined
Coke pmichaud: are dev.perl.org/perl6/doc/synopsis.html up to date? 21:01
chromatic By "go ahead", I mean "feel free to commit".
pmichaud Coke: no, and at one time there was a page that redirected to the correct location. 21:02
I'm not sure what happned with that.
The 'Synopsis' link in the left sidebar is a bit closer to what should be there. 21:03
Coke can probably add an .htredirect on synopsis.html that takes you to the same link as the sidebar. should I do that? 21:04
moritz Coke: please do so
pmichaud yes, please.
moritz I updated the link in the side bar some time ago
but I didn't know about .htredirect
allison Coke: can we switch which repository dev.perl.org is pulling from?
pmichaud I don't think that dev.perl.org is pulling from any repository at the moment. 21:05
moritz and I don't have commit access anyway, I have to submit all patches to webmaster@perl.org
allison Coke: (may be a question for Robert)
dalek kudo: 0c893bc | (Moritz Lenz)++ | src/ (2 files):
Implement Any.reduce in Perl 6, bacek++, pmichaud++
shorten dalek's url is at xrl.us/bei4fj
allison pmichaud: yes, it pulls from the svn.perl.org/perl6
Coke er... looking dev.perl.org's .htredirects file, I'm no longer confident that's a simple switch.
there are quite a few rules in there already. 21:06
pmichaud okay, I misspoke. I don't know that the perl6 dev pages are being auto-generated from a repository at the moment.
moritz looks like fun ;-)
pmichaud yes, the pages themselves are coming from the repository.
at one time the synopsis HTML files were being autogenerated from svn.perl.org/perl6/doc/trunk 21:07
Coke dev.perl.org/perl6/doc/design/syn/S06.html seems to just point to pugscode.
pmichaud I don't know if that step is still occurring. I think it isn't.
Coke uh... dev.perl.org/perl6/pdd/ ? 21:08
thinking we can just delete all the pdds from the perl6 site. 21:09
(since they don't seem to be linked to.)
pmichaud I agree.
maybe a redirect would be useful there, though.
but it could be a generalized redirect for the pdd/ tree, instead of each individual document.
allison pmichaud: combust allows a displayed html page to be generated directly from a pod file in some remote repository, that's probably the right way to go 21:10
Coke er, also dev.perl.org/perl6/code.html
allison: IMO, it's not.
pmichaud right now we have the HTML generation taking place on pugscode.org -- same place as where the documents themselves are stored. 21:11
Coke because then you end up with issues when the file moves, or new files are added, or...
(a smarter combust template might deal with that; i have yet to see one)
allison Coke: it worked well enough for the old parrot docs, but yes we did run into those problems
pmichaud it's much easier for us to fix things up on pugscode.org than it is to do via combust on dev.perl.org
Coke well enough? no, it was a complete PITA.
allison Coke: well, I'm happy to be on the new strategy :) 21:12
Coke looks like someone could go through svn.perl.org/perl.org/docs/live/de...cts,perl6} and do a bit of cleanup. 21:13
lots of ooooold stuff in there.
... including the lie that I am the project manager for parrot. Fixing that... 21:15
the people.html also doesn't seem to reference a bunch of people that it should, and also references a few people that should probably be moved to "old guard". 21:19
(who page is probably replacable with a wiki page.)
*whole page
also removing Jesse as perl6 pm. 21:20
moritz www.perlfoundation.org/perl6/index....l_6_people perhaps? 21:21
dalek rrot: r37279 | allison++ | trunk/config/gen/languages.pm:
[config] Removing pheme and dotnet from makefile generation step.
 21:22
pmichaud Coke: do you know/remember the steps for granting someone bugadmin privs in RT?
GeJ Good morning everyone.
Coke pmichaud: it's in docs/proj 21:23
21:23 PacoLinux joined, szabgab joined
pmichaud is it still there after today's updates? 21:24
;-)
Coke trac.parrot.org/parrot/browser/tru..._guide.pod 21:25
shorten Coke's url is at xrl.us/bei4ii
allison pmichaud: you have to sync back to an earlier version of the document
pmichaud: it's been updated for Trac now
pmichaud okay
allison pmichaud: (you may want to take a copy of the earlier doc for rakudo) 21:26
pmichaud I can browse it through tags, I suspect.
Coke pmichaud: sorry, didn't realize someone had gutted it. =-)
allison pmichaud: anything before 37106 will work 21:27
Coke moritz: yes, something like that.
purl i heard something like that. was that true?
pmichaud allison: one can also view old versions through the tags/ subtree
(in the current rev)
I find that a bit easier sometimes. 21:28
allison pmichaud: makes sense, and that way you get a view fixed to a particular release
pmichaud e.g.: svn.parrot.org/parrot/tags/RELEASE_...cs/project
shorten pmichaud's url is at xrl.us/bei4ix
pmichaud okay, the metacommitter guide says how to update the parrot queue... how do I do it for the perl6 queue? ;-) 21:29
allison pmichaud: should be the same, but substitute 'perl6' for 'parrot' 21:30
pmichaud starts playing around with random urls.
allison pmichaud: oh, no I remember...
pmichaud no "parrot" in rt.perl.org/rt3/Admin/Groups/Membe...?id=178234 :-)
allison pmichaud: there's no public way to navigate to the pages
Coke the perms are the same.
allison pmichaud: and the pages are random numbers
pmichaud the perms are the same for perl6 and parrot? okay, that makes sense. 21:31
Coke yes.
pmichaud that's what it looks like from what I'm seeing, yes. Thanks.
Coke which confused me the last time someone asked me to make them a perl6 bugadmin.
allison Coke: that makes more sense
Coke so says robrt, so say we all.
allison so say we all
Coke by your command. I gotta get home.
21:33 wayland76 joined
dalek rrot: r37280 | allison++ | trunk/t/perl/Parrot_Distribution.t:
[languages] Remove dotnet from hardcoded path check.
 21:39
21:43 davidfetter joined, korshak left
dalek rrot: r37281 | allison++ | trunk (3 files):
[dotnet] Migrating .NET bytecode translator to
 21:43
21:47 bsdz joined
chromatic Projects which are focused on "awesome" tend, over the long term, to be rather more successful than projects which worry about what others might be saying about them. 21:55
(from an LWN article)
wayland76 Well, that looks good for perl6 and enlightenment :) 21:57
21:57 ruoso joined 21:58 Whiteknight joined
diakopter I'm guessing it's merely a reflection of attention/focus... projects focused on external perspective don't have enough resources to focus on being "awesome" let alone have enough time to figure out what "awesome" is 21:59
GeJ if only e17 could have a once-a-month release schedule. 22:00
Tene e17 has packages in fedora right now.
:)
GeJ or even a once-every-solar-eclipse, I'd go with that.
chromatic I don't know how a serious project can fail to have a release every year.
Tene So they're doing something a bit better than Parrot might be. 22:01
moritz we *had* debian packages in parrot 22:02
allison packagers love projects that don't make new releases, no work :)
moritz and they hat FTBS bugs, and nobody responded
so they were thrown out
Tene FTBS?
GeJ well, renderman is only releasing snapshots of the sub projects he maintains. and since he's giving some away, every snapshot available has less components available than the previous one. :(
moritz "failed to build from source"
GeJ s/renderman/rasterman/
moritz I think all it takes is a maintainer how really cares 22:03
allison moritz: I've been trying to get updated debian packages accepted for a year and a half now
moritz allison: I should add "a debian maintainer who really cares" 22:04
allison moritz: there seems to be some sort of process confusion somewhere
Tene allison: what do you need to get that moving faster?
allison Tene: an act of god?
Tene: seriously, we've been producing stable debian packages for over a year 22:05
moritz how do I generate a debian package from the svn?
allison Tene: so, it's mainly a matter of getting a debian maintainer to sponsor them
22:06 donaldh joined
allison mortiz: you generate a debian package from a release tarball, instructions are in docs/project/debian_packaging_guide.pod 22:06
moritz allison: I'll take a look, thanks 22:07
Tene So we keep asking them, and they just ignore it?
allison moritz: I recommend doing 0.9.0, if you want to experiment
Tene: no, we have 3 debian sponsors, and they agree to review the packages, and then drop off the map again (they're all quite busy with other things) 22:08
Tene Ah.
chromatic You know, chip is a Debian Developer. 22:10
allison I'd say he falls in the same "busy with other things" category 22:12
chromatic I'm exploring guilt and backtalk as a motivational technique.
Tene Arson is a significantly underused motivational technique. 22:13
allison chromatic/Tene: has been known to backfire, make sure you wear your personal shield :)
22:26 wayland joined 22:32 kid51 joined 22:34 donaldh joined 22:37 rurban joined
rurban seen NotFound 22:37
purl NotFound was last seen on #parrot 1 hours, 48 minutes and 44 seconds ago, saying: Let's play again!
rurban NotFound: I just confirmed that my simple tt254-64bit-pbc.patch fixes 64bit reading 32bit any everything else. And I found why ALIGN_16 is broken on 64bit. 22:39
I also looked more deeply at your pbc_checker and it is very inmature. the various ignores are just specials, not violating the specs 22:40
also the 64bit logic is okay, because you forgot segment_unpack
22:41 Tene_ joined
rurban I'll nopaste the patch. I still have a problem when thawing the interp pmc, with an invalid hash type (!= PMC) though. 22:41
22:42 Limbic_Region joined
nopaste "rurban" at 93.210.227.68 pasted "make 64bit pbc reading almost work." (354 lines) at nopaste.snit.ch/15855 22:44
22:56 Patterner joined
dalek rrot: r37282 | jkeenan++ | trunk/examples/pir/befunge/Makefile:
Add SVN Id tag and svn:keywords property per report in TT 428; mikehh++.
 23:02
23:15 PacoLinux joined 23:26 alvar joined
dalek rrot: r37283 | jkeenan++ | trunk/lib/Parrot/Test/Pod.pm:
Add to list of files which should not be examined for POD syntax because we have deliberately bad syntax in them for use in other parts of the test suite.
 23:26
23:30 bacek_ joined
dalek rrot: r37284 | jkeenan++ | trunk/tools/dev/install_files.pl:
Convert inline TODO item to Trac ticket 434.
 23:30
rrot: r37285 | jkeenan++ | trunk/examples/pir/befunge/Makefile:
Trying to get this to pass t/distro/file_metadata.t re svn:keywords.
 23:38
pmichaud Tene_: cardinal repo now created -- sent email with details. 23:41
Tene_ pmichaud++
pmichaud github.com/cardinal/cardinal
I have to take son to soccer game -- bbl
Tene_ Thank you for doing my work for me.
dalek rrot: r37286 | jkeenan++ | trunk/examples/pir/befunge/Makefile:
Still(!) trying to get the properties right. Will delete and restore.
 23:46
moderator Parrot 0.9.1 Released | parrot.org/ | < 1 week to Parrot 1.0! 23:47
cotto allison, can you update the numbers at the end of docs/project/release_manager_guide.pod to whatever was agreed on? 23:49
(version numbers)
allison cotto: okay, doing now 23:50
dalek rrot: r37287 | jkeenan++ | trunk/examples/pir/befunge/Makefile:
Restoring Makefile, in the hope that the svn properties will now be correct.
cotto thanks 23:51
allison cotto: they're right, someone must have already updated
cotto ok. I thought that the next stable release after 1.0 would be called 1.6. My mistake. 23:52
chromatic The next stable release after 1.0 will be 1.1. 23:54
allison cotto: your release will be 1.4, yes
dalek rrot: r37288 | jkeenan++ | trunk/src/packfile/pf_items.c:
Applying patches submitted by Andy Dougherty in trac.parrot.org/parrot/ticket/430 and trac.parrot.org/parrot/ticket/431. Quieting some warnings and other fixes.
cotto Sorry. "stable", not stable. 23:55
Infinoid "supported" :)
allison "eggplants"
chromatic I like "supported".
Infinoid I didn't mean to reopen any cans of worms, guys.
cotto ditto
allison :) 23:56
anyway, the version numbers in the guide are right 23:57
kid51 trac is sloooooooow 23:58
dalek rrot: r37289 | allison++ | trunk/docs/project/release_manager_guide.pod:
[cage] Add a reference to support policy from upcoming release list.

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!