Parrot 2.0.0 "Inevitable" released! | parrot.org | Priorities: deprecated core PMCs, ops -> dynops, GC tuning and implementation | Roadmap: icanhaz.com/parrotroadmap | Latest modified TT's: icanhaz.com/parrotbugs
Set by moderator on 2 February 2010.
dalek rrot: r43730 | allison++ | trunk/ports/debian (3 files):
Prepare for 2.0 packages.
00:04
00:18 bacek joined
Coke hates IRC. 00:20
or irssi.
or something.
dalek rrot: r43731 | allison++ | trunk/ports/debian (2 files):
Adding the parrot-nqp binary.
00:36
TT #351 closed by coke++: [TODO] Merge parrotcode.org into parrot.org 00:49
TT #402 closed by coke++: make html's generated pages should mention the version
Austin Don't be a hater, Coke. 01:03
dalek rrot: r43732 | allison++ | trunk/ports/debian (6 files):
Add a patch for the missing DESTDIR flag in the 'install-doc' target.
01:09
rrot: r43733 | allison++ | trunk/config/gen/makefiles/root.in:
[install] Add a missing DESTDIR flag in the 'install-doc' target,

directory during packaging.
rrot-data-structures: cb7ecff | Whiteknight++ | benchmarks/ (12 files):
remove the old .nqp-based benchmarks. Replace with a series of pir-based benchmarks with a factored library file
01:12
Whiteknight yay! dalek is reporting parrot-data-structures! 01:46
01:47 patspam joined
dalek rrot: r43734 | allison++ | trunk/ports/debian/parrot-devel.install.in:
[debian] Install the P6Regex library required by parrot-nqp.
02:15
02:32 ruoso joined
dalek rrot: r43735 | allison++ | trunk/ports/debian/parrot-devel.install.in:
[debian] Adding further libraries needed by parrot-nqp.
03:21
rrot: r43736 | allison++ | trunk/ports/debian (2 files):
[debian] Update Standards-Version to current required for sid.
rrot: r43737 | allison++ | trunk/ports/debian (2 files):
[debian] Add standard README.source file explaining use of quilt
03:38
03:48 janus joined 03:58 silug_ joined 04:04 cottoo joined 04:07 Infinoid_ joined, FullMetalHarlot joined, confound_ joined, particle1 joined, cotto_w0rk joined, purl joined, theory joined, Austin joined, lucian joined, Whiteknight joined, TonyC joined, cxreg joined, knewt joined, wagle joined, workbench joined, frodwith_ joined, zostay_ joined, Tene joined, dngor joined, ingy joined, bacek_at_work joined, eiro joined, jjore_ joined, TimToady joined, Ryan52 joined 04:08 leto joined, bacek_ joined, janus joined, ruoso joined, payload joined, kurahaupo_mobi joined, darbelo joined, jhelwig joined, Patterner joined, mikehh joined, tewk joined, plobsing joined, he joined, preflex joined, hiroyuki_y joined, slavorgn joined, slavorg joined, baest joined, GeJ joined, Maddingue joined, ascent joined, Hunger joined, Coke joined, NotFound joined, athomason joined, sri joined, treed joined, krunen joined, integral joined, dukeleto joined, redbrain joined, Khisanth joined
dalek rrot: r43738 | allison++ | trunk/ports/debian/patches (2 files):
[debian] Add documentation patches to pass lintian checks.
04:10
04:11 patspam joined
dalek rrot-data-structures: 0521a05 | Whiteknight++ | benchmarks/ (7 files):
add rapidfire benchmarks to all types.
04:11
rrot-data-structures: 6f75ad9 | Whiteknight++ | benchmarks.sh:
a quick sh script to run the benchmarks
rrot-data-structures: d41b330 | Whiteknight++ | (2 files):
small fixes. RPQ was throwing exception during the new rapidfire benchmark. Fix that and fix some output formatting
04:13 TimToady_ joined 04:14 eiro_ joined, ingyfoo joined 04:19 dngor_ joined, wagle_ joined, cxreg2 joined, KatrinaTheLamia joined, Tene_ joined, confound joined, Infinoid joined, zostay joined, Austin joined, estrabd joined, leto joined, bacek_ joined, janus joined, ruoso joined, payload joined, kurahaupo_mobi joined, darbelo joined, jhelwig joined, Patterner joined, mikehh joined, tewk joined, plobsing joined, he joined, preflex joined, hiroyuki_y joined, slavorgn joined, slavorg joined, baest joined, GeJ joined, Maddingue joined, ascent joined, Hunger joined, Coke joined, NotFound joined, athomason joined, sri joined, treed joined, krunen joined, integral joined, dukeleto joined, redbrain joined, Khisanth joined, eiro joined, confound joined, Ryan52 joined 04:23 frodwith joined 04:24 cotto_working joined, particle joined, particle1 joined, cotto_w0rk joined, purl joined, theory joined, lucian joined, Whiteknight joined, TonyC joined, knewt joined, wagle joined, workbench joined, dngor joined, bacek_at_work joined, jjore_ joined, bacek_at_work joined
Austin Is it me, or is the C3 linearizer totally broken? 04:43
04:44 krunen joined
Austin message pmichaud Is there a known problem with c3-linearizing basic p6object classes? I declare "class Parent {...}" and "class Child {...}" and get an "ambiguous hierarchy" failure when I try to do P6metaclass.add_parent(Child, Parent) 04:52
purl Message for pmichaud stored.
nopaste "Austin" at 68.37.46.53 pasted "C3 lin problem?" (13 lines) at nopaste.snit.ch/19492 04:53
Austin message pmichaud See nopaste.snit.ch/19492 04:54
purl Message for pmichaud stored.
dalek rrot: r43739 | allison++ | trunk/ports/debian/patches (2 files):
[debian] Add a simple manpage for parrot-nqp.
04:59
rrot: r43740 | allison++ | trunk/ports/debian/control.in:
[debian] Adding ${misc:Depends} to all binary packages, as requested
05:16
rrot: r43741 | allison++ | trunk/ports/debian/rules:
[debian] Generate and install the manpage for parrot-nqp.
05:29 patspam1 joined
dalek rrot: r43742 | allison++ | trunk/ports/debian/changelog:
[debian] Add changelog entries for doc patches.
05:32
06:02 theory joined
bacek o hai 06:11
seen whiteknight 06:12
purl whiteknight was last seen on #parrot 4 hours, 25 minutes and 40 seconds ago, saying: yay! dalek is reporting parrot-data-structures!
bacek karma lurker...
purl lurker... has neutral karma
dalek rrot: r43743 | bacek++ | branches/gc_encapsulate/src/runcore/trace.c:
Return early from trace to avoid crash on invalid pointer.
06:22
rrot: r43744 | bacek++ | branches/gc_encapsulate/src/gc/gc_ms.c:
Implement more introspection in gc_ms_info.
rrot: r43745 | bacek++ | branches/gc_encapsulate (55 files):
Sinc branch with trunk.

  \tPBC_COMPAT
  \tconfig/gen/makefiles/root.in
  \tsrc/pmc_freeze.c
  \tt/native_pbc/annotations.pbc
  \tt/native_pbc/integer_1.pbc
  \tt/native_pbc/number_1.pbc
  \tt/native_pbc/string_1.pbc
  \ttools/build/nativecall.pl
purl i guess config/gen/makefiles/root.in is what generates parrot/Makefile
purl src/pmc_freeze.c is, like, full of scary
purl somebody said tools/build/nativecall.pl was supposed to be run manually, right?
rrot: r43746 | bacek++ | branches/gc_encapsulate/src/gc/api.c:
Implement total_unblock in terms of unblock_mark/unblock_sweep
06:36 bacek joined 06:52 iblechbot joined 07:56 fperrad joined
dalek rrot: r43747 | bacek++ | branches/gc_encapsulate/src/pmc/imageio.pmc:
Fix building after merge with trunk.
08:00
rrot: r43748 | bacek++ | branches/gc_encapsulate/src/gc (3 files):
Expose compating of string pool via GC_Subsystem.
rrot: r43749 | bacek++ | branches/gc_encapsulate/tools/dev/mk_native_pbc:
mk_native_pbc now depends on make corevm only
rrot: r43750 | bacek++ | branches/gc_encapsulate/t/native_pbc (4 files):
Rebuild native PBC
rrot: r43751 | bacek++ | branches/gc_encapsulate/src/interp/inter_cb.c:
Fix fetching interpeter in NCI callback.
rrot: r43752 | bacek++ | branches/gc_encapsulate/src/gc/api.c:
Temporary poke into GC MS to support timely destruction of PMCs.
08:16 eternaleye joined 09:32 barney joined 10:01 barney joined
dalek rrot: r43753 | bacek++ | branches/gc_encapsulate (15 files):
Sync branch with trunc.

  \tconfig/auto/icu.pm
  \tsrc/pmc/imageio.pmc
10:44
rrot: r43754 | allison++ | trunk/ports/debian/copyright:
[debian] Add a copyright notice for external NQP files in the Parrot
11:34
rrot: r43755 | allison++ | trunk/ports/debian/changelog:
[debian] Note added copyright notice for nqp-rx in changelog.
11:50
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#32080), fulltest) at r43755 - Ubuntu 9.10 amd64 (g++ with --optimize) 12:03
dalek tracwiki: v24 | allison++ | ChrootSetup 12:19
tracwiki: trac.parrot.org/parrot/wiki/ChrootS...ction=diff
12:24 payload joined
dalek tracwiki: v25 | allison++ | ChrootSetup 12:35
tracwiki: trac.parrot.org/parrot/wiki/ChrootS...ction=diff
tracwiki: v26 | allison++ | ChrootSetup
tracwiki: trac.parrot.org/parrot/wiki/ChrootS...ction=diff
tracwiki: v27 | allison++ | ChrootSetup
tracwiki: trac.parrot.org/parrot/wiki/ChrootS...ction=diff
12:51 payload1 joined
dalek tracwiki: v28 | allison++ | ChrootSetup 12:52
tracwiki: trac.parrot.org/parrot/wiki/ChrootS...ction=diff
13:00 payload joined 13:10 payload1 joined 13:19 kid51 joined 13:34 JimmyZ joined 13:36 JimmyZ joined
dalek rrot: r43756 | allison++ | trunk/ports/debian/control:
[debian] Storing the generated control file, so quilt dependency can be
13:44
rrot: r43757 | allison++ | trunk/ports/ubuntu/README:
[ubuntu] Add a note explaining the Ubuntu-specific files.
14:17
rrot: r43758 | allison++ | trunk/src/parrot_debugger.c:
[cage] Add a short description after the executable name, needed for
rrot: r43759 | allison++ | trunk/tools/dev/pbc_to_exe.pir:
[cage] Add short description, required for generating manpage.
14:19 kurahaupo joined
dalek rrot: r43760 | allison++ | trunk/docs/project/ubuntu_packaging_guide.pod:
[ubuntu] Fix minor typos in the Ubuntu packaging guide.
14:34
rrot: r43761 | allison++ | trunk/ports/debian/changelog:
[debian] Update timestamp before building source packages.
14:44 dolmen joined 14:46 cxreg2 joined 14:47 confound_ joined 14:48 FullMetalHarlot joined, Tene joined 14:53 dngor joined 14:54 joeri joined, confound joined, jjore_ joined, workbench joined, knewt joined, TonyC joined, lucian joined, purl joined, particle joined, cotto_working joined, wagle joined, Whiteknight joined 14:59 JimmyZ joined 15:27 Psyche^ joined, cognominal joined 15:32 confound joined 15:33 JimmyZ joined, Whiteknight joined, wagle joined, cotto_working joined, particle joined, purl joined, lucian joined, TonyC joined, knewt joined, workbench joined, jjore_ joined 15:40 confound_ joined 15:49 Whiteknight joined, wagle joined, cotto_working joined, particle joined, purl joined, lucian joined, TonyC joined, knewt joined, workbench joined, jjore_ joined, confound joined 16:16 confound_ joined, Whiteknight joined, wagle joined, cotto_working joined, particle joined, purl joined, lucian joined, TonyC joined, knewt joined, workbench joined, jjore_ joined 16:23 mikehh joined, Hunger joined 16:26 dukeleto joined, leto_ joined, baest joined 16:27 khisanth_ joined 16:30 zostay joined 16:33 jan joined 16:39 kurahaupo joined
kid51 Hmm, we seem to be having some IRC problems this morning. 16:41
16:50 zostay joined, khisanth_ joined, leto_ joined, confound_ joined, Whiteknight joined, wagle joined, cotto_working joined, particle joined, purl joined, lucian joined, TonyC joined, knewt joined, workbench joined, jjore_ joined 17:03 kjeldahl joined 17:07 kjeldahl joined 17:09 kjeldahl joined 17:19 theory joined
Whiteknight IRC has been less than stable this morning 17:23
kid51 Yes, but certainly not due to excessive traffic ;-) 17:25
purl, are you there? 17:26
purl no idea, kid51
Austin seen pmichaud 17:27
purl pmichaud was last seen on #parrot 6 days, 1 hours, 58 minutes and 3 seconds ago, saying: okay, excellent work. [Feb 1 15:29:13 2010]
kid51 ponders the mere one inch of snow that fell in Brooklyn and wonders whether Whiteknight and Austin and Infinoid are snowed in 17:28
Whiteknight I haven't even tried to leave my apartment yet
Austin It's already melting here. The cars go by with that "ssss" sound of wet road. 17:29
(40 degrees outside)
Man. "Lap dances for Haiti" in a Toledo strip club. Thanks, CNN, for keeping me informed. 17:38
17:39 iblechbot joined
Austin Whiteknight, do you have a working parrot (2.0+) ? 17:40
17:44 zostay joined, khisanth_ joined, leto_ joined, confound_ joined, Whiteknight joined, wagle joined, cotto_working joined, particle joined, purl joined, TonyC joined, knewt joined, workbench joined, jjore_ joined
Austin Well, that's certainly irritating 17:44
kurahaupo has no experience of snow at home. It's just after sunrise and still 21°C, 90+%RH 17:49
Austin Where are you, kura?
kurahaupo 178E37S 17:50
Austin Hmm. Not Guam. Maybe new Zealand? 17:51
kid51 According to google maps, that's in the ocean about 300 km east of Auckland
Austin And how do you pronounce kurahaupo? Is that "how-po" or "wa-po" or "hoppo" or ...? 17:52
kurahaupo -36.909642,174.712931
Sorry for typo, like I said it's just after sunrise. 17:53
That's "koo rah how po"
Austin Well, if I find an airplane with " 17:54
Auckland" on it, I'll send you some snow. :)
So Chalmers St and St Georges Rd? 17:55
17:56 eternaleye joined
kurahaupo Is name of ancestral polynesian canoe that brought first settlers here. Also name of street where I once lived 17:56
Austin Is it a class name, or an individual name?
I think my keyboard wants some new batteries. 17:57
kurahaupo wanders outside to get a better gps reading
Austin That is, "The settlers came here in a kurahaupo" or "The settlers came here in the SS Kurahaupo"
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#32084), fulltest) at r43761 - Ubuntu 9.10 amd64 (gcc with --optimize) 17:58
Austin Ted William St and Trevola?
kid51 That's what I'm seeing
purl seeing is believing
Austin Yeah, I didn't put in enough digits the first time.
17:59 lucian joined
kurahaupo there were at least two canoes by this name over 200 years but only one at a time (to my knowledge) 17:59
Austin So I've learned something new already, and it's barely one o'clock. 18:01
Yesterday, I learned that Parrot hates me.
kid51 Austin, it's nothing personal. 18:02
Austin It sure seems that way.
kid51 clock 18:03
Austin Kid, do you have a working parrot 2+?
kid51 Well, I can always do an svn up and build one.
Is that what you had in mind?
Austin No, I was hoping you were working on something relatively current.
kid51 It will be current in about 5 minutes. 18:04
kid51 is never more than a few days out-of-date
Austin Ahh. 18:05
kid51 r u having trouble building?
Austin No, I'm having trouble believing this code doesn't work.
Can you run nopaste.snit.ch/19492 through parrot-nqp and see what it does?
kid51 Now, I'm not familiar with parrot-nqp
Austin "parrot-nqp test.nqp" 18:06
kid51 Should that paste be sufficient unto itself? No shebang line or anything like that? 18:07
Austin Just put the nopaste into a file (test.nqp) and run the installed parrot-nqp command on the filename. 18:08
(if you're doing installs)
It should either emit nothing, or emit an error. 18:09
According to me, it emits an error.
kid51 Hmm, nopaste not working normally. 18:10
Austin I think nopaste has left the building with all this bouncing that irc has been doing
Got a number?
or a url?
kid51 nopaste?
purl it has been said that nopaste is at nopaste.snit.ch/ (ask TonyC for new channels) or paste.scsys.co.uk/ or App::Nopaste or tools/dev/nopaste.pl or at www.extpaste.com/ or paste.scsys.co.uk (for #catalyst, #dbix-class, #moose and others) or gist.github.com/ or paste or gtfo or tools/dev/nopaste.pl or trac.parrot.org/parrot/browser/tru...nopaste.pl
nopaste "kid51" at 71.246.123.210 pasted "Output of: ./parrot-nqp ~/learn/parrot/foo.nqp" (7 lines) at nopaste.snit.ch/19495 18:11
Austin Brilliant. I get the same. 18:12
nopaste "kid51" at 70.85.31.226 pasted "Just me seeing if 'nopaste.pl' is working." (1 line) at nopaste.snit.ch/19496 18:13
"kid51" at 70.85.31.226 pasted "More testing of nopaste.pl; ignore" (7 lines) at nopaste.snit.ch/19497 18:15
Austin TT#1426 - C3 linearizer fails on two plain (P6object) classes 18:17
kid51 must get out of the house
dalek TT #1426 created by Austin_Hastings++: C3 linearizer fails on two plain (P6object) classes 18:20
18:31 Whiteknight joined 18:48 cognominal joined 19:38 kid51 joined
plobsing should src/pmc/pccmethod_test.pmc be a dynpmc? 19:56
19:58 kurahaupo joined
kurahaupo Purl msg Austin kurahaupo@36.909535S,174.712728E 19:58
purl Message for austin stored.
20:02 Whiteknight joined
Whiteknight what is the dynop stuff waiting for, vtable cleanups? 20:05
and is there any reason why not to delete the moving_core_to_dynops branch now and create a new one when we are ready 20:07
?
mikehh no reason not to - I just created it as requested 20:10
it's probably better to create it when needed
redbrain what are you guys working on? 20:11
mikehh be closer to trunk :-}
redbrain trunk seems to be very stable 20:13
20:16 bacek joined
redbrain ok scratch that :P pastebin.com/mf567bf1 20:16
20:28 kid51 joined
redbrain hey in the examples folder for like pir or pasm they are the only ones i play with but they seem to have the =Head ... stuff like man page style i see perl code alot like that isnt it a pain to handle from the parser perspective? 20:30
or do you just parse for the =cut or somthing
plobsing not really hard. AFAICT it's handled in compilers/imcc/imcc.l:285-303 20:36
redbrain yeah i guess it isnt hard but i would find it annoying :)
i should read more into the documentation but if i was to embed parrot i havent found anything on how to like add in client defined objects into the runtime 20:40
plobsing redbrain: AFAICT you have 2 options, create separate dynpmc libraries or use NCI to define subs, which you can attach to objects to get methods 20:41
redbrain plobsing: thanks i think i need to look at the NCI stuff now there is so many interfaces to parrot! :) 20:46
21:03 ash_ joined
dalek rrot: r43762 | mikehh++ | branches/moving_core_to_dynpmcs:
remove unused branch - can be re-created later
21:05
rrot: r43763 | mikehh++ | branches/remove_deprecated_vtable_entries:
remove unused branch - can be re-created later
Whiteknight which vtable entries are even deprecated? 21:06
is there a finite list of those?
plobsing bitwise VTABLE functions are
as well as get_bigint, set_bigint_*, pow, i_pow, and nextkey_keyed 21:07
21:09 KingOfKarlsruhe joined
Whiteknight ok 21:16
bacek Whiteknight, good morning 21:22
Whiteknight, I fixed NCI in gc_encapsulate branch. But something didn't marked properly. E.g. CallContext
kurahaupo What's the current thinking on preferring vtable vs method? How much faster is a vtable call than a method dispatch? 21:24
bacek kurahaupo, 100x
kurahaupo And we're *removing* bitops as vtables!!? 21:25
bacek kurahaupo, there is nothing to do with methods. 21:26
We are moving them to dynops.
Which is quite different from methods :)
and faster
bacek leaving to $dayjob 21:27
kurahaupo (Whew!)
21:33 confound joined
kurahaupo wonders how redbrain can have 74 failures out of 37 tests? 21:56
Perhaps the test framework itself is (also) failing? 21:57
21:58 payload joined
Whiteknight vtables are C functions, but there are LOTS of them 21:59
so the VTABLE structures are very large
cotto Whiteknight, don't clobber the bigint or pow VTABLE functions. 22:01
Allison posted in the deprecation ticket that they should stay.
I'm actually about to close that ticket.
Whiteknight damnit 22:02
I have a particular dislike of the BigInt PMC and the related VTABLEs
the pow VTABLE seems useless at best to me 22:03
cotto I don't understand the reason but nuking the bitwise VTABLE functions will yield a much better decrease in the _vtable struct size. 22:04
kurahaupo Whitelnight: when is an "Int" not an "int" and all that?
Whiteknight I really hate the 3-argument arithmetic VTABLES
and pow is a particularly useless one of those
cotto It is kinda shocking how large that struct is. On x86 it's 984B. 22:05
Whiteknight, if you want to make that point in the ticket, go for it.
Whiteknight which TT #? 22:06
cotto trac.parrot.org/parrot/ticket/866 22:07
Given the level at which PIR lives, I think you could make a good case for nuking the pow VTABLE functions. I'm not sure about a solution to the bigint ones. 22:09
dalek rrot: r43764 | plobsing++ | trunk (7 files):
convert PCCMETHOD_Test to dynpmc
22:11
purl I don't know how to convert PCCMETHOD_Test to dynpmc.
cotto poisoned botsnack
purl :)
cotto crosses his fingers 22:12
We really shouldn't be installing pbc_merge. 22:15
plobsing why not?
cotto It's broken. 22:16
plobsing we install a lot of broken stuff
cotto That doesn't make it a good idea.
Austin kurahuapo: An Int is a PMC type, with methods and vtable ops and what not. An 'int' is a low-level integer, stored in a $I register. 22:20
Cotto: Have you seen TT#1425? (I think it's -5) 22:21
cotto yup
Austin Are you still working on pbc_merge? 22:22
(IOW: Any chance you can fix that for me?)
cotto yes. I'm thinking that it'll require some significant work to become usable.
Austin Hell, it's usable now. It just doesn't do fixups right. 22:23
Kurahuapo: #26 or #28? 22:25
22:25 hercynium joined
plobsing just got parrot to throw an exception while building and exception while building an exception... 22:26
weee
22:30 cognominal joined
plobsing for the record, the exception system isn't supposed to throw exceptions right? 22:30
Austin How else we gonna throw them? 22:31
plobsing no, if something goes wrong *inside* the exception system
Austin :)
If they can still be thrown...
Whiteknight comment saved 22:35
plobsing not sure, gdb decided to terminate after >30000 frames deep of trying to throw exceptions 22:36
Austin Well, there you have it. It's a gdb problem, not a parrot problem. 22:37
plobsing yeah, gdb is such a quiter. I'm sure I would have broken out of that infinite loop eventually
Austin Given that PARROT_ARG_NAME is a wholly separate flag, is it possible to make a call with a variably named arg? 22:53
I guess that would be 'foo'(x :named($P0)) 22:54
never mind
23:10 patspam joined 23:13 lucian_ joined
dalek rrot: r43765 | plobsing++ | branches/nativecall_pir_build (3 files):
remove miniparrot dependency on src/nci.str (and src/nci.c)
23:17
Austin Anyone have a minute to check a patch? 23:31
23:31 jan joined
plobsing sure 23:31
Austin Too late.
Ah.
nopaste "Austin" at 68.37.46.53 pasted "Patch for tt#1425" (75 lines) at nopaste.snit.ch/19507
Austin If you "prove t/tools/pbc_merge.t" *after* applying the patch, but *before* re-building, it should fail. After rebuilding it should pass. 23:32
Now it's football time. 23:33
dalek rrot: r43766 | plobsing++ | branches/nativecall_pir_build/config/gen/makefiles/root.in:
remove .str files on make clean
plobsing Hunk #1 FAILED at 30. 23:36
cotto same here 23:38
23:38 eternaleye joined
plobsing it's the plan line, updating by hand, it seems sane... rebuilding now 23:38
cotto ah. It looks like he was using the 2.0 version of pbc_merge 23:39
Whiteknight I have never even looked at the source of pbc_mergge 23:40
seems like the kind of thing I should gander at eventually
cotto It's not very big. 23:41
plobsing with a by-hand fixup of the plan line, it performs as advertised.
Whiteknight stiffles desire to say "PBCFile" should be a PMC with a merge() method, then pbc_merge could just be a wrapper around that method
plobsing not that I know what it is supposed to be doing
Whiteknight but luckily, I didn't say it
plobsing Whiteknight: aren't packfiles becoming PMCs soonish? 23:42
Whiteknight plobsing: yes
they already are PMCs, we just don't use them
at least, not that I am aware. Not an expert on the subject
I think the Packfile PMC is the in-memory object, not the actual on-disk file 23:43
plobsing Whiteknight: I doubt we use them, bc if we did they would probably get GCed out from under us
nopaste "plobsing" at 67.212.22.93 pasted "Austin's patch updated for trunk" (75 lines) at nopaste.snit.ch/19508 23:50
plobsing AFAICT, the patch is good, but I don't know too much about pbc_merge 23:55
cotto I'm wondering if the similar switch above the patched one also needs the fix. 23:56
That might fix the bug that originally got me looking at pbc_merge. That'd be nice. 23:59