|
#parrot Parrot 4.11.0 "[All together - Happy Birthday Lovebird]" --with-threads | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC Set by moderator on 11 January 2013. |
|||
|
00:21
kid51 joined
00:40
Liz joined
00:45
woolfy joined
00:56
kid51_ joined
00:59
woolfy joined,
Liz joined
01:05
Khisanth joined
01:11
Mike-PerlRecruiter_ joined
01:24
bluescreen joined
01:32
Liz_ joined
01:33
woolfy joined
01:34
woolfy_ joined
01:48
Liz joined,
woolfy joined
01:53
kid51_ joined
01:56
japhb joined
02:05
Liz joined
02:06
woolfy joined
02:35
woolfy joined
02:36
Liz joined,
kid51 joined
02:44
kid51_ joined
02:56
woolfy joined,
Liz joined
03:09
Liz joined
03:11
woolfy_ joined,
Liz_ joined
03:19
woolfy joined,
Liz joined
03:20
MikeFair joined
03:41
MikeFair joined
03:47
Liz joined
03:48
woolfy_ joined
04:00
woolfy joined
04:01
Liz joined
04:43
woolfy joined
|
|||
| dukeleto | ~~ | 07:06 | |
| cotto | hio dukeleto | 07:10 | |
| dukeleto | cotto: how goes it? | 07:23 | |
| cotto: looks like I am cutting a release tomorrow | |||
| cotto: how goes it with you? | 07:29 | ||
| dalek | rrot: eaecde7 | mr.Shu++ | examples/benchmarks/sort.pir: added a simple sort benchmark |
07:31 | |
| rrot: 45be9e7 | mr.Shu++ | examples/benchmarks/sort.pir: added command line option |
|||
| rrot: 1248f2c | mr.Shu++ | examples/benchmarks/sort.pir: updated docs |
|||
| rrot: 7183b5b | rurban++ | examples/benchmarks/sort.pir: Merge remote-tracking branch 'mrshu/simple_sort_benchmark' into mrshu/simple_sort_benchmark-gh917 |
|||
| rrot: 2dbfd6a | dukeleto++ | MANIFEST: Run tools/dev/mk_manifest_and_skip.pl so that sort.pir gets added to the manifest |
|||
| rrot: 0150820 | mr.Shu++ | examples/benchmarks/sort.pir: added a simple sort benchmark |
|||
| rrot: 98aa4af | mr.Shu++ | examples/benchmarks/sort.pir: added command line option |
|||
| rrot: 5bc2f3e | mr.Shu++ | examples/benchmarks/sort.pir: updated docs |
|||
| rrot: 83e4639 | dukeleto++ | MANIFEST: Run tools/dev/mk_manifest_and_skip.pl so that sort.pir gets added to the manifest |
|||
| rrot: 4066aa6 | mr.Shu++ | examples/benchmarks/sort.pir: added default number of integers to sort |
|||
| rrot: 5260e25 | dukeleto++ | / (2 files): Merge branch 'mrshu/simple_sort_benchmark-gh175' |
|||
| rrot: e144f6a | dukeleto++ | CREDITS: Add @mrshu to CREDITS |
|||
| rrot: 285bd0a | dukeleto++ | examples/benchmarks/sort.pir: Cleanup FixedIntegerArray sorting benchmark |
|||
| cotto | dukeleto, just started a new gig and am getting used to having a bit less time to allocate. | 07:39 | |
| dukeleto | cotto: what are you coding in these days? | 07:40 | |
| cotto: i know how that goes | |||
| cotto | perl, so that's a nice change | 07:41 | |
| dukeleto | cotto: woot | ||
| cotto | goldenguru.com/team/ | ||
| dukeleto | cotto: nice. I talked to Gryphon a while ago | 07:42 | |
| cotto | dukeleto, what are you up to these days? | 07:44 | |
| dukeleto | cotto: too many things :) | 07:51 | |
| cotto: one thing is brewpony.com, which will launch soon | 07:52 | ||
| cotto: and I am also an instructor for OST: www.oreillyschool.com/ | 07:53 | ||
| cotto | congrats! | 07:55 | |
| dukeleto | cotto: thanks! | 07:56 | |
| cotto: i would like to talk with you about m0 soon, but it is close to bed time... | |||
| cotto: are you thinking about going to LFNW/OSCON or OSbridge this year? | 07:57 | ||
| cotto | dukeleto, I'm in the process of replying to your email. | ||
| yup. It'll be easier now that I have income. | |||
| dukeleto | cotto: yay! :) | 07:58 | |
| cotto | dukeleto, you have a bedtime? | ||
| dalek | rrot/gh925_ffa_sort: 3f65128 | dukeleto++ | / (2 files): Make FixedFloatArray sorting actually work and add a test #925 |
07:59 | |
| dukeleto | cotto: no, but I am close to passing-out time :) | ||
| dukeleto | now we can all sort our FixedFloatArrays, and the world was a slightly better place. | ||
| cotto | that'll do it | 08:00 | |
| dalek | rrot/gh925_ffa_sort: aeb9015 | dukeleto++ | / (2 files): [examples] Add benchmark for sorting FixedFloatArrays |
08:05 | |
| rrot/gh925_ffa_sort: 997d1d0 | dukeleto++ | examples/benchmarks/sort_ffa.pir: [doc] Improve synopsis for sort benchmark |
|||
| rrot/gh925_ffa_sort: 08e2d38 | dukeleto++ | ChangeLog: Update changelog |
08:08 | ||
| rrot: b52608b | dukeleto++ | src/pmc/fixedfloatarray.pmc: Add a sort method to FixedFloatArray, #925 |
08:09 | ||
| rrot: 3f65128 | dukeleto++ | / (2 files): Make FixedFloatArray sorting actually work and add a test #925 |
|||
| rrot: aeb9015 | dukeleto++ | / (2 files): [examples] Add benchmark for sorting FixedFloatArrays |
|||
| rrot: 997d1d0 | dukeleto++ | examples/benchmarks/sort_ffa.pir: [doc] Improve synopsis for sort benchmark |
|||
| rrot: 08e2d38 | dukeleto++ | ChangeLog: Update changelog |
|||
| rrot: 99676a6 | dukeleto++ | / (5 files): Merge branch 'gh925_ffa_sort' |
|||
|
08:48
travis-ci joined
|
|||
| travis-ci | [travis-ci] parrot/parrot#794 (master - 285bd0a : Jonathan "Duke" Leto): The build was broken. | 08:48 | |
| [travis-ci] Change view : github.com/parrot/parrot/compare/f...5bd0a07d6f | |||
| [travis-ci] Build details : travis-ci.org/parrot/parrot/builds/4160076 | |||
|
08:48
travis-ci left
09:27
travis-ci joined
|
|||
| travis-ci | [travis-ci] parrot/parrot#795 (master - 99676a6 : Jonathan "Duke" Leto): The build was broken. | 09:27 | |
| [travis-ci] Change view : github.com/parrot/parrot/compare/2...676a6663ee | |||
| [travis-ci] Build details : travis-ci.org/parrot/parrot/builds/4160497 | |||
|
09:27
travis-ci left
09:42
Psyche^ joined
|
|||
| dalek | rrot: 6496004 | rurban++ | tools/release/update_version.pl: [tools] rm MANIFEST.generated from release/update_version.pl |
11:36 | |
| Heuristic branch merge: pushed 36 commits to parrot/rurban/sockopt-gh909 by rurban | 11:37 | ||
|
11:48
sorear joined
|
|||
| dalek | rrot: b1ef06f | rurban++ | t/native_pbc/Test.pm: [t] omit TEST_PROG_ARGS warning in t/native_pbc/Test.pm |
12:01 | |
| rrot: 3754509 | rurban++ | t/native_pbc/ (12 files): regenerated all native_pbc with 5.0.0, ppc64 skipped |
|||
|
12:30
brrt joined
12:48
brrt left
13:11
Mike-PerlRecruiter_ joined
14:07
contingencyplan joined
14:16
PacoAir joined
15:06
bluescreen joined
15:44
Khisanth joined
15:58
dmalcolm joined
16:16
bluescreen joined
16:18
bluescreen joined
16:19
bluescreen joined
16:21
bluescreen joined
16:23
bluescreen joined
16:26
bluescreen joined
|
|||
| rurban | leto++ ffa_sort | 16:28 | |
|
16:28
slavorg joined
16:29
bluescreen joined
17:44
PacoAir joined,
PacoAir_ joined
|
|||
| dukeleto | looks like I broke some tests when I added .sort() to FFA. No good deed goes unpunished, I guess. | 18:05 | |
| wow. The travis-ci web interface is completely unusable | 18:07 | ||
| benabik | dukeleto: Just now noticing that? | ||
| Actually, it now appears to be blank to me, which is a whole new level of unusable. | 18:08 | ||
| dukeleto | benabik: that is when the js doesn't load | 18:09 | |
| benabik: that is why there is travis-lite.com/ , which I am now trying | |||
| travis-lite.com/parrot/parrot | |||
| benabik | Oh, hey, I can read stuff on that. Although I think travis-ci parses and summarizes the test output, which is useful. | 18:10 | |
| dukeleto | benabik: the fact that nothing renders on travis-ci with js turned off is a bit of a travesty | 18:11 | |
| benabik | I don't have JS off... | ||
| Although I do block cookies, which might be acting identically here. | 18:12 | ||
| dukeleto | benabik: yes, i know. Sometimes some bit of js doesn't load and the whole page doesn't load, as well | ||
| benabik | Ah. | ||
| dukeleto | looks like I might need to back-out adding .sort to FFA to get a release out today. Haven't I learned this lesson before? Don't add stuff the night before a release. | 18:13 | |
| One day I might even learn that lesson. | |||
| benabik: can you reproduce the failing fixedfloatarray tests ? | 18:14 | ||
| benabik | dukeleto: Compiling... | 18:15 | |
| dukeleto | smolder.parrot.org/app/projects/tap.../34325/151 | 18:16 | |
| benabik: it works with clang on Linux locally. I am sure I did something stupid. | |||
| benabik: my test could be written better, but I was lazy. Perhaps this is the price. | 18:17 | ||
| rurban | my smokers passed | 18:19 | |
| dukeleto | benabik: looks like travis is on Clang 3.1 but I have 2.8 locally | ||
| benabik | dukeleto: OS X/gcc = t/pmc/fixedfloatarray.t ..................... ok | 18:20 | |
| dukeleto | rurban: can you test it on a machine with clang 3.1? | ||
| rurban: or 3.x ? | |||
| rurban | let me see, I have 3.2 and 3.3 for sure | ||
| dukeleto | Am I doing anything really dumb in either of these commits? github.com/parrot/parrot/commit/b5...fae8114fda | 18:21 | |
| and github.com/parrot/parrot/commit/3f...2fd3b4f4a7 | |||
| benabik | Rebuilding with clang 4.1 (based on LLVM 3.1svn) | 18:23 | |
| Huh. Compiling with clang has finally made what those warnings are obvious... libffi has a PACKAGE macro, as does parrot) | 18:24 | ||
| rurban | oh | ||
| dukeleto | ooh boy, two macros enter, but only one can leave... | ||
| benabik | clang really does have significantly better error reporting. | 18:25 | |
| rurban | what libffi version do you have? | ||
| clang++ even more | |||
| benabik | Um... /usr/include/ffi ? | ||
| 2.1-pyobjc, fficonfig.h claims | 18:26 | ||
| dukeleto: Passes on OS X/clang 4.1 or 3.1 I dunno what Apple's doing with versions. | 18:27 | ||
| rurban | I have 3.0.10 | 18:28 | |
| head /usr/include/x86_64-linux-gnu/ffi.h | |||
| I believe I also tested with 3.0.11 and 3.0.9 | 18:29 | ||
| tests passed with clang version 3.2 (trunk 167594) | |||
| benabik | It looks like PyObjC installs a manually created fficonfig.h, which has the offending macro. I guess if I really care, I can `brew install libffi` | 18:30 | |
| rurban | and passed with clang version 3.3 (trunk 171459) | ||
| benabik | homebrew++ | 18:31 | |
| rurban | So we better rename our PACKAGE macro | ||
| Even if it's a PyObjC upstream bug | 18:32 | ||
| benabik | I view that as an upstream issue. `#define PACKAGE "libffi"` is not something that should happen in a public include file. | ||
| benabik doesn't care. | 18:33 | ||
| ... Homebrew outputs little glasses of beer to my terminal. | 18:34 | ||
| rurban | can we #undef PACKAGE? | 18:40 | |
| benabik | I don't see why not. | 18:42 | |
| Hm. Switching to clang loses ccache. Hm. | 18:45 | ||
| dukeleto | benabik: but clang is so much faster, that isn't so much of a big deal :) | 18:46 | |
| rurban: i had to #undef DEBUG or something like that in PL/Parrot, because both Postgres and Parrot wanted to define it. Seems reasonable | 18:48 | ||
| rurban | No prob. We use NDEBUG | ||
| For me clang is 2x faster | |||
| benabik | dukeleto: But wouldn't clang+ccache be even faster? ;-) | ||
|
19:29
tuxit joined
|
|||
| dalek | rrot: 9baf8c3 | mrshu++ | CREDITS: Updated CREDITS to hold more info about @mrshu |
19:43 | |
| rrot: 94e9219 | dukeleto++ | CREDITS: Merge pull request #928 from mrshu/patch-1 Updated CREDITS to hold more info about @mrshu |
|||
|
20:27
bluescreen_ joined
20:41
perlite_ joined
20:51
lue joined
|
|||
| lue | I've gotten a weird new insufficient permissions problem running `sudo make install`, during the Pod generation part. perldoc can't seem to open packfile-c.pod for writing. Running `make install` gets that file, but then fails with ops/bit.pod | 20:54 | |
| Every single file in the parrot repo belongs to me and is writable by me, at least according to ls -l . I have a sneaking suspicion this is due to me upgrading my Perl 5 not too long ago, but I'm not sure what to check to see if this is the case. | 20:55 | ||
|
21:05
Reini joined
|
|||
| Coke | are you using perlbrew? | 21:12 | |
| lue | I don't think so. (Unless something is using it without me knowing) | 21:13 | |
| Coke | you'd have upgraded your perl with "perlbrew install ..." | 21:14 | |
| nice thing about that is that you can easily switch back to the old perl if something like this happens, for testing. | |||
| arnsholt | There's a simple way to check | 21:15 | |
| lue | No, upgraded through portage (5.12 -> 5.16, in case it matters) | ||
| arnsholt | lue: What does "which perl" give you? | ||
| lue | /usr/bin/perl | ||
| arnsholt | Yeah. No perlbrew | ||
| 'tis odd though. What's the mode on ops/? | 21:16 | ||
| Only thing I can think of that blocks even root from reading a file would be the directory lacking an execute bit | 21:17 | ||
| lue | docs/ops is rwxr-xr-x | 21:18 | |
| arnsholt | Weeeird. What happens if you su and try to run the same command as make install? | 21:19 | |
| lue | same problem (so it's not a weird sudo issue) | ||
| arnsholt | What's the error message you get? | 21:20 | |
| (And what's the command, for that matter) | |||
| lue | gist.github.com/4542183 | 21:21 | |
| Also, I only run the make install command as root, to put it all in /usr/local | |||
| packfile-c.pod is what it first fails on, unless I run make install as myself (after which make install fails unconditionally on the ops/ stuff) | 21:22 | ||
| arnsholt | That's pretty weird | ||
| Have you tried make distclean/git clean -xd? | 21:23 | ||
| lue | O.O I just deleted packfile-c.pod because it was empty (maybe I touched it and forgot to remove it), and ran `make install` as normal user, and it all works. WTF | ||
| (oh, and I've run git clean -dfx many a time already) | 21:24 | ||
| arnsholt | Weird. What's git status say after you killed that pod file? | ||
| lue | (and I just installed it all with sudo make install) | 21:25 | |
| working directory clean | |||
| arnsholt | Weirdness | ||
| Oh well, at least it works now =) | 21:26 | ||
| lue | I have no words for what happened. I've deleted empty packfile-c.pod before in all this (although it's the first time it was write-protected too) | 21:27 | |
| Reini | I am at a car service wifi, bad bandwidth. | ||
| lue: your problem was that you do sudo make install, which did doc as root. | |||
| lue | but, that's never been a problem before this time. | 21:28 | |
| Reini | sudo rm -rf doc/ops; make installable docs, sudo make install should fix it | ||
| lue | it works now, so I'll keep that in mind the next time this happens, Reini | 21:29 | |
| Reini | I recommend to do make all installable; sudo make install | ||
| these targets are new. we forgot them before. | 21:30 | ||
| docs are better than no docs. we left out all ops and packfile docs | |||
| lue | OK. I'm compiling NQP now, so I Parrot's installed fine for me this time. But like I said, I remember the 'installable' targets | 21:32 | |
|
22:04
PacoAir joined
23:14
PacoAir joined
23:19
benabik joined
23:57
kid51 joined
|
|||
| kid51 | smolder reports are showing failures in t/pmc/fixedfloatarray.t | 23:58 | |
| e.g., smolder.parrot.org/app/projects/rep...ails/34387 | |||