Camelia, the Perl 6 bug

IRC log for #parrot, 2011-07-19

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:02 jay joined #parrot
00:06 kid51 cotto_work: I notice that both you and Coke are getting PASS on Win32 boxes where you configure with: "--cc=cl" "--link=link" "--ld=link" (e.g., http://smolder.parrot.org/app/​projects/report_details/18173; http://smolder.parrot.org/app/​projects/report_details/18152)
00:09 cotto_work yes
00:15 rurban_ joined #parrot
00:17 rurban left #parrot
00:17 rurban_ is now known as rurban
00:20 Coke (pass) yes. we did the manual update on the offending perlhist.txt file.
00:24 * Coke responded to the email.
00:26 kid51 It looks like "--cc=cl" "--link=link" "--ld=link" is our best bet on Windows
00:35 cotto_work I usually have good luck using Strawberry/mingw too.
00:37 NotFound_b The failures I have in os links seems to be due to the fat filesystem and/or the XP Home version.
00:38 NotFound_b The ones in socket, no idea.
00:39 kid51 If ./parrot_config --dump | grep THREAD    yields:      HAS_THREADS => '1', does that mean that this box qualifies as "Parrot is multi-threaded" in PLATFORMS?
00:40 Felipe ./parrot_debugger is not used?
00:42 kid51 Felipe:  Well, I've never used it, but then, I probably wouldn't know how. But that's probably why we have soh_cah_toa's GSOC project.
00:44 Felipe okay, because it's crashing when passing any parameter.
00:51 cotto_work looks like a couple tests in t/dynpmc/os.t fail on strawberry/mingw.  smolder is running and will submit soon
00:51 dalek parrot: 43c34c4 | jkeenan++ | PLATFORMS:
00:51 dalek parrot: Add 2 entries for darwin/ppc.
00:51 dalek parrot: review: https://github.com/parrot/parrot/commit/43c34c4b4e
00:51 cotto_work I'm out, but will test in the morning if it's not known to pass.
00:54 kid51 Coke: re renaming that file: Do you think that's something we can risk tonight?  Or should we wait until after release?
01:03 woosley joined #parrot
01:06 woosley left #parrot
01:08 bluescreen_ joined #parrot
01:12 nopaste "kid51" at 192.168.1.3 pasted "Implementing suggestion re changing name of perlhist.txt." (663 lines) at http://nopaste.snit.ch/62455
01:14 kid51 I'll apply that paste in an hour unless there is an objection.
01:15 kid51 is now known as kid51_at_dinner
01:36 nakajima joined #parrot
01:44 cotto ~~
01:44 woosley joined #parrot
01:48 Felipe ~~
02:02 bluescreen__ joined #parrot
02:08 dalek parrot: 4c9445a | jkeenan++ | / (5 files):
02:08 dalek parrot: Change name of file used for testing to force update.
02:08 dalek parrot: review: https://github.com/parrot/parrot/commit/4c9445a9cb
02:08 dalek parrot: 8a06578 | jkeenan++ | t/library/perlhist.txt:
02:08 dalek parrot: Make sure the old file comes out.
02:08 dalek parrot: review: https://github.com/parrot/parrot/commit/8a06578a06
02:13 kid51_at_dinner cotto: ping
02:13 kid51_at_dinner is now known as kid51
02:13 cotto kid51, pong
02:13 kid51 cotto: re release guide paragraph 3.7:  How do I know whether "new entries to PBC_COMPAT have been added since the last supported release"?
02:14 cotto kid51, all changes are dated
02:15 kid51 The most recent I see is: 10.0    2011.01.18  cotto   released 3.0.0
02:15 cotto kid51, There you go.
02:16 kid51 However, lines 12-18 of PBC_COMPAT suggest that there are at least 6 different conditions that could require a change.
02:17 cotto kid51, we probably forgot.  Bumping the version for 3.6.0 is a good idea.
02:17 kid51 Does a human have to remember to enter a change in that file whenever any one of those conditions has changed?
02:17 kid51 So we'll go to version 11.0
02:17 cotto Yes.  It would be difficult to automate.
02:17 kid51 k
02:19 everbrew joined #parrot
02:28 theory joined #parrot
02:29 kid51 Alright, away we go.
02:30 bluescreen left #parrot
02:30 bluescreen_ left #parrot
02:30 bluescreen__ left #parrot
02:30 Topic for #parrot is now Release is underway; no commits to master till further notice
02:32 dalek parrot: ff8079e | jkeenan++ | docs/project/release_manager_guide.pod:
02:32 dalek parrot: Correct typo in release manager guide.
02:32 dalek parrot: review: https://github.com/parrot/parrot/commit/ff8079ec70
02:37 everbrew left #parrot
02:42 soh_cah_toa where can you download the plumage tarball? i can't find it anywhere
02:48 Felipe 'night
02:55 kid51 Hmm ... I never knew I was listed as Project Manager in RESPONSIBLE PARTIES!
02:56 kid51 We'll discuss that *after* the release.
03:07 kid51 cotto: ping
03:09 Drossel left #parrot
03:11 soh_cah_toa is it possible to install plumage to somewhere besides /usr/local/bin? i tried using PREFIX=foo and --prefix=foo with `parrot setup.pir` but that doesn't seem to be working
03:12 cotto kid51, pong
03:12 kid51 cotto: Experiencing problem at 3f & 3g
03:13 kid51 When I ran make distro_tests, I got an incompatible bytecode error during make
03:13 cotto kid51, did you run make reconfig?
03:13 Kulag joined #parrot
03:14 kid51 Ah, I missed that.
03:15 kid51 So I just ran 'make reconfig'.  Will I have to *re-run* tools/relesae/update_version.pl?
03:15 cotto kid51, nope.  reconfig only clobbers artifacts of the build and configuration
03:16 kid51 Okay, then I shall proceed (again) to 3g
03:19 NotFound_b soh_cah_toa: It gets installed in the parrot bin directory. If you want to change it you must reconfigure and reinstall parrot
03:21 Kulag left #parrot
03:22 soh_cah_toa NotFound_b: ok. so whatever directory parrot is installed to, that's where plumage goes. /usr/local/bin by default
03:22 soh_cah_toa NotFound_b: to install parrot into somewhere else, do you use the --prefix switch or set the PREFIX variable w/ Configure.pl?
03:22 kid51 sh tools/dev/mk_packfile_pbc
03:22 kid51 error:imcc:syntax error, unexpected PARROT_OP ('print')
03:22 kid51 in file '-' line 1
03:22 NotFound_b soh_cah_toa: --prefix will do
03:22 soh_cah_toa ok
03:22 kid51 error:imcc:syntax error ... somewhere
03:22 kid51 in file '-' line 1
03:22 kid51 syntax error ... somewhere
03:23 cotto kid51, is that the first bit of output you get?
03:23 kid51 Yes -- from sh tools/dev/mk_packfile_pbc
03:24 kid51 PackFile_Header_validate: This Parrot cannot read bytecode files with version 10.0.
03:24 kid51 current instr.: 'parrot;PGE;Exp;__onload' pc 2975 (compilers/pge/PGE/OPTable.pir:27)
03:24 kid51 make: *** [runtime/parrot/library/PGE.pbc] Error 1
03:24 kid51 ... after: sh tools/dev/mk_native_pbc
03:25 cotto kid51, did you update PBC_COMPAT before running make reconfig?
03:25 nopaste "kid51" at 192.168.1.3 pasted "Current release 'git status'" (28 lines) at http://nopaste.snit.ch/62456
03:26 NotFound_b make reconfig does not do a realclean, doesn't it?
03:26 kid51 I did, then I tried to revert
03:26 kid51 How should I proceed?
03:26 cotto NotFound_b, it does realclean and runs Configure.pl with the same arguments originally passed to it
03:27 kid51 Suppose that I revert PBC_COMPAT, do make realclean, perl Configure.pl, make
03:28 kid51 Will that put me in a position to proceed
03:28 kid51 ?
03:28 cotto Updating PBC_COMPAT is part of the release.
03:28 kid51 I understand that.
03:28 kid51 But I appear to have gotten really tangled up.
03:29 kid51 Failing to do 'make reconfig' at precisely the right point has screwed me up completely.
03:30 cotto if you know that the tests pass, you can reset to HEAD, run the update script, bump PBC_COMPAT and configure/build
03:30 Kulag joined #parrot
03:32 kid51 reset everything to HEAD?
03:32 cotto git reset --hard HEAD
03:33 cotto that will lose local changes, so don't do that if you have anything hard to reproduce
03:35 kid51 Do I run: tools/release/update_version.pl    before or after   make reconfig
03:35 cotto before
03:37 Kulag left #parrot
03:37 Kulag joined #parrot
03:39 kid51 Having just run 'make reconfig', do I have to rerun Configure.pl as implied by 3f?
03:40 cotto reconfig runs Configure.pl
03:40 kid51 I'll take that as a "No"
03:41 cotto right
03:41 kid51 So that's one thing that has to be corrected in the release guide
03:42 dalek tracwiki: v3 | soh_cah_toa++ | GettingStartedWithPlumage
03:42 dalek tracwiki: Added a few examples, fixed some grammar mistakes, and further elaborated in several paragraphs.
03:42 dalek tracwiki: http://trac.parrot.org/parrot/wiki/GettingS​tartedWithPlumage?version=3&action=diff
03:42 dalek tracwiki: v182 | soh_cah_toa++ | Languages
03:42 dalek tracwiki: Added a small paragraph about Plumage and a link to its wiki.
03:42 dalek tracwiki: http://trac.parrot.org/parrot/wiki/L​anguages?version=182&action=diff
03:42 Drossel joined #parrot
03:42 NotFound_b I don't see handling for the -a option  in frontend/parrot/main.c Maybe it got forgotten in the frontend changes?
03:43 kid51 Now patching PBC_COMPAT to add version 11.0.
03:44 Kulag left #parrot
03:44 kid51 Next question: How do I determine whether running sh tools/dev/mk_packfile_pbc [is] necessary?  (3 g 4th graph)
03:45 kid51 3g 4th graph implies this may or may not be necessary; how do I determine?
03:46 cotto "graph"?
03:46 kid51 paragraph
03:46 cotto I don't see any implication that it's not necessary.
03:46 cotto nm
03:46 cotto brain failure
03:46 kid51 "if necessary"
03:47 cotto yeah
03:47 Drossel left #parrot
03:47 kid51 This is where I got the IMCC syntax error before
03:47 kid51 Should I dare it?
03:47 NotFound_b kid51: see my last line about the -a option
03:48 cotto That's a special script,
03:48 NotFound_b The syntax error is due to reading the heredoc as pir, not pasm
03:48 kid51 NotFound_b: I don't know anything about the frontend changes.
03:49 NotFound_b Anyway, the -a oprion doesn't not work, regardless of the cause.
03:49 kid51 So, should I simply run:  sh tools/dev/mk_packfile_pbc  ?
03:50 nakajima left #parrot
03:51 cotto It appears to be broken.  Are you comfortable turning that pasm snippet into pir?
03:52 NotFound_b kid51: Isn't going to work, no matter how you run it.
03:53 kid51 I have no idea how to proceed here.
03:53 kid51 Tell me what to paste that can help you help me.
03:53 nopaste "cotto" at 192.168.1.3 pasted "fix mk_packfile_pbc" (43 lines) at http://nopaste.snit.ch/62457
03:53 NotFound_b kid51: if make test pass, just ignore it
03:54 cotto just the last chunk of that is necessary
03:54 nopaste "cotto" at 192.168.1.3 pasted "fix mk_packfile_pbc (the important bits)" (15 lines) at http://nopaste.snit.ch/62458
03:55 kid51 Trying: sh tools/dev/mk_packfile_pbc
03:56 * kid51 breathes sigh of relief
03:56 kid51 $ sh tools/dev/mk_packfile_pbc
03:56 kid51 t/native_pbc/integer.pbc updated
03:56 kid51 t/native_pbc/number.pbc updated
03:56 kid51 t/native_pbc/string.pbc updated
03:56 kid51 Next: trying: sh tools/dev/mk_native_pbc
03:56 NotFound_b We should delete the full t/native_pbc directory. There is no point in mantaining it.
03:57 cotto If we can show that it's useless, I'd love to drop it.
03:58 NotFound_b cotto: skip_all is a clear sign of useless to me.
03:58 kid51 FAIL
03:58 cotto NotFound_b, indeed
03:58 kid51 compilers/imcc/imcc.l:1287: warning: no previous prototype for 'determine_input_file_type'
03:58 kid51 PackFile_Header_validate: This Parrot cannot read bytecode files with version 10.0.
03:58 kid51 current instr.: 'parrot;PGE;Exp;__onload' pc 2975 (compilers/pge/PGE/OPTable.pir:27)
03:58 kid51 make: *** [runtime/parrot/library/PGE.pbc] Error 1
04:01 contingencyplan joined #parrot
04:01 NotFound_b kid51: delete runtime/parrot/library/PGE.pbc
04:01 cotto kid51, that's with PBC_COMPAT bumped?
04:02 kid51 $ grep -nC1 'released 3.0.0' PBC_COMPAT
04:02 kid51 31-11.0 2011.07.19      jkeenan released 3.6.0
04:02 kid51 32:10.0 2011.01.18      cotto   released 3.0.0
04:02 kid51 33-9.0  2010.10.19      gerd    released 2.9.0
04:02 theory left #parrot
04:03 soh_cah_toa left #parrot
04:04 kid51 cotto: Do you agree with NotFound re PGE.pbc to be removed?
04:06 bubaflub left #parrot
04:06 NotFound_b kid51: is just a test to see if it gets recreated with the wrong number again
04:07 kid51 sh tools/dev/mk_native_pbc
04:07 kid51 tools/dev/mk_native_pbc: line 35: ./parrot_config: No such file or directory
04:08 cotto kid51, there's no way that deleting it could mess anything up
04:08 kid51 "No way"  ... but it just did, apparently.
04:09 cotto you just have to run make to rebuild it
04:09 NotFound_b kid51: you should make after that
04:09 kid51 Plain 'make'?
04:09 kid51 Trying: make clean && make
04:09 Drossel joined #parrot
04:10 NotFound_b Looks like PBC_COMPAT has a lot less versions than it used to have. There is some police about deleting minor changes?
04:11 kid51 NotFound_b: That's specified in docs/project/release_manager.pod
04:11 kid51 NotFound_b: That's specified in docs/project/release_manager_guide.pod
04:11 kid51 Though I did not have to deal with any minor changes.
04:13 NotFound_b I suppose that someone applied that policy retroactively at some point, then.
04:13 kid51 Better results this time, it appears.
04:13 kid51 Linked: winxed  !!!
04:14 NotFound_b make realclean correctly deletes PGE.pbc for me.
04:14 kid51 Damn!
04:15 Kulag joined #parrot
04:15 contingencyplan left #parrot
04:15 nopaste "kid51" at 192.168.1.3 pasted "Result of sh tools/dev/mk_native_pbc" (27 lines) at http://nopaste.snit.ch/62459
04:15 Drossel left #parrot
04:17 nopaste "kid51" at 192.168.1.3 pasted "Results of the automatic rerun" (5 lines) at http://nopaste.snit.ch/62460
04:21 Drossel joined #parrot
04:21 Kulag left #parrot
04:21 NotFound_b The guide does not mention bootstraping ops. I think the last time the issue got mentioned the conclusion was that it's needed.
04:22 kid51 See 3(a)(i)
04:23 cotto NotFound_b, I made it unnecessary by adding better laziness to tools/build/update_version.pl
04:24 NotFound_b Ah, good.
04:24 cotto Updating takes a second now instead of several minutes
04:24 cotto assuming that the update process doesn't involve adding any new ops
04:30 Drossel left #parrot
04:34 daniel-s joined #parrot
04:47 kid51 Do I take it that t/native_pbc/annotations.pbc does NOT get added to git?
04:51 NotFound_b kid51: give that is not in the repo right now, I suppose so.
04:52 dalek parrot: 4039e9d | jkeenan++ | / (18 files):
04:52 dalek parrot: Release data for Parrot 3.6.0.
04:52 dalek parrot: review: https://github.com/parrot/parrot/commit/4039e9d0b5
04:57 Kulag joined #parrot
05:01 Drossel joined #parrot
05:03 Kulag left #parrot
05:07 cotto kid51, are we there yet?
05:07 Drossel left #parrot
05:09 Kulag joined #parrot
05:13 Drossel joined #parrot
05:14 Kulag left #parrot
05:20 Kulag joined #parrot
05:22 Drossel left #parrot
05:30 Drossel joined #parrot
05:30 Kulag left #parrot
05:31 dalek website: jkeenan++ | Parrot 3.6.0 "P�jaros del Caribe" Released
05:31 dalek website: http://www.parrot.org/news/2011/Parrot-3.6.0
05:42 Topic for #parrot is now Parrot 3.6.0 "Pájaros del Caribe" released | http://parrot.org | Log: http://irclog.perlgeek.de/parrot/today | #parrotsketch meeting Tuesday 1930 UTC
05:44 kid51 http://trac.parrot.org/parrot/ (wiki front page) is utterly out-of-date
05:45 NotFound_b left #parrot
06:00 kid51 Not done: 11f (make html; update docs.parrot.org); news to places like Perlmonks, Slashdot; awaiting new Wikipedia data; update release_manager_guide.pod on basis of tonight's experiences
06:00 kid51 left #parrot
06:21 Ademan left #parrot
06:23 szbalint_ is now known as szbalint
06:29 nopaste left #parrot
06:30 TonyC left #parrot
06:31 TonyC joined #parrot
06:36 nopaste joined #parrot
06:53 Drossel left #parrot
06:55 Kulag joined #parrot
07:02 Kulag left #parrot
07:13 Kulag joined #parrot
07:18 fperrad joined #parrot
07:19 Kulag left #parrot
07:20 Kulag joined #parrot
07:42 mj41 joined #parrot
08:14 rurban_ joined #parrot
08:17 rurban left #parrot
08:17 rurban_ is now known as rurban
08:53 Drossel joined #parrot
08:54 Kulag left #parrot
09:05 contingencyplan joined #parrot
09:09 preflex left #parrot
09:11 preflex joined #parrot
09:12 daniel-s left #parrot
09:14 Drossel left #parrot
09:15 Kulag joined #parrot
09:55 woosley left #parrot
09:59 Coke (cc=cl) Yah, I've always used strawberry when it was available. I'm using strawbery+msvc now mainly just to test this config.
10:24 JimmyZ joined #parrot
10:30 Kulag left #parrot
10:34 Kulag joined #parrot
10:35 Felipe morning
10:37 JimmyZ morning Felipe
10:37 JimmyZ left #parrot
10:40 Kulag left #parrot
10:51 preflex left #parrot
10:52 contingencyplan left #parrot
10:53 preflex joined #parrot
11:37 JimmyZ joined #parrot
11:45 Kulag joined #parrot
11:47 JimmyZ_ joined #parrot
11:50 JimmyZ left #parrot
11:50 JimmyZ_ is now known as JimmyZ
11:57 JimmyZ left #parrot
12:03 lucian joined #parrot
12:18 whiteknight joined #parrot
12:21 whiteknight good morning, #parrot
12:21 atrodo Good morning whiteknight
12:21 whiteknight kid51++ on the release
12:22 benabik ~~
12:22 whiteknight good morning atrodo, benabik
12:22 whiteknight atrodo: Did you get all the data you needed for ipfy?
12:22 atrodo whiteknight> Yep, sorear++ got me what I needed
12:23 atrodo I'm still not going to have any actually built runs for over a month and half, but the data points are known
12:24 whiteknight okay
12:24 whiteknight that's better than nothing
12:25 atrodo Yep.  Unfortunately, the bug will always prevent that machine from producing results for that month and a half.
12:25 whiteknight atrodo: For those graphs on ipfy.com, can we put in some targets? Like a blue line that shows where we would like to be?
12:25 atrodo On the plus side, I should be able to easily add a new machine to run the tests as well
12:25 benabik atrod: "the bug"?
12:25 benabik atrodo: ^^
12:26 atrodo benabik> Rakudo was missing a write barrier for the new GC, which meant that it wouldn't build until that bug was fixed
12:26 benabik Oh, ew.
12:26 benabik I'd suggest a couple of lines for "last release", "average of last year", or similar.  So when we hit speed regressions it's very noticable.
12:26 atrodo benabik> And it looks like it was only on lower memory sized machines
12:27 benabik And we can easily see if we get better too.
12:27 atrodo whiteknight, benabik> Yea, I'm hoping to have something show up sometime.  Probably a chosen release(s) that show up at every day
12:29 atrodo fascinating.  whiteknight, any chance you can hit the test service hook button again?
12:31 bluescreen joined #parrot
12:33 atrodo whiteknight> n/m, i just realized I can just replay old packets
12:35 whiteknight atrodo++ The work you're doing on this is extremely valuable. Thanks for it
12:42 Util kid51++ # A release worthy of Long Term Support
12:49 Felipe :)
12:54 Util msg dukeleto The pics are in folders Parrot_Perl6_BOF and Parrot_Perl6_Hackathon_Dinner at: https://picasaweb.google.c​om/104678127090280812061/
12:54 aloha OK. I'll deliver the message.
13:03 smash joined #parrot
13:03 smash hello everyone
13:06 kid51 joined #parrot
13:07 dalek TT #1919 closed by whiteknight++: Removal of dead functions in src/packfile/api.c
13:07 dalek TT #1919: http://trac.parrot.org/parrot/ticket/1919
13:10 ambs joined #parrot
13:11 kid51 Why is it that since the release the "Revision" entries on our Smolder reports have dropped the git commit number?
13:13 smash kid51++ # release
13:13 kid51 Thanks!
13:13 atrodo kid51> That is the output of "git describe --tags", which will describe a revision with a tag if it is a tag, or describe how far away from a tag it is
13:13 Coke so, if was a release...
13:13 Coke ... it weighs as much as a duck...
13:14 kid51 So the next commit should have us showing a commit number once again, right?
13:14 atrodo kid51> Correct
13:14 atrodo kid51> And git will happily take "RELEASE_3_6_0" as a revision
13:15 kid51 Thanks
13:18 kid51 atrodo: verified: http://smolder.parrot.org/app/​projects/report_details/18243 (smoke of branch)
13:22 whiteknight seen darbelo?
13:22 aloha darbelo was last seen in #parrot 20 hours 46 mins ago joining the channel.
13:27 Felipe kid51++ :)
13:31 dalek parrot: cb9e9b8 | jkeenan++ | / (115 files):
13:31 dalek parrot: Merge branch 'master' into tt1047/type_sizes
13:31 dalek parrot: review: https://github.com/parrot/parrot/commit/cb9e9b8969
13:31 dalek parrot: a5afabb | jkeenan++ | / (2 files):
13:31 dalek parrot: Update Copyright date.
13:31 dalek parrot: review: https://github.com/parrot/parrot/commit/a5afabb4a3
13:40 wagle left #parrot
14:03 wagle joined #parrot
14:03 jay Just did a fresh get and build and install, confirmed that the parrot I'm about to run was installed today:
14:03 jay jay@bayesman:~/Desktop/NQR$ parrot -V
14:03 jay This is Parrot version 3.6.0-devel built for amd64-linux.
14:03 jay Copyright (C) 2001-2011, Parrot Foundation.
14:03 jay Is 3.6.0 really still -devel?  Or is this the new "stable"?
14:04 kid51 Alright, now I'm going to try to understand the 'make docs' part of the release (11f)
14:07 whiteknight kid51: impossible!
14:07 whiteknight jay: If you got it from the git repo, it's -devel
14:07 whiteknight jay: if you got it from the release package, it's not
14:07 kid51 WTF Does this mean?  "Run "make html" in a release copy of parrot, and save the resources/ and html/ directories created in docs/. ssh into the parrotvm"
14:07 jay Ok, git... thanks.
14:08 kid51 Do I do this on my local drive, and then copy them somehow to the server?
14:08 kid51 Or do I ssh into some parrot.org server and do the work there?
14:09 whiteknight kkid51: you need to ssh into the parrotvm, and copy the resources/ and html/ directories to the server
14:10 kid51 whiteknight: Is that done with this?   ssh parrot@ftp-osl.osuosl.org
14:11 whiteknight that's the ftp server where the tarball is uploaded
14:11 kid51 Then, with this?   ssh jkeenan@parrot.org
14:12 whiteknight Do you have ssh access to that box?
14:12 whiteknight if so, yes
14:12 whiteknight if not, you'll need to poke somebody who does
14:13 kid51 Yes, I do.
14:14 kid51 Note to add to docs:  ssh <user>@parrot.org
14:15 kid51 mkdir releases
14:15 kid51 # back at your box
14:15 whiteknight kid51: As you go through this headache, think to yourself "if only I knew an awesome perl5 programmer who could automate most of this work"
14:16 kid51 scp parrot-a.b.c.tar.gz users@parrot.org:releases/
14:16 kid51 # back on parrot.org
14:16 kid51 cd releases/
14:16 kid51 tar xzvf parrot-a.b.c.tar.gz
14:16 kid51 cd a.b.c./
14:16 kid51 perl Configure.pl && make html
14:16 kid51 # that's where I am now
14:17 whiteknight you're building parrot on parrot.org?
14:17 whiteknight oh, you're just building html. Gotcha
14:17 kid51 AFAICT, (a) that's what the instructions say; and (b) make html doesn't entail building parrot
14:17 whiteknight the instructions are clearly not clear
14:18 kid51 That last cd should have been:
14:18 kid51 cd parrot-a.b.c/
14:21 kid51 What is: webroot for docs.parrot.org ?
14:21 whiteknight only one person looks at these instructions, only one day per month. They obviously don't get improved, corrected, or sanity-tested enough
14:21 whiteknight kid51: I don't remember specifically. I assume it's something like /var/www/docs/...
14:23 kid51 whiteknight: /var/www/docs.parrot.org/htdocs/parrot
14:24 whiteknight I was close.
14:25 kid51 Unfortunately, I don't appear to have permissions to make /var/www/docs.parrot.org/htdocs/parrot/3.6.0/
14:25 whiteknight are you a sudoer?
14:25 kid51 drwxrwsr-x 32 coke parrot 4096 Jun 21 07:16 .
14:25 kid51 I have not tried that before
14:26 kid51 No, I cannot sudo.
14:26 kid51 Is there someone who could make me part of the 'parrot' group?
14:28 kid51 Coke or whiteknight: Could you make me part of the 'parrot' group?
14:33 kid51 Well, I guess I'll have to open a support ticket
14:39 whiteknight I have very limited permissions on that box, if I recall correctly
14:39 whiteknight of course, I can't really log in from here anyway
14:43 hercynium joined #parrot
14:49 kid51 This part of the release is such a bear that docs for 3.4.0 apparently never made it to the web.
14:52 whiteknight doesn't surprise me
14:52 whiteknight not many people have ssh access to that box AND the permissions necessary to upload the files
14:56 rblackwe_ is now known as rblackwe
15:01 Coke msg allison licensing question for you - if I use a GPL'd piece of code in my web application without making any modifications to it, have I fulfilled my obligations? If it's JS or CSS, you've already got a copy. If it's server side code, I'm not distributing the code, but running it, and then "distributing" the output.
15:01 aloha OK. I'll deliver the message.
15:02 Coke I can add you to the group.
15:03 allison Coke: yup, you're fine
15:03 allison Coke: this is actually why the AGPL was invented
15:03 allison Coke: because the GPL doesn't have any requirements at all on using the code server-side
15:05 Coke allison++
15:06 Coke kid51: I have root on that box, but no idea what to run.
15:06 Coke whiteknight: you probably have "sudo bash" also.
15:06 Coke I have meetings right now. have to run.
15:07 kid51 Coke: I'll email you what I wrote in support ticket
15:08 kid51 email fwded.  Thanks.
15:32 bluescreen left #parrot
15:47 bluescreen joined #parrot
16:03 darbelo joined #parrot
16:06 benabik Whee...  Re-merged my branch with master and now it won't compile.  I hate tracking down Makefile dependencies.
16:10 benabik Ah.  The Makefile has a default .pir.pbc rule!  And it doesn't have $(PARROT) as a dependency.
16:11 benabik That's fairly useless.
16:15 rurban_ joined #parrot
16:17 rurban left #parrot
16:17 rurban_ is now known as rurban
16:21 dmalcolm joined #parrot
16:21 mj41 left #parrot
16:44 pjcj left #parrot
16:45 dalek rakudo/nom: a331d3c | moritz++ | src/core/Match.pm:
16:45 dalek rakudo/nom: fix Match boolification
16:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a331d3c099
16:46 cotto ~~
16:47 jay left #parrot
16:48 dalek NQR: 78f0898 | (Jay Emerson)++ | / (2 files):
16:48 dalek NQR: Starting to update grammar/actions for proper lvalues
16:48 dalek NQR: review: https://github.com/NQRCore/NQR/commit/78f0898b26
16:48 kid51 left #parrot
16:49 bluescreen left #parrot
16:59 bluescreen joined #parrot
16:59 dukeleto ~~
17:05 dalek website: rohit_nsit08++ | corellaScript : testing and objects
17:05 dalek website: http://www.parrot.org/content/co​rellascript-testing-and-objects
17:05 whiteknight left #parrot
17:24 cotto_work ~~
17:46 Coke benabik: deep in review, but not all makes support giving implicit rules explicit dependencies.
17:46 Coke ... or not so deep.
17:46 benabik Coke: Yeah, I don't think GNU make supports it even...
17:47 benabik Which makes the rule not so useful, since it triggers well before building parrot itself.
17:47 Coke benabik: also, try t/*/checkdepend.pl
17:51 whiteknight joined #parrot
17:51 dukeleto kid51++ # making the release announcment a treasure hunt
17:56 Coke anyone know the command on parrotvm for "add a user to a group" ?
17:56 Coke it's not "vi /etc/group". ;)
17:56 atrodo You beat me to it
17:57 atrodo sudo vi /etc/group ?
17:59 Coke usermod -a -G parrot jkeenan also does nothing.
18:00 pjcj joined #parrot
18:04 Coke msg kid51 I can't add you to the group. Looks like I should do "sudo usermod -G parrot jkeenan", but doing a "groups" jkeenan before and after shows you in no groups.
18:04 aloha OK. I'll deliver the message.
18:11 dalek rakudo/nom: 01b163b | moritz++ | src/core/Str.pm:
18:11 dalek rakudo/nom: implement :g for Str.match
18:11 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/01b163b637
18:13 user_8696 joined #parrot
18:16 benabik Why was one of our goals not setting dogs on fire?
18:22 whiteknight whatnow?
18:23 benabik #ps topic
18:24 whiteknight well, it's very doable as a goal, though not very ambitious
18:24 cotto_work I'm almost certain that I set 0 dogs on fire this week.
18:24 whiteknight I go most weeks without setting any on fire
18:25 whiteknight although there are a few really bad weeks that raise the average
18:28 bluescreen left #parrot
18:37 contingencyplan joined #parrot
18:42 NotFound I must go out for a while, will probably come late to #ps
18:43 user_8696 left #parrot
18:43 user_8696 joined #parrot
18:45 bluescreen joined #parrot
18:47 dukeleto wow, there is a Summer of Code In Space now: http://sophia.estec.esa.int/socis2011/
18:50 whiteknight how did they find enough students up there?
18:55 dukeleto i hear it is easier to herd cats in zero-G, since they way so little
18:55 dukeleto s/way/weigh/
18:56 dukeleto also, english is dumb
18:57 whiteknight english--
18:57 whiteknight We should all speak in winxed
18:57 pmichaud http://gist.github.com/1093407  # latest bench marks
18:57 pmichaud http://gist.github.com/1093408 # latest bench marks, different machine
19:05 cotto_work #ps in 25
19:07 kid51 joined #parrot
19:22 dalek rakudo/nom: 5ee4881 | moritz++ | src/core/ (2 files):
19:22 dalek rakudo/nom: unanchor regexes
19:22 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5ee48814dd
19:23 soh_cah_toa joined #parrot
19:23 pyrimidine joined #parrot
19:24 tcurtis ~~
19:26 dalek rakudo/nom: c2d9fe3 | moritz++ | t/spectest.data:
19:26 dalek rakudo/nom: comment out arrayhash.t again, it only passed by sheer luck before
19:26 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c2d9fe3883
19:29 benabik Hmmmm... newPOST doesn't generate to PIR.  There's some new work to do.
19:32 cotto_work #ps in now
19:35 Coke pmichaud: lower is better?
19:37 pmichaud Coke: lower is better, yes.
19:37 Coke I always look for one of those on a benchmark page, because I invariably guess wrong.
19:37 Coke (well, I guess i varied today0
19:39 whiteknight it's good to see those numbers not increasing. We haven't done anything big this month that would be expected to decrease it
19:56 user_8696 left #parrot
19:57 fperrad left #parrot
20:06 soh_cah_toa aloha: going to fosscon?
20:06 aloha soh_cah_toa: going to fosscon is kid51 or soh_cah_toa or possibly_pmichaud
20:08 pmichaud s/possibly//
20:08 soh_cah_toa good
20:10 soh_cah_toa there is one particular name missing that begins with a "w" and rhymes with "bright light"
20:10 soh_cah_toa ;)
20:11 whiteknight grrr
20:12 soh_cah_toa yay, you got it
20:12 whiteknight do you have any idea how long we've been trying to buy a house? This deal is going through come hell or high water, and the sellers say it is happening during FOSSCON
20:12 soh_cah_toa i know you got lots of house selling stuff going on. i'm joking
20:12 soh_cah_toa sort of ;)
20:13 whiteknight I'm losing my capacity to joke. The hell of realestate is eating away like a worm at my insides
20:14 soh_cah_toa i don't envy you there. i've heard all about the "wonders" of selling a house
20:14 whiteknight i am slowly going crazy yzarc gniog ylwols ma i
20:19 whiteknight is now known as brightlight
20:20 soh_cah_toa ha!
20:21 brightlight I might like this nickname better
20:23 wagle left #parrot
20:24 tadzik wow, you're actually brigther!
20:24 * tadzik screenshots
20:24 tadzik http://i.imgur.com/KjUw1.png
20:24 Coke is now known as LiteBrite
20:24 brightlight hah. Awesome
20:24 LiteBrite is now known as Coke
20:25 tadzik cotto_work: and you got darker :)
20:25 tadzik oopsie, Coke that is
20:25 brightlight is now known as whiteknight
20:29 wagle joined #parrot
20:30 whiteknight left #parrot
20:39 cotto_work pmichaud: are there any Perl6 lists I should send the PDS Doodle to?
20:39 pmichaud perl6-compiler, likely.
20:39 pmichaud We're also planning to have a "module hackathon" on the 30th.
20:40 pmichaud http://rakudo.org/2011/07/1​9/upcoming-rakudo-releases/
20:46 wagle left #parrot
20:54 lichtkind joined #parrot
20:54 cotto_work d'oh.  That message should have said "Q2".  I'll hope that nobody notices.
20:57 lucian left #parrot
21:03 dalek TT #2159 created by Andrew Dougherty <doughera@…>++: OpenBSD 4.9/i386 segfaults (fwd)
21:03 dalek TT #2159: http://trac.parrot.org/parrot/ticket/2159
21:06 cotto_work Mentioning something is a great way to make sure people don't notice it.
21:06 cotto_work doughera++
21:07 wagle joined #parrot
21:07 * Coke is running that on darwin now.
21:08 Coke blah. any trac admins about?
21:09 Coke .. and now the email hits
21:10 cotto_work Coke: what needs doing?
21:10 dalek rakudo/nom-indirect-name-lookup: fcd142d | moritz++ | src/ (3 files):
21:10 dalek rakudo/nom-indirect-name-lookup: first attempt at indirect name loookup.
21:10 dalek rakudo/nom-indirect-name-lookup:
21:10 dalek rakudo/nom-indirect-name-lookup: Does not work yet, method term:sym<name> in Actions.pm needs to learn
21:10 dalek rakudo/nom-indirect-name-lookup: not to touch indirect name lookups first
21:10 dalek rakudo/nom-indirect-name-lookup: review: https://github.com/rakudo/rakudo/commit/fcd142d9ce
21:16 dalek parrot: e0231e5 | doughera++ | src/gc/fixed_allocator.c:
21:16 dalek parrot: avoids sign extension errors by using ptrdiff_t
21:16 dalek parrot:
21:16 dalek parrot: Trac # 2159
21:16 dalek parrot: review: https://github.com/parrot/parrot/commit/e0231e52b4
21:20 Coke cotto_work: nothing now
21:21 Coke hey, look at that, I made the patch come from andy.
21:21 bluescreen left #parrot
21:27 ambs left #parrot
21:37 bluescreen joined #parrot
21:45 bluescreen left #parrot
21:48 Patterner left #parrot
21:48 Psyche^ joined #parrot
21:49 Psyche^ is now known as Patterner
22:01 hercynium left #parrot
22:11 Coke I see a LOT of 'git' is not recognized as an internal or external command,
22:11 Coke operable program or batch file.
22:12 Coke when running 'nmake smoke'. I think we're assuming that if -d .git, we can `git`, which ain't the case.
22:13 dukeleto Coke: yes, our config/build system makes that wrong assumption
22:15 Kulag left #parrot
22:15 Kulag joined #parrot
22:18 Coke dukeleto: #2160
22:22 dalek TT #2159 closed by coke++: OpenBSD 4.9/i386 segfaults (fwd)
22:22 dalek TT #2159: http://trac.parrot.org/parrot/ticket/2159
22:22 dalek TT #2160 created by coke++: .git doesn't imply git executable
22:22 dalek TT #2160: http://trac.parrot.org/parrot/ticket/2160
22:23 kid51 Alright, I doodled!
22:29 jsut joined #parrot
22:34 jsut_ left #parrot
22:43 cotto_work all your doodles are belong to parrot
22:50 Kulag left #parrot
22:54 Kulag joined #parrot
23:02 dmalcolm left #parrot
23:15 Felipe hehe
23:33 khisanth_ joined #parrot
23:33 Khisanth left #parrot
23:33 darbelo left #parrot
23:34 khisanth_ is now known as Khisanth
23:35 bubaflub joined #parrot
23:36 bubaflub ~
23:40 dalek parrot: 0d3a111 | jkeenan++ | docs/project/release_manager_guide.pod:
23:40 dalek parrot: Modify on basis of 3.6.0 experience, particularly re server permissions, make reconfig, etc.  Some POD reformatting.
23:40 dalek parrot: review: https://github.com/parrot/parrot/commit/0d3a111dc3
23:40 kid51 msg cotto Am committing to docs/project/release_manager_guide.pod.  Please review -- though I have more edits to make.
23:40 aloha OK. I'll deliver the message.
23:40 kid51 left #parrot
23:45 smash left #parrot
23:46 cotto_work msg kid51 Is there any reason for "RELEASE MANAGER GUIDE" to be all caps?  iirc allison figured out that the only reason we do that was because man pages did it.
23:46 aloha OK. I'll deliver the message.
23:48 cotto_work I like the idea of splitting that file into two parts; the first being an introductory section for first-time release managers and the being a more sparse version that assumes the knowledge in the introduction.
23:52 cotto_work that may be overdoing it

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced