Parrot 3.7.0 "Wanda" | parrot.org | Log: irclog.perlgeek.de/parrot/today | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 6 September 2011.
whiteknight good evening, #parrot 00:19
dalek sella/harness_test: 321f1f6 | Whiteknight++ | src/harness/ (6 files):
Cleanup Harness files and add documentation
sella/harness_test: 9f20a03 | Whiteknight++ | t/harness:
Don't need to explicitly get the view in t/harness
sella/harness_test: 437bedf | Whiteknight++ | t/harness:
Update t/harness to recognize that tap_harness tests are winxed
sella/harness_test: 3eaff10 | Whiteknight++ | src/harness/TestExecutor.winxed:
Catch exceptions thrown by the parser and properly mark the test as aborted.
cotto_work chromatic++ # "His plan is more conservative than my idealism likes, which means it's probably a wise approach." 00:25
bubaflub whiteknight: how goes the housing situation? 00:45
00:52 patspam joined 01:05 patspam joined 01:06 whiteknight joined
whiteknight AARG! I hate Class PMC 01:06
hate it
I override find_method to return a closure trampoline. Class helpfully caches that 01:07
so when I call methods of the same name on a separate object, the method references the original instance
hatehatehatehate
so all I need to do is jam a call to class.clear_method_cache() inside every single method invoke in my class 01:16
I need somebody to help hold Parrot down for a minute. I'm going to get the knife
soh_cah_toa_ no fair, you hold it down. i get the knife ;) 01:17
dalek sella/harness_test: 1093087 | Whiteknight++ | src/ (3 files):
MockObject method closures are being cached, along with the list of expectations. Multple mocks of the same type, might share expectations. Add a method to get the proxy class object from Proxy.Factory. In MockObject.Factory, give a reference to it to the mock Controller. When invoking a method on a mock, explicitly clear the method cache.
01:21
sella/harness_test: 7d62cce | Whiteknight++ | src/harness/Harness.winxed:
In Harness, for backwards compatibility we need to be able to take a single test in addition to an array of them
sella/harness_test: f1cea2a | Whiteknight++ | / (2 files):
Some cleanups and error handling in TapParser. More tests
01:23 JimmyZ joined
whiteknight soh_cah_toa_: you have any idea how frustrating it is, when I'm doing testing with mockobjects and my tests are failing for no apparent reason 01:25
Coke whiteknight: I think everyone in this channel knows how frustrating parrot failing is. ;) 01:30
soh_cah_toa_ it gets even better when the results are something like "got: foo, expected: foo"
Coke tries to do some work on his tap server.
cotto ~~ 01:31
soh_cah_toa_, then you edit the output with hexedit and find that there's a non-printing character messing it up 01:32
or better, there's not 01:33
whiteknight soh_cah_toa_: that's exactly what it was
soh_cah_toa_ cotto: ah, never thought about trying that
dalek sella/harness_test: dca4e9b | Whiteknight++ | t/tap_harness/Harness/TapParser.t:
Several more tests for Harness.TapParser
01:34
03:21 schmooster joined
Coke I pushed out to github.com/coke/muddle - just a shell at the moment, will eventually turn into the smolder-like-thing. 03:22
github.com/coke/muddle/blob/master...mpledb.sql is probably the most interesting file at the moment. 03:23
will hack as time permits. which is not now. zzz
03:52 schmooster joined 04:10 fperrad joined 05:13 contingencyplan joined 05:23 SHODAN joined 05:43 woosley joined 07:26 mj41 joined 07:42 pjcj joined 08:06 wagle joined 08:48 jsut joined
dalek kudo/nom: ecce797 | moritz++ | src/core/Parcel.pm:
Parcel.fmt
08:53
09:37 ixnay joined
ixnay Good timezone, people 09:37
moritz good zonetime, ixnay 09:38
dalek kudo/nom: d97997e | moritz++ | src/core/Enum (2 files):
implement .fmt in Enum and EnumMap
09:39
ixnay so, this is a channel dedicated to parrot birds, not something else parrot, corret? 09:41
moritz ixnay: it's dedicated to a virtual machine called "parrot" 09:42
ixnay awh, thats a downer :P
moritz it's the reason for me to be here :-)
though parrot birds are cute too
ixnay yeah, I'm getting one and looking for advice and such
moritz then this isn't a good place for you, I fear 09:43
ixnay anyhow, I'll continue my search for a bird-dedicated channel
moritz try irc.freenode.org or so
ixnay Have fun, people, whatever it is you do :P
yeah, I'm googling =)
moritz irc.perl.org is very programming specific
ixnay Bye =)
moritz ixnay: have fun
09:43 ixnay left 10:28 Coke joined 10:55 mj41 joined 11:14 schmooster joined 11:35 Psyche^ joined 11:38 benabik joined
benabik o/ 11:54
12:00 woosley left
Coke . 12:13
12:27 whiteknight joined
whiteknight good morning, #parrot 12:27
tadzik good morning whiteknight 12:28
whiteknight hello tadzik 12:29
12:33 bluescreen joined
benabik o/ whiteknight 12:38
12:39 ambs joined
whiteknight good morning, benabik 12:40
benabik whiteknight: How was your weekend? I got distracted by a wedding. :-)
whiteknight benabik: alright. Mostly relaxing.
benabik Relaxing is good.
Coke benabik: your own? 12:54
benabik Pretty good. Saw a bunch of family. Weddings are useful for that. :-D
moritz hopes not benabik's own, because it's a bad sign when you see your own wedding as a distraction :-) 12:55
benabik moritz: Mine was years back. :-D If my wedding was this weekend, my wife would kill me for being on IRC right now. Honeymoons are not times for chat rooms.
moritz wise words 12:56
12:58 JimmyZ joined
dalek rrot/whiteknight/kill_threads: 9280de6 | jimmy++ | src/interp/inter_cb.c:
Merge pull request #163 from niner/whiteknight/kill_threads

Whiteknight/kill threads
13:02
13:10 Coke joined
dalek sella: 8a9c108 | Whiteknight++ | src/harness/testfile/PIR.winxed:
Try to fix a weird precidence error
13:18
sella: 684e19b | Whiteknight++ | src/harness/Harness.winxed:
Add in a Harness method that will be coming on a branch, to help with forward compatibility
sella: 2c0ec08 | Whiteknight++ | src/harness/testfile/PIR.winxed:
Fix TestFile.PIR, it was reading the wrong executable name. I probably need to grab this from the config hash instead
13:44
13:50 contingencyplan joined 13:57 bluescreen joined
dalek Heuristic branch merge: pushed 47 commits to winxed/dyns_opt by Whiteknight 14:36
whiteknight just got the laziest, least compelling cold call from a recruiter ever 14:59
despite the fact that the economy is in the shitter and the unemployment rate is phenominal, I'm still getting recruitment calls on a daily basis 15:00
15:00 Coke joined
benabik Unemployment in the computer field is very low, especially in some areas of the country. 15:00
Also, some of it is "hey, unemployment is high, let's cycle through people until we get some we like" 15:01
whiteknight That's what I've been telling people for years: if you want a job, go into the technology fields and be good at what you do 15:04
moritz the last half is the hard half
whiteknight like, if you're a teacher and you're great at what you do you can still be fired if there are budget cuts and you're low on the seniority list
cotto whiteknight, not the least competent: seldo.tumblr.com/post/10140794798/t...t-to-do-it 15:12
also, ~~
Coke whiteknight: that happens in our district frequently. low seniority teachers have been fired and hired back. multiple teachers, multiple instances per teacher. 15:13
it's crazy.
whiteknight cotto: okay, that one would be worse 15:14
15:18 davidfetter joined
dukeleto parrotvm is running out of inodes again because of smolder 15:21
15:21 dmalcolm joined
cotto dukeleto, cron script? 15:21
there's no strong reason to keep those reports around for very long 15:22
cotto heads to $dayjob
15:24 soh_cah_toa joined
dukeleto i don't want to touch smolder. 15:25
anybody want to write a cronjob for parrotvm for smolder? I can't volunteer for it.
soh_cah_toa just an fyi, i'm gonna be a little late to #ps today b/c of my physics class 15:39
i was planning on just using my laptop in the library but i totally borked it last night while screwing around w/ $PATH - now i can't login :( 15:40
normally, being late wouldn't really matter but i just didn't want anyone to think i was abandoning my release manager duties
off to class...
dukeleto Ramereth++ in #osuosl just wrote a cron to delete smolder reports >= 2weeks old, which normalized our inode usage. Thank him if you get a chance.
whiteknight dukeleto: did you see my libgit2 pull request this morning? 15:42
I made some fixes to Rosella master. I figured out why your pir tests were crapping. It was my fault 15:43
dukeleto whiteknight: i didn't! 15:45
dukeleto loves pull requests in the morning 15:46
Coke dukeleto: I don't think restarting the instance frees the indoes. 15:47
I'm pretty sure you have to do ugly find -?time magic and find the oldest files and delete them.
I will work faster on getting it shut off.
question: does anyone ever download the .tar.gaz reports /from/ smolder?
dukeleto Coke: that is what Ramereth did. see recent parrot-dev email 15:48
Coke *gz
dukeleto: perfect.
I will endeavor to get things running faster. hard this week due to dayjob deadlines.
dukeleto whiteknight: i see that your pull request changes the namespace in the nci file, and removes on the functions in the api, most probably because you had a different version of libgit2 installed 15:49
Coke no one responded to the smolder email, btw. So You'll get what I give ya, and be happy with it. ;)
whiteknight oh, I didn't think I committed those files
Coke sorry, kid51++ # actually already did stuff.
whiteknight at least, I tried not to
dukeleto Coke: also, you don't have parrotvm access anymore 15:50
Coke dukeleto: no, but I can gets it. Just glad I don't have to! ;)
dukeleto whiteknight: no worries, thanks for the very useful pull request 15:51
whiteknight dukeleto: I have a pull request in to Winxed to add a -X commandline option so we can fix those tests
it can't find libgit2 on my system without -X/usr/local/lib
cotto_work ~~ 15:52
dukeleto whiteknight: interesting. winxed somehow finds libgit2 on my machine 16:03
whiteknight: mine is installed in /usr/local/lib as well 16:07
whiteknight maybe I just don't have it set up right
cotto_work Could it be a configuration issue? 16:14
dukeleto whiteknight: what kind of machine are you on? perhaps /usr/local/lib is not in your ld.so.conf ? (unlikely) 16:17
16:23 sgt joined
whiteknight dukeleto: x86-64 ubuntu vm 16:24
I may have purposefully screwed up a setting that I don't rememer. It's no big deal 16:25
16:29 eternaleye joined 16:35 bluescreen joined
cotto_work #ps in 2:47 16:42
17:00 HSlayer joined
NotFound whiteknight: Weird precedence error? 17:01
whiteknight NotFound: seemed weird to me. Could just be the code. I ripped it all out anyway 17:02
NotFound whiteknight: string s = some[thing] resulting in get_string_keyed... is completely intentional. 17:03
whiteknight yeah, I figure that. I thought that the Parrot vtables would fall back
NotFound Ah, yes, maybe ParrotInterpreter should implement it. 17:04
17:09 fperrad joined
whiteknight I need to prototype some of the new ops for PCC changes 17:40
17:42 dukeleto joined 18:09 fperrad joined 18:10 zby_home joined, davidfetter joined
dalek nxed: c74f8ed | NotFound++ | winxed (2 files):
add -X command line option - Issue 11, whiteknight++

Also refactor a bit the installed driver
18:11
whiteknight NotFound++
dalek nxed: 0de3efd | NotFound++ | pir/winxed_installed.pir:
update generated driver
18:13
rrot: 405ca66 | NotFound++ | ext/winxed/driver.pir:
update winxed snapshot to 0de3efd454
18:24
18:29 mj41 joined 18:35 AzureStone joined 19:09 Coke joined 19:25 tcurtis joined
tcurtis ~~ 19:26
whiteknight hello tcurtis 19:27
tcurtis hello whiteknight
whiteknight I've been digging through lalrskate again. It's very nice code 19:28
tcurtis whiteknight: thank you, though I don't fully agree. Some of the transformations are rather ugly. 19:30
cotto_work #ps at now o'clock 19:31
19:36 benabik joined 19:41 schmooster joined
benabik Wow, empathy is really annoying to use in some ways. It wants to grow the chat window to the full width of my screen no matter what I set it to... The text cursor is still 100px away from the edge and it keeps growing and growing... Past the edge of the screen even! 19:42
19:50 preflex_ joined 20:21 soh_cah_toa joined
Coke cotto_work: my question was "you don't want to warnock the smolder thread, right?" 20:38
cotto_work Coke: certainly not 20:40
I'm quite happy to have smolder only retain the last 2 weeks (or less) of reports. 20:42
Coke ... wow. no, MY smolder thread. 20:45
the one called... "smolder". ;) 20:46
bubaflub Coke: i'd love to play with your new smolder client / help ya out, could you add a roadmap or maybe some github issues? 20:58
cotto_work Coke: sorry. I read too quickly. 21:01
Coke bubaflub: ... nicely done :P 21:12
Yes, I'll do that. at this point, feedback on the db schema would be helpful. 21:13
github.com/coke/muddle/blob/master...mpledb.sql
hurm. that's behind.
pushed. 21:14
basic roadmap - add file upload capability, so we can get a bunch of data from existing test runs. add front end pages to view either summary info about the project, summary about the run, particular information about a test (in a run or over time.) 21:16
bubaflub Coke: yeah, i don't suppose we need a dedicated users table with different auth levels and what not 21:21
Coke: or rather, later we can transition to something like that if need be
Coke: upload via FTP or HTTP POST or some other means?
Coke http post 21:24
pretty much the same as current smolder.
bubaflub ok, makes sense. i'll be a bit busy this week but i should have some time to add docs, tests, ask annoying questions, etc.
Coke (so, probably a simple, public user/pass for the submittal, just to add a roadbump.)
\\o/
21:36 Drossel joined 21:47 benabik joined 22:02 Limbic_Region joined
cotto_work anyone here on windows x64? 22:19
dalek kudo/nom: cca1166 | jnthn++ | t/spectest.data:
Label a few more tests as needing ICU.
22:27
rrot/win64-workaround: 0950a00 | cotto++ | / (2 files):
first try to force sizeof(void*) == sizeof(INTVAL) on 64-bit windows
22:29
23:01 bubaflub joined 23:27 aloha joined 23:52 whiteknight joined
cotto_work whiteknight: can you poke win64-workaround with a stick? It needs some fiddling to figure out how to get the right information into the right place at the right time. The goal is to automagically set sizeof(intval) == sizeof(void*) == long long on msvc/x64. 23:54
cotto_work heads home
whiteknight okay I'll look tomorrow
I wonder, is the time better spent forcing the two values to be the same size, or fixing Parrot to not assume they are in the first place? 23:56