Parrot 3.0.0 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Goals: Deprecations as data, merge/cleanup, AutomatedHLLTesting, clean up false test failures
Set by moderator on 20 January 2011.
cotto_work whiteknight: how much effort would it be to remove that indirection? 00:03
whiteknight not too hard. the hard part is doing it without making bacek sad
:)
chromatic I think we say "Every GC must implement these APIs." Everything else you do is up to you. 00:04
whiteknight I think the first thing we can do is replace the indirect calls with macros. Then, we just change what the macros are at config time 00:05
00:06 dmalcolm left
chromatic Isn't that still a layer of function calls? 00:06
whiteknight We can do what we do with the context accessors 00:07
have functions and macros which look like functions.
if NDEBUG, the macros are just macros, but with DEBUG they redirect to functions
00:08 thundergnat left
chromatic Some of these functions need PARROT_API though. 00:08
00:23 hercynium left
dalek rrot: 9e06e07 | mikehh++ | MANIFEST (2 files):
re-generate MANIFEST and MANIFEST.SKIP
 00:24
00:29 plobsing joined 00:38 cosimo joined
whiteknight chromatic: definitely not PARROT_API. PARROT_EXPORT maybe 00:44
(I mean, the two are #defined the same, but they have different meanings) 00:45
chromatic Hopefully no extensions call into the GC directly?
whiteknight jnthn has been using the fixed-size allocator for 6model 00:46
it's faster most of the time than malloc
but otherwise, nothing should be using GC directly
01:06 bubaflub joined
bubaflub hi dukeleto 01:06
particle free the mallocs! 01:14
01:17 Coke left
bacek_at_work I don't mind switching GC selection to compile-time. 01:21
It will not speed-up parrot dramatically though.
Our PCC is extremely slow. 01:22
01:27 Coke joined 01:44 kid51 joined
cotto_work Nothing bad seems to happen if I uncomment PackFile_destroy in the Eval PMC. Does this mean that trac.parrot.org/parrot/ticket/1292 can be closed? 01:54
plobsing probably 01:55
01:56 bubaflub left
plobsing evals aren't merged into initial_pf like other bytecode and any dependant subs should reference the Eval PMC 01:57
kid51 Slow day: No change in any tickets in past 12 hours. I got no Parrot mail since 700 this morning.
bacek_at_work cotto_work, nope. Comment is correct. We can't destroy used PackFile 02:00
cotto_work bacek_at_work: make test passes. What's a good indicator that the ticket can be closed, apart from pervasive use of the Packfile PMCs?
02:01 contingencyplan left
bacek_at_work cotto_work, looks like we've lost destroying of main interp. 02:04
cotto_work That'd explain why VTABLE_destroy on some PMCs isn't covered. 02:05
The use of ATEXIT_DESTROY looks backwards. 02:07
Coke kid51: not counting commit msgs? 02:08
cotto_work It's been like that for years. Did we ever have interp destruction?
chromatic Yes we did. 02:10
cotto_work nm. really_destroy is called. I just misread how it was called. 02:15
02:17 whiteknight left
kid51 How would you describe 'miniparrot' in 6-8 words? (i.e., in the 'make help' output) 02:20
Coke A relic of an older time. 02:22
kid51 Heh
Coke bare minimum parrot executable?
cotto_work a parrot binary with no configuration data
kid51 If so, does it merit inclusion in 'make help'? Is there any reason for someone to call it on the command-line?
'make miniparrot', that is 02:23
cotto_work I'm doubtful
kid51 fine, we already have 63+ lines in the help!
Coke not really.
it's very frustrating to check rakudo when it takes like 10m to build a fresh parrot+rakuod. 02:24
kid51 Good, then I'm just about ready to close this ticket
cotto_work kid51: great
02:27 mtk left 02:31 mtk joined
dalek TT #1922 closed by jkeenan++: Review Parrot's 'make help' 02:46
TT #1922: trac.parrot.org/parrot/ticket/1922
03:05 plobsing1 joined 03:10 plobsing left 03:15 theory left
dalek TT #1392 closed by jkeenan++: build parrot failed on CentOS 4.5 03:18
TT #1392: trac.parrot.org/parrot/ticket/1392
03:38 kid51 left 04:20 theory joined 04:40 bacek left 05:11 contingencyplan joined 05:15 p6eval left 05:16 sorear left 05:17 TimToady left, dalek left 05:28 p6eval joined, sorear joined, workbench left, dngor left 05:29 TimToady joined, dalek joined 05:33 nwellnhof_ joined 05:36 dngor joined, nwellnhof left, nwellnhof_ is now known as nwellnhof 05:39 dalek left 05:41 dalek joined 05:42 Patterner left 05:44 dalek left 05:45 Psyche^ joined, Psyche^ is now known as Patterner 05:50 rurban_ joined 05:51 p6eval left 05:53 rurban left 05:54 rurban_ is now known as rurban
dukeleto ~~ 05:55
06:27 nwellnhof left 06:31 theory left 06:32 chromatic left 06:55 dalek joined, dduncan_ joined 06:56 dduncan_ left, p6eval joined 07:01 cogno joined 07:02 theory joined 07:06 theory left
cotto bacek_at_work, ping 07:19
msg bacek_at_work Can you add your availability to doodle.com/4phzuaiahkgd3ie5 for the PDS? 07:22
aloha OK. I'll deliver the message.
07:23 cogno left 07:28 cogno joined 07:33 particle left 07:35 allison left
dalek rrot: 9b2a899 | jimmy++ | include/parrot/platform_interface.h:
added missing codes after branch mergence
 07:35
07:35 allison joined 07:53 cogno left 08:09 cogno joined 08:19 fperrad joined 09:10 cogno left 09:20 jjore left 09:21 jjore joined
mj41 osuosl.org/about/news/supercell-announcement 09:54
10:13 mtk left 10:19 mtk joined
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#4380) fulltest) at 3_0_0-151-g9b2a899 - Ubuntu 10.10 i386 (g++-4.5) 11:10
11:11 ambs joined 11:25 ambs left 11:31 JimmyZ joined 11:33 ambs joined
dalek tracwiki: v148 | gerd++ | Languages 11:38
tracwiki: trac.parrot.org/parrot/wiki/Languag...ction=diff
11:58 cogno joined 12:20 cogno left 12:21 cogno joined 12:46 JimmyZ left 13:12 cogno left 13:27 cogno joined 13:39 whiteknight joined 13:43 cogno left
whiteknight good morning, #parrot 13:47
13:50 rurban_ joined 13:51 simcop2387 left 13:52 rurban left, rurban_ is now known as rurban 14:01 simcop2387 joined
dalek rrot: 16acf9c | Whiteknight++ | NEWS:
update NEWS to mention that Exceptions are subclassable
 14:04
whiteknight msg nwellnhof when you get to a good point can you update NEWS with info about the stuff you've merged? 14:06
aloha OK. I'll deliver the message.
14:15 plobsing1 left
dalek TT #1985 created by whiteknight++: Add windows-related temporary files to .gitignore 14:17
TT #1985: trac.parrot.org/parrot/ticket/1985
rrot: 53d72c1 | Whiteknight++ | docs/project/release_manager_guide.pod:
gerd++ has 3.4 upon request
whiteknight msg dukeleto I'm going to delete the exceptions_merge_refactor branch. I've merged in a different branch this week which makes exceptions subclassable 14:26
aloha OK. I'll deliver the message.
14:31 particle joined 14:34 JimmyZ joined
JimmyZ good morning whiteknight 14:34
I am confused between ' Exceptions are subclassable' and 'Exception PMCs are not subclassable from PIR' 14:35
14:38 TiMBuS left 14:39 TiMBuS joined
Coke you can write subclasses in C. 14:41
well, in .pmc 14:42
whiteknight jimmyZ: Exceptions are subclassable now from PIR 14:43
that happened in a branch I merged this week
JimmyZ well, the message is 'are' and the content is 'are not'
whiteknight JimmyZ: Where do you see that? Everything might not be updated 14:44
JimmyZ whiteknight: oh, then I think that's a typo
whiteknight probably
JimmyZ whiteknight: github.com/parrot/parrot/commit/16acf9c692
Coke probably meant "are now" 14:45
common english typo.
JimmyZ hehe :)
whiteknight my english is not so good 14:46
JimmyZ++ for finding my mistake
JimmyZ my english is poor :(
dalek rrot: ee5be9c | Whiteknight++ | NEWS:
Fix typo discovered by JimmyZ++
 14:47
14:47 mtk left
whiteknight JimmyZ: Good enough to find my mistakes! 14:47
JimmyZ just confused, hehe
14:52 mtk joined 14:57 mtk left 15:02 mtk joined 15:25 plobsing joined 15:34 plobsing left 15:37 cognominal joined
particle if you haven't seen osuosl.org/about/news/supercell-announcement already, facebook is sponsoring a testing lab at ousosl for oss. i've already submitted feedback at supercell.osuosl.org/feedback for parrot to participate. 15:38
cotto particle++ 15:41
15:44 plobsing_ joined 15:46 szabgab left
whiteknight wow. awesome 15:46
15:53 davidfetter joined 16:00 cognominal left 16:02 szabgab joined 16:05 JimmyZ left
cotto_work ~~ 16:10
whiteknight goodmorning cotto_work 16:11
cotto_work hi whiteknight
tadzik hello parrots 16:12
whiteknight hello tadzik 16:13
tadzik: how are classes going?
cotto_work hello tadziks
tadzik whiteknight: good and bad. The phisycs is alredy doomed, but there is still hope for electronics, esp. after todays labs
whiteknight tadzik: ah yes. I remember physics. It was an extremely tough course 16:14
I didn't really understand all that stuff about waves and optics and electromagnetics until after I graduated 16:15
and now that I understand all that stuff, finally, I have a career that doesn't use any of it :(
tadzik ah 16:22
16:32 bluescreen joined 16:33 theory joined 16:35 davidfetter left 16:36 dmalcolm joined
Coke aloha, neicza? 16:37
aloha Coke: Sorry, I don't know.
16:39 dip left
NotFound whiteknight: but you can use an optic mouse with more authority ;) 16:43
whiteknight NotFound: :) 16:48
16:59 bluescreen left 17:28 cogno joined 17:49 plobsing_ left 17:57 mtk left 17:58 mtk joined, mtk left
cotto_work It looks like the code to call Parrot::Manifest::print_gitignore is never called. 17:58
17:59 mtk joined
cotto_work I'm pretty sure it's highly broken. 18:00
whiteknight awesome 18:01
I will edit .gitignore myself if I didn't think the changes would be overwritten 18:02
cotto_work whiteknight: go for it 18:04
I also doubt it's worth fixing the code to automatically generate .gitignore.
dukeleto ~~ 18:05
.gitignore is not auto-generated
LIES LIES LIES
cotto_work you can also delete the scary "generate by" line
and set it to not be ro
dukeleto: is there any value in keeping the gitignore-related perl code in Parrot::Manifest? 18:06
dukeleto hmmm, maybe it is auto-generated 18:07
cotto_work dukeleto: it's not
If you use the generating code, it's pretty thoroughly incorrect.
dukeleto cotto_work: that is what i vaguely remember 18:08
cotto_work: i think .gitignore should be manually edited and any code that used to generate it should get deleted 18:09
cotto_work +1
dukeleto cotto_work: i think the code that generates .gitignore is from SVN days, and is most likely worthless 18:10
particle: thanks for getting on the supercell train :)
whiteknight: exceptions are subclassable now?!?
sorear .gitignore used to be automatically generated from svn:ignore 18:11
this no longer makes sense
dukeleto sorear: yes 18:13
cotto_work: shall I take the axe to the .gitignore-generating code?
cotto_work dukeleto: sounds like there's no reason to keep it around 18:14
18:14 mtk left 18:15 mtk joined
dukeleto cotto_work: this code is hilarious. All it does it cat .ignore, remove comments, and put it back into .gitignore 18:15
lulz
cotto_work It's nice to have hilarious code that we can safely excise. 18:16
cotto_work hands dukeleto a scalpel
dukeleto hmmm, but the manifest.skip code uses the gitignore code. blarg 18:17
dukeleto sees the light 18:20
cotto_work turns off the flashlight
tadzik hello again 18:24
18:27 chromatic joined
dukeleto tadzik: wazzup 18:30
whiteknight dukeleto: to the best of my knowledge, Exceptions are now completely subclassable 18:32
dukeleto: ExceptionHandlers are still not
tadzik dukeleto: the crazy week's over, whew! 18:34
18:36 cogno left
tadzik I have some time to look at Deprecations now 18:37
cotto_work tadzik: great! 18:38
dalek rrot: 3f5d600 | dukeleto++ | / (3 files):
[tools] Remove ancient/dumb .gitignore-generating code

This code hasn't done anything useful since we switched to Git. Everyone can now add stuff to .gitignore as they see fit.		 18:40
tadzik my first idea was that they should maybe be organized using tags, rather than categories we have now. This way some deprecation could be assigned to many categories easily, what do you think?
if there's a need for it of course, not sure, hence I'm asking you 18:41
whiteknight dukeleto: if I do "*.obj" in .gitignore, that will cover all .obj files in the repo, correct?
dukeleto: as opposed to "/*.obj", which would only do it for the root directory?
dalek rrot: b87413e | Whiteknight++ | .gitignore:
ignore windows-specific build files and temporaries in .gitignore
 18:44
dukeleto whiteknight: yes, i believe so 18:45
18:45 ambs left
dukeleto tadzik: what are the current categories? 18:45
tadzik dukeleto: PMCS, Opcodes, Bytecode, Debugger, PIR syntax, Functions, Compiler tools (has subcategories), Parrot library and Experimental 18:47
18:48 cogno joined
dukeleto tadzik: i like the idea of tags 18:49
tadzik: so that a single deprecation can be given multiple tags
tadzik: such as "PIR syntax" and "PMCs" or something like that 18:50
tadzik: for the first iterations of your script, i would say that convert all previous categories into a tag, and then we can add more tags in the future to deprecations
tadzik: and document somewhere that a deprecation can have multiple tags 18:51
tadzik sure thing, the change will be documented, maybe I'll dedicate a wiki page to it
18:53 nwellnhof joined
nwellnhof ~ 18:53
tadzik o/ 18:56
dukeleto Looks like rfw got a part time job out of GCI: sahanadev.wordpress.com/2011/01/20/...e-code-in/ 19:02
that is also a very nice description of what went wrong during GCI
19:09 hercynium joined
cotto_work dukeleto: nice find 19:11
where'd you see that?
dukeleto cotto_work: the twitterwebs 19:12
cotto_work lols at rfw being a gsoc mentor because he's too young to be a student
dukeleto cotto_work: yep, that has happened a few times
cotto_work: i met a GSoC mentor 2 years ago that was *13* 19:13
cotto_work must be a smart kid 19:14
19:16 cogno left
cotto_work seen rfw 19:17
aloha rfw was last seen in #parrot 11 days 11 hours ago saying "i thought i scared you".
nwellnhof but i think we can say that gci was an overwhelming success for parrot. 19:19
19:19 Kristaba joined 19:21 Eclesia joined
Eclesia hi 19:21
sorear hello.
cotto_work nwellnhof: yes. overwhelming and successful
nwellnhof i'm hacking on the OS dynpmc right now and its test coverage got increased to 94.1% from 30.2% by the gci students. 19:22
dalek rrot/nwellnhof/unicode_dynpmcs: 3312ddd | nwellnhof++ | / (7 files):
Implement some file functions with Unicode support

getcwd, mkdir, chdir, rmdir, unlink		 19:23
dukeleto nwellnhof: yes, GCI was a hug success. Many of our PMCs went from abysmal to >90% coverage 19:32
huge success, even
cotto_work nwellnhof++ for that commit 19:37
whiteknight nwellnhof: did you merge that imcc branch thingy you were working on? 19:43
or, unicode_filenames I think it was
nwellnhof whiteknight: not yet. i hit another issue yesterday, but that should be resolved now. 19:44
it didn't get much testing, but it shouldn't be a problem to merge it at that point of the development cycle. 19:45
cotto_work nope. We've got lots of time for testing between now and the next release. 19:46
whiteknight: what's your plan for NotFound's rethrow bug? 19:48
whiteknight cotto_work: no plan. It isn't a bug 19:49
NotFound cotto_work: The problem was found by plobsing, not me.
whiteknight it's behavior by design. We can talk about the sanity of the design
cotto_work ok
whiteknight rethrow creates an iterator for the handlers that are active at the time of the rethrow 19:50
cotto_work Is it something that's likely to come up later? It looked surprising as far as I could grok the test.
whiteknight if we rethrow then add more handlers to the context, then rethrow again, the second rethrow won't see the newly added handlers because they aren't in it's iterator
it took me a very long time to understand as well. I'm not sure I like the behavior, but I'm at a loss about how to reimplement things 19:51
cotto_work We should at least document the weirdness. 19:52
Thanks for looking into that.
dukeleto: is there a nice way to find which branch a commit was made to through github's web interface? 19:53
whiteknight cotto_work: probably not once the branch has been merged 19:54
20:03 particle1 joined 20:06 particle left 20:07 cognominal joined 20:10 particle1 is now known as particle 20:11 plobsing joined
tadzik wklej.org/id/461661/ -- how does that look? 20:12
needs some cleanups of course, but is that what you expected? 20:13
cotto_work tadzik: it's a good start. 20:15
The urge to expand it is very strong.
I'd add some fields so that the affected PMCs, ops and functions are machine-readable 20:16
also fields for the data included at trac.parrot.org/parrot/wiki/ParrotD...ionsFor3.0 20:17
tadzik well, for now I'm thinking about the format
the bad side is that some deprecations got lost: there are 53 entries, while there were 57 in DEPRECATED.pod 20:18
cotto_work as long as you don't have to write a new parser and there aren't too many gotchas about the format, wfm
we can figure that out later
tadzik oh well 20:19
cotto_work If it were me I'd just pick 5 or so representative deprecations to avoid having to edit a long list 20:20
tadzik the actual parser is so bloody stupid I'm not even suprised: it's actually parsing pod2text output line by line, matching the strings depending on their indentation
whiteknight we probably want to separate out experimental items into a separate data set
cotto_work while prototyping
which parser?
tadzik I just couldn'd find any module for Parsing pod which would be easier to work with than this
the one I wrote 20:21
dukeleto tadzik: you making progress on DaD ?
whiteknight tadzik++
tadzik dukeleto: yeah, somewhat
20:22 cognominal left
whiteknight oooh, we could easily have a test that reads trac and makes sure all tickets listed in that structure are open and are set to DEPRECATION 20:22
tadzik prr, let's have a proper list first
cotto_work whiteknight: now you're thinking with data
whiteknight and the reverse, we could have a test that proves all deprecation tickets in trac are represented in that structure
ZOMG DATA!
tadzik I need to figure out the four items that are lost 20:23
dukeleto tadzik: awesome. just manually add those to the DEPRECATED.JSON, and we have DaD !
whiteknight what is this, JSON?
dukeleto tadzik: the script was just to bootstrap us. Now we need a script to read DEPRECATED.JSON and spit out a nice human-readable description 20:24
whiteknight: yep
whiteknight with this, can we get rid of DEPRECATED.pod?
dukeleto tadzik: also, can you lower case all the tags?
cotto_work It's an interesting idea to use the title as the identifier of the deprecation.
tadzik wait wait! :) 20:25
dukeleto: I can, sure
I'm down to 56 items, one lost
dukeleto cotto_work: interesting is a word :) Perhaps having a hash, where the parrot eligible version is the key, and the values are the deprecations associated to that
cotto_work: but for now, this is much better than what we had
tadzik++
whiteknight tadzik++
cotto_work tadzik++ 20:26
dukeleto whiteknight: i think DEPRECATED.pod will become a generated file
cotto_work I'm excited.
Our users might even not hate us.
whiteknight dukeleto: That's fine, I suppose. I would almost rather not have it at all, but that's a small nit 20:27
maybe only generate it during make install
dukeleto whiteknight: well, the point of having DaD is being able to use that data to do stuff, like scripts that search source code for deprecated uses, or making a nice HTML page of currently deprecated stuff 20:28
whiteknight: i don't strongly care, but we should have a human-readable version of deprecations that is generated from deprecated.json
tadzik great, now I have 56 items: one more than in DEPRECATED.pod somehow 20:31
whiteknight tadzik: I thought you said there were 57 in DEPRECATED.pod
tadzik 57 =items, 2 are not deprecations 20:32
so 55 in DEPRECATED.pod, 56 in the generated JSON 20:34
whiteknight awesome
tadzik worrying 20:35
looks like flawed pod2text output 20:36
dalek rrot: e29f45d | nwellnhof++ | / (11 files):
Merge branch 'master' into nwellnhof/unicode_filenames
rrot: c6bc20c | nwellnhof++ | / (35 files):
Merge branch 'nwellnhof/unicode_filenames'
whiteknight nwellnhof++ 20:37
nwellnhof git is really black magic. it transmitted 902 bytes for that branch merge.
parrot should now be ready to handle unicode filenames for .pir, .pbc, .pasm 20:38
dukeleto nwellnhof: if the content already existed, but is getting re-arranged, very little needs to go over the wire
while(1) nwellnhof++ 20:39
20:41 plobsing left
dukeleto nwellnhof: does that mean that nqp-rx can deal with unicode filenames as well? 20:41
nwellnhof dukeleto: every parrot user should automatically get all those unicode improvements. 20:42
dukeleto nwellnhof: awesomesauce
nwellnhof everyone is encouraged to give unicode filenames a try. i couldn't write tests yet because we can't rely on perl for that, and we still need things like deleting files with unicode filenames from pir. 20:48
tadzik alright, the list is ok 20:50
wklej.org/id/461692/
can be used, tuned, fixed, etc
so after all there are 56 deprecations 20:57
dukeleto tadzik: awesome! 21:02
tadzik: do you want to commit the new deprecation.json to parrot.git ?
tadzik: or make a github pull request for it?
tadzik: we also need a bit of a doc update. There are probably various places that talk about DEPRECATED.pod that need to be deprecated.json 21:03
tadzik dukeleto: I can create a patch, if I think about cloning the huge Parrot repo again I fell allergic to pull requests :)
or I can quickly set up a git repo at home (: 21:04
dukeleto tadzik: heh, no problema
tadzik: patch is fine
tadzik dukeleto: btw do you have any news on my CLA?
dukeleto particle: do you have news about tadzik's CLA ? 21:05
tadzik: i think we just need to give you a bit
tadzik: i will give you a commit bit to the parrot github org now 21:06
tadzik why, Thank You
21:15 whiteknight left
tadzik wklej.org/id/461703/txt/ -- here's a patch 21:16
well, DEPRECATED.json needs some tuning, let me make a few cleanups
wklej.org/id/461710/ -- new patch, with a few cleanups 21:21
21:25 plobsing joined
dukeleto tadzik: you are now laden with a commit bit to all parrot repos! Use it wisely and don't break stuff :) 21:32
tadzik dukeleto: I'll commit this to a branch for now
dukeleto tadzik: sure, works for me 21:33
tadzik: we can make sure that all the docs get updated in that branch, then merge it
tadzik: thanks for working on this!
tadzik: it is really important
dalek rrot/tadzik/deprecated-json: 56207d3 | tadzik++ | DEPRECATED. (2 files):
DEPRECATED.pod is now DEPRECATED.json, adjusted rest of the files due to the change
 21:34
tadzik branch pushed, I think
dukeleto tadzik: yep, see above :)
tadzik oh crap, the commit msg lies 21:35
dukeleto tadzik: OH THE HUGE MANATEE
dalek rrot/tadzik/deprecated-json: ff6a5bc | tadzik++ | / (5 files):
s/DEPRECATED.pod/DEPRECATED.json
 21:36
dukeleto tadzik: also, can you update the TT that talks about Deprecations as Data and add a link to that branch?
tadzik I think so
dukeleto tadzik: TT 1888
tadzik: trac.parrot.org/parrot/ticket/1888
tadzik updated 21:39
dukeleto tadzik++ 21:40
tadzik sleepytime for me now. o/ 21:43
cotto_work 'night 21:44
dalek TT #1986 created by fperrad++: readline on Socket is broken 21:46
TT #1986: trac.parrot.org/parrot/ticket/1986
21:49 davidfetter joined 21:51 rurban_ joined 21:52 fperrad left, rurban left, rurban_ is now known as rurban
cotto_work dukeleto: did you ever merge that change from gerd that I reverted before the release? 21:56
21:57 hercynium left 21:58 davidfetter left
dukeleto cotto_work: nope 22:05
LuaJIT Roadmap for 2011: article.gmane.org/gmane.comp.lang.l...eral/74072 22:06
Qualcomm is sponsoring ports of Lua. Very interesting. 22:08
22:08 mtk left
Coke does that include the lua on parrot parts? 22:17
22:19 Eclesia left 22:36 plobsing left
cotto_work looks like whiteknight may have merged it 22:39
22:50 bacek joined
cotto_work hi bacek_at_work 22:55
er, bacek
22:59 fbrito joined
cotto_work dukeleto: based on your experience, do you think yaml would be a better choice for deprecation data? 23:00
Coke parrot has a json reader. 23:03
your argument is invalid. ;)
cotto_work I suspect we'll eventually want yaml 23:04
a yaml parser, rather
Coke yes, but we don't have that now. 23:05
and json is just fine.
heading offline.
23:12 pmichaud_ joined, pmichaud left 23:17 theory left
dukeleto json and yaml are isomorphic 23:23
cotto_work: jsync.org/ 23:24
cotto_work: i think we still need to think about the structure of deprecation data 23:25
cotto_work: we need to think about use cases. What kind of code is going to query the deprecation data?
cotto_work dukeleto: yes. I'm also thinking about how to make it minimally annoying to edit. 23:26
dukeleto cotto_work: yep, YAML is nicer for humans to edit, for sure
cotto_work: parrot doesn't have a YAML parser yet, though
cotto_work No, but we could probably get one in short order.
dukeleto cotto_work: what language would you write it in? 23:27
cotto_work for $dayjob? PHP.
nm. misread 23:28
I don't know. I'd have to play with it for a while.
sorear there's only one reasonably complete JSYNC implementation, AFAIK, and it's not usable from Parrot
cotto_work does it have tests? 23:29
sorear yes
dukeleto cotto_work: it is written by Ingy, of course :)
github.com/ingydotnet/jsync-pm 23:30
i think we should probably just stick with YAML, but JSYNC sounds really cool 23:31
cotto_work +1 to both sentiments
sorear mm, two then 23:32
I wasn't counting ingy's - it used to be much less complete, and stagnant 23:33
dukeleto sorear: what is the other implementation? 23:37
sorear: i just saw your comment in #perl6 23:38
23:39 kid51 joined 23:44 whiteknight joined 23:50 vmspb joined
cotto_work hi vmspb 23:50
vmspb hi 23:51
dalek TT #540 closed by jkeenan++: installed versions of dynext/*.so still link to -lparrot in build ... 23:58
TT #540: trac.parrot.org/parrot/ticket/540

Powered by Cro and the Raku® Programming Language.

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

Thank you!