The Return of the Journal : pugs.blogs.com/ | pugscode.org | pugs.kwiki.org | paste: paste.lisp.org/new/perl6 or sial.org/pbot/perl6
Set by GammaRay on 31 December 2005.
00:01 azuroth joined, arcady joined 00:15 \xe6var joined
svnbot6 r8769 | putter++ | Parser.hs - added q_to_eof(), the beginnings of file-level source filters. 00:26
r8769 | putter++ | Currently, the best one can do is use Mumble; mumble_macro q_to_eof(); ... .
r8769 | putter++ | Which isn't much better than heredocs. But the complexity cost is low - it's
r8769 | putter++ | only 6 lines of simple code, only one of which touches the existing parser.
r8769 | putter++ | This simplicity is also why it wasn't integrated as q:to_eof/uncertain_semantics/.
00:33 scook0 joined 00:35 mjl69 joined
Juerd rafl: I still haven't received my camera, and fear it may be lost in transit 00:58
rafl: Do you perchance have a tracking number? 00:59
01:22 rantanplan_ joined 01:29 Khisanth joined 01:31 \xe6var joined 01:32 putter joined
putter r8766 smokes of normal, pil2js, and pir are up on m19s28.vlinux.de/cgi-bin/pugs-smokeserv.pl , under their various "repository snapshot / linux". fyi. 01:34
01:36
luqui: we should chat sometime about regexp engine designs. 01:38
01:45
tewk: how did the numbers.t :2<1001> etc parser code go? is it committed? would you like someone to help out? 01:47
I assume you saw bsb++ refactoring of number()... 01:48
02:15 \xe6var joined 02:19 rantanplan_ joined 02:21 vel joined 02:25 Cryptic_K joined 02:41 feng joined 02:42 alexe_home joined 02:56 rantanplan_ joined 03:01 stevan_ joined 03:25 putter joined
putter hey stevan 03:25
03:31 dominix joined
putter www.smock.de/jsmock/en/33-hints.html Hints for Computer System Design - a classic 03:42
03:43 stevan__ joined
audreyt woot 03:45
putter ?
audreyt posted mg.h patch to p5p
putter googles
audreyt fixed 5.8.8rc1's swig interop problem yesterday
putter ah :)
audreyt perlcabal.org/~autrijus/tmp/mg.h.diff
so I can use svk again 03:46
good thing that this is caught in RC before it bites another person :)
audreyt proceeds to fix other svk/perl588 incompats 03:47
putter sigh. one word and boom. thus the argument for biological-analog software.
audreyt mm biological-analog. 03:48
"I cut my finger but my brain doesn't turn to blue screen"
putter *chuckle* :)
audreyt what's project oxygen doing nowadays anyway.
putter no idea
audreyt it's been over a year with no press release or confs or anything 03:49
okay, the svk side is fixed as well 03:50
"svk sy" no longer segfaults ;)
03:50 dduncan joined
dduncan hello there 03:51
audreyt dduncan: greetings!
dduncan I am now running Mac OS X 10.4.4 Tiger for the first time
yesterday it was 10.3.9
so far so good
audreyt wonderful
so you can get svk installed? :)
dduncan I still have to reinstall GHC and svk and such under the new system
however, svk should be easier to do than before 03:52
because the current package installers were Tiger-only before
I had to do the package source to make it work under panther
audreyt nodnod
dduncan I don't anticipate problems, though
audreyt putter++ # having a smoke ready for me so I don't have to run the loop first 03:53
okay, back to triaging.
03:53 tcliou joined
audreyt ooh, hs' is made public. I can finally talk about it freely here :D 03:54
# www.haskell.org//pipermail/haskell-...14036.html
dduncan audreyt, a comment on your latest journal entry 03:55
putter actually, it's more than just ARPA. US funding for research is starting to look.. almost european. ;)
dduncan I saw the 5.8.8 change summary too
and it looked mainly like the updates were putting in newer versions of bundled modules 03:56
audreyt that happens every release
but look at "Performance Enhancements ^" and "Performance Enhancements ^"
dduncan can you name anything specific in the core that was particularly better? 03:57
the other stuff I saw mentioned some refactoring, and that the binary is smaller than before
is that it mainly, or are there actual feature/behaviour changes that are quite noticible? 03:58
audreyt O(1) weakrefs; no redundant SVGV creation; aggresive "static" and "const"ing of core; utf8 under debugger no longer results in 100x slowdown
dduncan I've used weakrefs and never understood that there was a speed problem with them
audreyt if you used it in ithreads context, then you would
<- deployed ithread application under mod_perl2 in Win32
dduncan okay, if that 0(1) thing relates to ithreads ... 03:59
audreyt anyway, this is a point release, so obviously there would be no features
dduncan I haven't done any multithreaded stuff yet
not even minor features?
putter :)
dduncan well, no matter
audreyt no, I think the point of point releases is that minor features is only added if the old misfeature is too broken
and as we are at 5.8.8, that probably doesn't happen
dduncan I expect to test it soon enough
audreyt oh, if you want to recompile svk with it (not likely, but in the off chance), I just posted a patch to p5p5 04:00
perlcabal.org/~autrijus/tmp/mg.h.diff
it should be merged in RC2
(without that, swig fails)
dduncan in the short term, I'm going to just install svk using the system perl (5.8.6 now) as last time (the svk installer didn't work when linking to my custom perl 5.8.7, in a nonstandard location, while I use the custom for everything else) 04:02
audreyt that'd work merrily.
wow, the #haskell people are crazy at this alioth thing 04:03
I swore they were at 3rd place yesterday, now it's 2nd
# shootout.alioth.debian.org/benchmar...p;lang=all
audreyt ponders having a exapmles/shootout/ 04:04
anyway, lunch+triage first. bbiab & 04:06
putter haskell - its never a problem to optimize... if you can bring the entire hs community to bear. ;) 04:07
audreyt I think that's true for every free language :)
putter :) I wish I could see diffs over time of the implementations... 04:10
audreyt (otoh, nonfree languages also have communities, and sometimes they can even write optimizers without access to code)
putter: it's all on hawiki
haskell.org/hawiki/RecentChanges
grep for "Entry"
04:11 Amnesiac joined
putter ahh, nifty 04:11
ok, have cold, wiped. good night all & happy hacking. 04:18
04:41 dominix left 05:09 vel joined 05:10 vel left 05:33 hcarty joined 05:34 hcarty left
meppl guten morgen 05:55
azuroth morning 06:10
meppl good morning azuroth 06:30
azuroth what's on the plates today?
06:35 Cryptic_K joined 06:39 feng123 joined 06:49 Kattana joined 06:59 iblechbot joined 07:11 elmex joined 07:20 Cryptic_K is now known as protoss, protoss is now known as K 07:21 dduncan left 07:43 kakos joined 07:49 DesreveR joined 07:52 lampus_ joined
svnbot6 r8770 | audreyt++ | * Adverbial number forms eg. :16<deadbeef> works, as does :16[14,15]. 07:58
r8770 | audreyt++ | Still doesn't do the horribly complicated nested forms and
r8770 | audreyt++ | multiplication and exponentiation; TODOing 3 test out of 65.
r8771 | audreyt++ | * Remove hex() and oct() from prims, per S02. 08:04
r8771 | audreyt++ | * Update hex.t and numbers.t accordingly.
r8772 | audreyt++ | * hexadecimal.t: Use the new \x[abcd] form instead of \x{abcd}.
r8772 | audreyt++ | All tests pass there now.
Kattana Good to see development carrys on in the (night|morning), even on the weekend. 08:17
audreyt :) 08:21
Kattana I guess it will be a while before I catch up with the big fish, atm I am standing on the bottom, drowning, and everything is getting dimmer with each little bubble of info I see. 08:29
audreyt "perl6 can cause oxygen starvation" 08:43
Kattana If only that had been in a faq or somewhere, I would have taken a deep breath first. 08:47
08:56 elmex joined 09:30 marmic joined 09:43 GeJ joined 10:13 elmex joined, G2 joined 10:20 feng joined 10:42 feng123 joined
rafl Juerd: I'll ask my flatmate, who shiped it, tomorrow. 11:07
11:08 bsb joined 11:15 chris2 joined 12:55 DesreveR is now known as r0nny 13:00 GeJ_ joined 13:17 eric256_ joined
Juerd rafl: Okay, thanks 13:22
13:35 joepurl joined 13:36 kane_ joined 13:44 GeJ joined 14:17 stevan__ joined 14:35 lypanov joined 14:42 qmole joined 14:47 frederico joined 14:59 chihchun joined, Limbic_Region joined 15:23 kane_ joined
Limbic_Region anyone else having build problems on Win32? 15:24
Juerd refrains from silly win32 bashing 15:28
Only three users on feather right now. Why are people using it less and less? 15:29
Alias_ feather?
wolverian feather.perl6.nl
Limbic_Region Juerd - I only need feather access from work which is where I can't connect to 15:30
webmind Juerd, I think the unusability with high load might have an impact ?
Juerd webmind: It might, but no-one except you has reported it to me, and I've yet to experience it.
wolverian Juerd, I use it all the time to test pugs.
Juerd Besides, the load hasn't been high in a while.
Limbic_Region Juerd - and I am not a Win32 or a *nix bigot - I use whatever is available
webmind Juerd, care for the logs ?
Juerd webmind: If that shows that it's unusably high, sure
But high load and inavailability aren't the same thing.
webmind well they tell you load and cpu usage 15:31
Juerd Feather had a load of over 20 for days, a while ago, and was perfectly usable then.
webmind ok
Juerd Linux is quite capable of managing load
webmind the k-line on efnet maybe then ?
hmm, not sure agree on that for workstations though 15:32
even with nice levels it somestimes just is crap
15:32 xern joined
wolverian I haven't had any problems with gnome after I upgraded to 768mb ram. before I'd get regular lockups. I think it's just I/O that people experience 15:33
Juerd Is it klined?!
WTF?
wolverian I hope it's not, as I'm ircing on efnet fro mit 15:34
Juerd I can connect it to irc.efnet.nl quite properly.
wolverian s,fro mit,from it,
Juerd So no, that c8an't be a reason.
webmind: Why are you the only one who experiences this and mentions it?
Have I done so badly to indicate that I'm available for comments and complaints? :(
webmind Juerd, no I think it has more to do with human nature and not complaining 15:35
or actually more a cultural thing
but I'm just good at complaining :) 15:36
ah efnet works now yes.. there used to be a kline
a week back or so
Juerd With so little actual use, it's becoming less worth the costs of maintaining its existince.
15:37 aufrank joined
webmind hmm 15:37
aufrank good $time_of_day everyone!
webmind good day
Juerd Hi
aufrank I've got 1, maybe 2, bugs with the make install process. I haven't had any success eliminating them on my own 15:38
Juerd And it's important to note that there haven't been many newcomers anymore
lisppaste3 aufrank pasted "make install bug" at paste.lisp.org/display/15902 15:40
aufrank pasted "maybe a bug?" at paste.lisp.org/display/15903 15:45
webmind Juerd, well set a poll in the motd
or when it's really a problem mail to the list ?
or find a way to attract more perl6 coders 15:46
Juerd webmind: "Why are you not using feather?" in feather's motd doesn't quite work ;)
webmind ofcourse it could be that people are just using it more on their own hardware
Juerd I have a list of mail addresses of users. I should start using it.
webmind yep :) 15:47
Juerd webmind: That's certainly possible.
rafl Juerd: Re eu-geek-tour: Any news? Do you already know what you want to visit? 15:48
Juerd rafl: I think I'll just do it, except for the cebit part. That's expensive and makes the entire tour last way too long. 15:49
rafl Juerd: I'm also not that interested in cebit, so that's fine with me.
Juerd: Do you have a schedule?
Juerd And going from .de to .nl and then back again is a little silly :)
rafl Juerd: Indeed. 15:50
Juerd rafl: Yep, hold on
I should clip my nails.
rafl That's what I'm telling me for almost two weeks now. 15:51
Juerd It becomes harder to use a laptop keyboard as nails grow.
Why do we have nails anyway?
rafl To scrape ourselfs?
wolverian ask the intelligent design proponents.
integral to peel the foil off the top of meals-for-one
Juerd tnx.nl/4433BNBF 15:52
rafl Or to chew them maybe..
Alias_ To praise the lord
Juerd integral: I have a leatherman for that.
webmind Juerd, imagine your fingers without them.. I think it would be less solid typing
lypanov is the darcs repo up to date?
Alias_ To claw out the eyes of the heathen masses
webmind also they help in opening packages and astuff
rafl Juerd: So why don't you use it for that instead of punishing other people as you did in Berlin?
webmind -a
Juerd rafl: Because I'm a hacker. 15:53
I use whatever tool is effective :)
rafl Juerd: There are other tools that are more effective for that. And it would be even more effective for creating social connections not to harm anyone.. 15:54
Juerd rafl: There weren't that many tools around, IIRC :) 15:55
rafl: And I don't think social connections were harmed :)
rafl Juerd: Actually it was funny.. right. 15:57
Juerd rafl: Anyway, what do you say re geektour :)
Especially something on feb 27 or feb 28 would be nice :)
rafl Juerd: I'll join. Do we have space for Timo?
Juerd My car moves 4 people comfortably, 5 less so 15:58
And I don't have any passengers registered yet
So, yea
webmind: Will you join? :)
rafl Great. Then we also have someone that'll drive with us from Bochum to Chemnitz.
Juerd: So please register me. I'll ask Timo tomorrow.
Juerd: No idea that to do on 2{7,8}.2. 15:59
webmind Juerd, not sure yet
Juerd rafl: Do you have a driving license?
rafl Juerd: We could can do a chemnitz perl mongers meeting in 6.3.
Juerd: No.
webmind first thing I'm going to plan now is to let someone cut in my eyes
wellaaround them 16:00
Juerd rafl: Especially Chemnitz -> Amsterdam is WAY too much for me alone.
webmind and then plan around that
rafl Juerd: Timo as well.
Juerd rafl: Also no license?
16:00 \xe6var joined
rafl Juerd: Right. 16:00
Juerd rafl: 6/3 is needed for the LONG drive
rafl Juerd: How long will that take? 16:01
Juerd rafl: 7 hours if all goes well
rafl Juerd: OK, thats long. But we don't have any events on that day, so we can do a lot of breaks.. 16:02
16:02 penk joined
Juerd rafl: Right, but still, even with breaks, it's quite a lot for one driver :) 16:03
I'm willing to do it only if other people promise to not get drunk the day before :P
rafl our $driver = Storable::dclone($Juerd);
Juerd Because with moody passengers, it's harder even to stay focussed
rafl Hm... is it enough if one of is doesn't get drunk? ;-) 16:04
Juerd webmind: Okay, here's the plan - you get your eyes cut halfway March, and your driving scrap in a few weeks from now, and you will join the geektour. How does that sound? ;)
webmind Juerd, can't 16:05
Juerd Darn
webmind I actually suspended my lessons until after the operation
Juerd Oh 16:06
Then, get your eyes cut next week, take up the lessons, get your license the first week of Feb, and join :)
webmind the current condition of my eyes wasn't really helping
and that's being very positive about and our medical health care system and the burocracy of the cbr (have to get medical testing now from them) 16:07
Juerd How much will you be affected by the operation, by the way?
rafl Juerd: OK, I think I can promise not to get drunk, even if it's hard as there will be free drinks for me..
webmind I will probably able to use both my eyes at the same time on a regular basis
Juerd webmind: I meant more: hoe ingrijpend is het?
Dunno how to express ingrijpend in English :) 16:08
webmind I think I'll be able to see quite directly
they'll be replacing to muscles
but I'll be under a heavy narcoses
Juerd Won't your brain need time to adjust?
webmind and maybe I need to get used to the moving them around.. dunno yet
Juerd For the eye coordination...
webmind yeah
maybe
I'll ask this when I get the chance 16:10
probably this week
got some more questions anyway
Juerd Why wasn't this done before, by the way?
Didn't plan on driving? ;)
webmind only recently looked into it
that too 16:11
but my timing with this is horribly
Juerd Timing is always horrible when something is needed
webmind well I probably could get my licensen with my curretent eye coordination 16:13
but I notice I'm mainly cheating
which results in unsafe driving
don't like that
ofcourse if they pierce my eye while stiching the muscle back I won't be able to drive at all.. but ey :) 16:16
Alias_ shudders 16:17
Of course, we all have our own problems that we've gotten used to 16:18
Alias_ still hasn't entirely come to grips with having to wear glasses
webmind uhuh 16:19
I recently discovered implant lenses 16:20
low risk and no trouble with replacing
first going to do this though
Alias_ I need to wait until my eyes stop shifting
Also, the rate of these eye-related inventions is improving at such a speed I'm happy to delay for a few years 16:21
webmind uhuh
Alias_ As long as I'm not playing a sport for which glasses is a problem, I'm relatively happy
I still get a bit icky about the loss of field of vision though 16:22
every now and then I take off my glasses and remember what it was like to have a full 160 degrees
well, 140-160
It adds a certainly connection to reality... 16:23
Putting the glasses back on feels like it adds a hint of depersonalisation
Anyways...
Alias_ & # food
16:26 bsb joined
Juerd rafl: The entire tour, beginning and ending in Dordrecht, NL, is approx 2000 km, and that'll cost only around 150 euros in LPG :) 16:26
rafl: Including car costs, this brings the trip costs up to approx 200 16:27
rafl: Divided by 4 (hoping on 3 passengers total), that gets us around for 50 euros each :)
rafl: That's dirt cheap... 16:28
Try that by train ;)
Though for you it'll be more expensive because you have to go to .be by train and to .de by train
rafl: Do you know what good options there are for sleeping in Bochum? 16:30
Alias_ Juerd: That seems unusually cheap 16:42
Does that factor in maintenance/devaulation etc?
Generally car costs are around 20-40euro'cents per K 16:43
16:43 bsb left
Juerd Alias_: It factors in maintenance, not devaluation. 16:45
Actual costs for my car are 21 cent per km.
Provided Dutch LPG. German LPG is more expensive, for example
16:51 cookys joined 16:55 j0sephi joined
svnbot6 r8773 | audreyt++ | * Text.Parser.Kwid: Initial check in of a Haskell Kwid parser. 17:50
Alias_ shudders
svnbot6 r8774 | audreyt++ | * Remove unused imports from Pugs.CodeGen.PIL2 and Pugs.Compile 17:53
17:56 r0nny joined 18:00 if joined
svnbot6 r8775 | audreyt++ | * "make haddock" now works again. 18:00
18:00 if left
svnbot6 r8776 | audreyt++ | * ditto for drift-derived files. 18:03
lypanov squeaks 18:04
chris2 hands lypanov a bonbon 18:16
lypanov eats
lypanov makes some coffee
18:55 elmex joined 19:05 lypanov_ joined, lypanov_ is now known as lypanov 19:13 feng joined 19:16 Lartuv joined 19:23 _maydayjay_ joined 19:42 Cryptic_K joined 19:55 _maydayjay_ joined 19:58 _maydayjay_ joined 20:11 lampus_ is now known as lampus
Juerd Anyone interested in Mastering Regular Expressions, 1st edition? 20:12
Alias_ sigh 20:14
Juerd ?
Alias_ tries to think of how prove is different to make test 20:15
I have some test scripts that WORK with prove but FAIL in make test
Juerd Environment?
Alias_ Just one very specific case, most of the rest of it all works still
20:15 lypanov left
Alias_ And don't see how... I'm literally running them one after the other 20:15
Juerd Is it Perl? 20:16
Alias_ what?
Yes
Juerd make test may be including blib/lib, while prove may not.
Alias_ prove -br
So prove is using blib/lib too 20:17
Juerd Still, though, dump \%ENV and compare :)
Alias_ point... doing so 20:18
20:24 Medvekoma joined
Alias_ hmm... how interesting 20:34
Juerd++ # Dunno if it will help, but it might be a start
aha! 20:36
Juerd++
what an esoteric bug
Juerd Yay, karma :) 20:37
jabbot Juerd: :) has neutral karma
Juerd :)++
Alias_ heh
Juerd Yay, karma :)
jabbot Juerd: :) has neutral karma
Alias_ prove sets up the included directories differently to make test
Juerd jabbot++ # You're useless, but I like you anyway.
Alias_ prove's ones don't forvive run3'ing 20:38
s/forvive/survive
theorbtwo That's probably a bug.
Alias_ I'm in RT as we speak
wow it has a lot of bugs 20:39
and spam apparently 20:40
Juerd Of course.
Doesn't everything that is big have lots of bugs?
Alias_ no 20:41
Juerd The more code, the more chance for error.
Alias_ Assuming constantly bad code
Juerd Not really.
Alias_ has about 20 bugs for 107 distributions
15 of those are small PPI things I haven't gotten around to
Juerd But are those 107 distributions used as much as RT? 20:42
Alias_ Test::Harness has more bugs than my total number
rafl Juerd: 50 EUR sound reasonable.
Juerd: I think I can organize something in Bochum.
Juerd rafl: Does that equate to 'free'? :)
rafl: BTW, in .nl, I think we can sleep in my office 20:43
rafl: It's showerless (as it's an office building), but I think we can survive that.
rafl Juerd: I also think so.
Juerd It's a 1 hr drive from my office to Amsterdam 20:44
Something I do every month anyway :)
And recently, several times per week.
I can show you feather if you want ;)
Though it's not interesting if you've been in a datacentre before
rafl Juerd: I work in a super computing centre.. 20:46
Juerd Then it'd be uttely boring :)
Google for "aopen h360" and you know what feather looks like :P
aopen.jp/products/housing/images/h360-300.jpg # like that 20:47
Hm, no, wrong model
Alias_ hmm
Actually, I'm not sure if make test's behaviour is correct either
Juerd www.i-tech.com.au/images/products/AO89001_hg.jpg # like that
Alias_ aaaah 20:50
Now I see...
Because prove doesn't set PERL5LIB, it was run3'ing something that launched with the older version of the code, that didn't have the bug
Juerd It must have been a path issue. 20:54
Alias_ What was hiding it was that the old and new code was so similar
3 lines different, which resulting in a subtle change not fully covered in the tests that SHOULD have double checks that behaviour
3 element failure 20:55
Actually, 4 element
old and new code + prove behaviour + run3 + bad coverage
21:02 match joined 21:21 kane_ joined 21:29 bsb joined 21:44 amv joined 21:47 SamB joined 21:49 Lartuv joined 21:59 Lartuv joined 22:07 Lartuv joined 22:10 gantrixx joined 22:24 sweinig joined
Juerd Alias_: Nice way to spend your sunday. 22:25
Alias_ hmm? :)
Juerd Alias_: Hunting bugs in bug hunters.
Alias_ PITA is Da Big Weekend Project
That particular problem is par for the course
There's way bigger problems
22:31 oylenshpeegul joined 22:37 mjl69 joined 22:48 kakos joined 22:50 revdiabl1 joined 22:51 revdiabl1 is now known as revdiablo 22:58 drbean_ left 23:06 Discipulus joined 23:08 Cryptic_K left, marmic joined 23:10 drbean joined 23:15 oylenshpeegul left 23:45 Cryptic_K joined 23:48 Cryptic_K is now known as K 23:49 scook0 joined 23:57 Skud joined
Skud seen audreyt 23:58
jabbot Skud: audreyt was seen 15 hours 14 minutes 34 seconds ago

Powered by Cro and the Raku® Programming Language.

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

Thank you!