pugscode.org <<Overview Journal Logs>> | You have safely opened the door to many Perl 6 hackers. | pugs.kwiki.org | smoke: xrl.us/fmw2 | Mac OS X r1607 (435/4138) | Win2k r1600 (437/4135)
Set by stevan on 6 April 2005.
stevan pugscode.org <<Overview Journal Logs>> | You have safely opened the door to many Perl 6 hackers. | pugs.kwiki.org | smoke: xrl.us/fmw2 | Mac OS X r1615 (432/4142) | Win2k r1600 (437/4135) 02:37
jabbot pugs - 1615 - adding a few more mutli-dim tests 02:42
pugs - 1616 - * Try to include File::Spec even if we d 02:52
stevan morning autrijus 02:56
autrijus greetings stevan 02:59
how's going? 03:01
stevan good, and you?
autrijus just up and going to $work
stevan I am watching TV and getting ready to sleep :) 03:02
autrijus I think I need to fix multidim today.
stevan autrijus: that would be wonderful
autrijus :D
stevan I need it to finish CGI :)
so I would be very appreciative 03:03
autrijus okie.
ingy just spent ~30 hours in bed 03:11
autrijus good. I'd like that too
ingy actually I'm still in bed so counting... 03:12
crysflame hi, ingy
cls_bsd ingy: where's your Spoon-0.23? :D
ingy on cpan no?
hmmm 03:13
cls_bsd no, only 0.22
although I have patches from Spork and gugod@ :p
ingy wtf
cls_bsd inc/.../Makefile.pm from Spork, and Makefile.PL from gugod 03:15
Khisanth Spoon and Spork ... so where is Fork? 03:20
cls_bsd fork() will die() 03:21
ingy Spork *is* a subclass of Spoon 03:22
stevan but shouldn't Spork also be a subclass of Fork?
or at least a trait/role or something 03:23
cls_bsd WebService-GoogleHack, good 03:28
autrijus the Array data type is now going to change from List to IntMap... you should notice no external effects 03:29
except operations go from O(n log n) to O(n)
jabbot pugs - 1617 - * discard MkHash newtype from HV 03:32
ingy cls_bsd: Spoon-0.23 uploaded 03:45
cls_bsd ingy: great :) thakn you
ingy it appears to actually have worked this time 03:46
I *swear* I uploaded the on Monday
cls_bsd Yes, looks fine. :) 03:47
ingy++
Alias_ Did I hear someone mention CGI? 03:49
Is someone planning on rewriting CGI with a new API for Perl 6?
Can I put in my vote for PLEASE making sure it's clean and works well with mod_perl2 from the start?
ingy Alias_: fyi, Perl 6 will not suffer from the one implementation per module name problem of Perl 5 03:51
so there can be more than one CGI.pm 03:52
Alias_ I know, but we are going to have one primary blessed version per name :)
autrijus "we" means "alias" :)
each entity is free to have one primary blessed version per name
Alias_ I thought "we" meant "The CPAN QA team"
autrijus sure, CPAN QA team is such an entity.
Alias_ But if someone must be the artibrar of taste, I accept your wise nomination 03:53
cls_bsd perl6-Bible..
autrijus I think perl-qa is fine for that
but I think multiple blessers is totally the way to go.
ingy agrees
autrijus i.e. I'd subscribe to Randy Kobes's list for Win32 Apache modules, etc 03:54
Alias_ I don't care who does the blessing, so long as the "normal" way of doing things is simple
And code never changes in meaning because of context (like mod_perl2 wanted)
Modules looks like the most dangerous place for second system dangers to take hold ATM 03:55
But then I never like new ideas 03:56
Especially ones that don't come with well though out 5 page documents that explain the effect on all the stakeholders and use profiles :)
autrijus: Oh, BTW, what mechanism do you use in M:I for stripping modules for inclusion in /inc? 03:58
autrijus Alias_: same code as PAR::Filter::PodStrip
Alias_ autrijus: I hope to do Perl::Compress, to make a compressing thingy 03:59
autrijus: Can you easily put an optional hook in there somewhere once it's done?
if ( installed(Perl::Compress) ) { ... } else { current code }
autrijus sure, that'd be easy 04:00
as long as there's no self-parsing code 04:01
Alias_ self-parsing code?
autrijus seek DATA, 0, 0; my @code = <DATA>;
Alias_ um...
PPI don't touch __DATA__ 04:02
jabbot pugs - 1618 - * Array is now an IntMap.
autrijus good good
then it's all fine :)
Alias_ I presume I can remove __END__ though :)
jabbot pugs - 1619 - * more use of monadic fixpoints in loops 04:12
shapr w00 fixpoint! 04:19
autrijus I like fixpoints and how it magically works in ContT :D 04:21
shapr yeah, fixpoint is another one of those simple and elegant ideas that does neat stuff. 04:22
Alias_ writes his first cron job in 3 years 04:26
It's a lot easier than it used to be
ingy ?
shapr I couldn't get my head around fixpoint for a long time, but after enough reading I saw that it's just an all-monad safe Haskell implementation of Y used for a higher order recursion combinator. Not so scary after all.
Alias_ I have horribly memories of adding manual entries to the crontab
maybe more like 6 years
Now it seems that dron.daily is nice and easy
cron.daily
ingy has it changed in 6 years?
Alias_ I don't think the cron.daily/hourly were used as much 04:27
At least, not on whatever distro I was using 6 years ago
puetzk ingy: only in that most distros now supply a stock crontab that does run-parts on some dirts
dirs
Alias_ yeah, I don't think cron itself has changed, just that the distro are provide default setups that are more friendly 04:28
ingy someone tell thorbtwo that I added a Perl6::Bible->get_raw method for him 04:50
I think I will release Perl6-Bible immeditately after every pugs release 04:51
(if needed)
autrijus did you include PDD and PA? 04:54
that's parrot and pugs docs 05:00
crysflame ingy: heh, cool 05:51
21:50 < ingy> someone tell thorbtwo that I added a Perl6::Bible->get_raw method for him
theorbtwo: ping
it's 22:52 my time 05:52
05:54 tmoertel_ is now known as tmoertel
Daniel_Nee Hi, Autrijus 06:07
gaal morning all 06:08
Daniel_Nee Cindy told me this morning, there're some e-mail attachments missing since the changes made on Monday, she would like us to get it solve this afternoon. 06:10
theorbtwo Pong. 06:15
Mornin, all. 06:17
shapr theorbtwo: greetings! 06:20
theorbtwo: Sadly, I spout that kind of random confusion so often no one believes me anymore. 06:21
Like, my name is Erisson, some people are beginning to notice the connection. 06:22
theorbtwo Connection? 06:26
shapr hail eris?
theorbtwo Oh. 06:28
crysflame theorbtwo: ingy message for you ping 06:29
theorbtwo Pong.
Saw it.
crysflame :)
06:36 castaway_ is now known as castaway 06:38 mugwumptest is now known as mugwump
shapr mugwump: Are you using a perl6 irc client now? 07:01
mugwump no, should I be? :)
shapr just curious, I saw the mugwumptest 07:02
mugwump ? does such a beat exist, I wonder
*beast*
castaway you could try porting irssi :) 07:03
shapr I could wrap lambdabot/hircules into Pugs. That's kind of cheesy though.
mugwump sure
pugs -e 'eval_perl5("irssi.pl")'
shapr grins 07:04
castaway cheat :)
mugwump Actually a wrapper around the haskell curses library would probably go a long way to that sort of thing 07:05
crysflame win 5 07:08
shapr places a bet on camel 6!
mugwump & # food 07:10
cls_bsd camel6 07:15
theorbtwo notes that it's the camel 4th ed that will cover p6.
shapr When castaway says you're her pal, is that actually short for palantir? 07:16
castaway doesnt recall ever having said that :) 07:17
castaway ruffles shapr.
shapr gets ruffled - www.scannedinavian.org/~shae/foto/n...aircut.jpg 07:18
castaway iek, hairy monster :) 07:20
note to self: install sql navigator in a VM and suspend it (doesnt count evaluation days if still running ,) 07:21
theorbtwo Mornin, meta. 07:23
metaperl gm theorbtwo
where does your name come from BTW?
theorbtwo It was just "theorb", but there were too many of them, so I became "theorbtwo". 07:24
"two" because everybody else at the time was postpending digits. 07:25
theorbtwo wonders why he keeps saying "postpending" instead of "appending". 07:26
shapr postfixing?
theorbtwo see also perlmonks.org/?node=theorbtwo's+name+space 07:33
Darren_Duncan hello, anyone here? 07:36
shapr snores quietly 07:37
Darren_Duncan quick question ...
I'm starting to release my perl 6 modules on cpan, but I wanted to ...
Alias_ err... we can do that? 07:38
Darren_Duncan say the right thing in my makefile.pl so that CPAN does not index parts of the distro, such as the perl 6 modules themselves, because there would be a name conflict with the perl 5 versions
I noticed some distros say no_index or 'private' in their Meta.yml
Alias_ you need no_index entries in the METAL
META
Darren_Duncan and I'm wondering what to put in my makefile so that 'make dist' puts no_index in the Meta file? 07:39
Alias_ erm... not sure
Module::Install does it for me
Darren_Duncan I tried grepping Makemaker but nothing found
Alias_ are you doing modules with both P5 and P6 code? Or just P6 code?
wolverian Darren_Duncan: I don't know if that's a good idea.
Darren_Duncan so how do distros that don't use Module::Install do it?
Alias_ by hand I imagine 07:40
wolverian besides; freepan.org/modules.html
Darren_Duncan the Makefile is pure perl 5
Alias_ Make the Meta file from scratch yourself and make sure makemaker doesn't overwrite it
Darren_Duncan the modules themselves are pure perl 6
Alias_ umm... why are you putting them on CPAN then?
Darren_Duncan essentially, I want to do for myself what Pugs used to do with its /modules directory
Alias_ if P6 doesn't support CPAN, and CPAN doesn't support P6
wolverian currently the plan is to use FreePAN. 07:41
Darren_Duncan this is meant to be a 'current state of things' snapshot
Alias_ I would suggest that at the moment it could only be a bad thing to release that way, via CPAN
Darren_Duncan even if I used Freepan, I wanted to still have gzipped snapshots the old way on cpan also
for now, I don't expect the perl 6 code to be executed ... think of these like documentation-only releases 07:42
... 07:43
metaperl Locale::KeyedText works in p6 doesn't it?
Darren_Duncan it requires object support
afaik, it does not yet 07:44
what I want is to have a distro channel that is like CPAN in that I can put my own gzips up there and they are listed, for Perl 6 stuff 07:45
since Freepan is new, I don't want to put all my eggs in one basket, I prefer to maintain a copy both there and on the old CPAN
my anticipated problem is because there are perl 5 and 6 modules with the same NAME pod in them, and the CPAN indexer can get confused unless I tell it to hide the perl 6 one; people would view it through the MAKEFILE link 07:47
I like the backup capabilities that CPAN provides; put work up there and its mirrored world wide
Alias_ There is _so_ much potential for confusing the indexer, even with no_index on... 07:49
There could be other things in there you aren't expecting
version_from, for example 07:50
or who knows what...
mugwump Yes, please don't do that. We'll have a mirror shortly. The system is already on a mirrored box, and has nightly backups
theorbtwo jabbot: seen corion 07:51
jabbot theorbtwo: I havn't seen corion , theorbtwo
Darren_Duncan Okay, so say I keep the perl 6 stuff off cpan then ...
castaway he was on PM just now
doesnt seem to join here til evenings
Darren_Duncan does Freepan have the capacity for me to do this ... 07:52
I want to upload 4 tgz distro-style files representing snapshots of my perl 6 modules at different points in time, with additional ones occasionally 07:53
they are snapshots in the same way that traditional CPAn releases are snapshots of a developed product 07:54
mugwump please, load them in as unpacked source files. Set the property freepan:version in the root of the project to "tag" the version
There is no difference between a tgz snapshot and one extracted from the source files which are exactly the same as that tarball's contents 07:55
s/extracted/generated/
That way, CPAN doesn't have to be so massive
Alias_ mugwump: space is cheap
mugwump Recently Jesse Best tried loading all of backpan into a subversion repository (well, all the A's :) and it was smaller 07:56
than the version on mini-ca^Hpans
space efficiency is good. "disk is cheap" is IMHO always been an excuse for people "not tidying their room" 07:57
Alias_ My way is tons more space efficient for me
None of these pesky LICENSE files lying around
just one that gets put into the tarball before upload
Darren_Duncan is there a requirement that all parts of a tag are aligned to a SVN release number?
Alias_ And a common MANIFEST.SKIP file
oh, and only one 99_pod.t
And did I mention that the release packager automatically checks newlines and POD and updates the copyrights and versions and writes the Changes file checks that every .pm file /t is indexed 07:58
etc etc etc
How am I supposed to maintain my obviously far more efficient method of storing modules, if you insist on unrolling everything 07:59
Alias_ bows
Darren_Duncan eg, if I wanted to upload 4 versions of a ChangeLog, one at a time, each version being a 'newer release' of the last, can I then say that each of the first 3 uploads is tagged the same as how all the other files were as of several weeks ago?
I want to make tags so it looks like the ChangeLog was there in the past
so if a dist was made from a certain tag, it would include a 'so far in this distro' changelog 08:00
the files I refer to contain more information than the commit comments
so commit comments aren't s substitute
the change log is summary style, like that released with the Pugs distro 08:01
mugwump sure. well, ideally you'd get it right the first time ;) 08:02
but in lieu of that, I can suggest two options
I can blaze your repository, and you can upload the revisions to it as you intended them to be
or, we can decide that the "last" revision that has freepan:version set to a particular value is considered to be the release repository version 08:03
Darren_Duncan alternately, I suppose I could write the ChangeLog file to summarize the past 3 and current 1 release, but the file itself only actually exists in the current 1 release 08:04
mugwump or, maybe instead of using these tags, we can change things so that we use the "cheap copying" subversion trick, ie, have a standard layout of trunk/, /releases/, etc
Darren_Duncan and I'll do the normal tagging for the first 3
mugwump sure. that sounds simplest.
Darren_Duncan then you don't have to blaze anything
mugwump :)
Darren_Duncan I'll do that then ... tomorrow 08:05
and while I'm at it, I'll add a readme and other such files
FYI, the collected release 4, or separate LKT 4 + SRT 2, will match the CPAN perl 5 releases I did a couple days ago 08:06
while Freepan has almost-current versions, they don't have a few bug fixes
will fix that
question ... 08:07
does the Freepan web site contain any management utilities, such as for tagging, or do I need to use my 'svn' client program to do it?
nevermind, I'll investigate it tomorrow 08:08
good night
and thanks for all your feedback
now I know not to make some stupid mistakes
mugwump Darren_Duncan: most of this 'crazy' property setting stuff will be something the freepan script does for you 08:11
Darren_Duncan I'll learn it as I go along
mugwump setting the password is the craziest one, it sends a digest of your password to the server in a property, then the server actually rejects the update with the message "password successfully changed"
Darren_Duncan I actually haven't hooked my svn to freepan yet ... it last committed to the Pugs repository
mugwump so that the digested password isn't extractable :) 08:12
jabbot pugs - 1620 - Add some more lc and uc tests.
Darren_Duncan Anyway, thanks to whom it concerns for Freepan ... it looks to be a valuable tool
mugwump That's the idea ... remember any good ideas how it can be better will be gladly accepted on the wiki 08:13
Darren_Duncan I'll have to learn where it is before I can suggest where it will go
mugwump (discuss them here too, of course ;))
wiki.freepan.org
Darren_Duncan oy
I note that that opening page seems to be full of broken image links 08:14
a row of them by the search box 08:15
only the tree in the corner shows up
checking in another browser ...
mugwump oh, sure. yeah I see it
Darren_Duncan the second browser (Firefox) shows text in the place of the images 08:16
the one with the question marks for images is Safari
mugwump Wow, safari doesn't even grok html 1.0 08:17
that's impressive
I thought alt tags were universally understood ... amazing
Darren_Duncan it's supposed to be fully up to date
maybe I'll report that as a bug 08:18
mugwump I've got a better idea, bang your head against a brick wall
it will probably result in more action ;)
mugwump is so cynical it hurts ...
Darren_Duncan I love Wikis ... making it easy to make links that don't go anywhere 08:20
mugwump yeah. It sure makes it look like there's a lot of information, though 08:21
Apache-- # stinking pile of shite, another crappy configuration snafu 08:23
Darren_Duncan now, according to the Freepan wiki, it is supposed to be complementary to cpan
mugwump sure 08:24
Darren_Duncan freepan shows works in progress, and cpan is for packages tied up with string
so while I plan to use freepan, what's the best way to distribute neat packages, today and in the future?
made of perl 6 code
or shall I not even try for the near future? 08:25
mugwump Are you volunteering to write a little script to check out an SVN version and make a tarball? :-) 08:26
Darren_Duncan I don't know whether it already does that or not
off topic - mugwump, what part of the world are you in? 08:27
mugwump well, that's the plan as to how it will work. It would have rules based on what the freepan:type of the module is that would look at MANIFEST & MANIFEST.SKIP to build the tarball
Right now, Taipei
Darren_Duncan so it's day over there
mugwump Tomorrow I'm flying back to Wellington, New Zealand 08:28
Darren_Duncan not too far away from where you are, relatively speaking
mugwump where are you based?
Darren_Duncan I'm based near Victoria, BC, Canada
so it's 1:30am here
anyway, good night 08:29
for me
mugwump night 08:30
Darren_Duncan but good bye anyway
mugwump 8:30pm NZ time now
Darren_Duncan I hear ya
and out ...
theorbtwo wonders why G is merged, instead of M. 08:42
castaway wonders what G and M are 08:44
scw M -> modified
theorbtwo scm.sipfoundry.org/svndoc/re28.html doesn't list modified. 08:45
scw sure, M only appears at `svn st' 08:46
theorbtwo Hm.
jabbot pugs - 1621 - Fix "undefined" var error to "undeclared 08:52
pugs - 1622 - * small cleanup on HTTP::Headers about t 09:22
09:30 bsb is now known as bsb_
bsb_ /msg nickserv link bsb bsb_ 09:30
09:31 bsb_ is now known as bsb
autrijus rehi. 10:10
theorbtwo Greetings, autrijus of the unreliable connection. 10:15
autrijus greetings.
castaway autrijus: You use SubEthaEdit, right?
autrijus I'll prolly disconnect in another half hour or so
castaway Can the client share documents, as well as the server side? 10:16
autrijus doing buildworld/installworld now
castaway: not sure what you mean.
castaway the side that initiates the connection is the client, the other is the server. once the connection is there, does it matter who has the documents? 10:17
(hmm, TCMMMStatus both ways, so probably it doesnt)
autrijus I do not think it does.
castaway nods
kungfuftr moo 10:19
theorbtwo Autrijus, have you tried the haskell module loading stuff with my newest changes? I'm not sure where to go from here. 10:22
autrijus theorbtwo: I had not. maybe you need to load dependency packages? 10:23
I'll take a look after dinner
<- freshly off $work
theorbtwo Nod.
castaway lucky for some :)
autrijus using perl6 and haskell at $work usually means easy and fast debugging :) 10:24
theorbtwo Hm, I did disable some of the hardcoded loading you did, but I tried pairing SHA1 down to the point where it was pretty much nothing, and still got missing symbols.
And those symbols are present in objdump --syms ./pugs.
clkao another perl5 bug hit me today
autrijus you'd need to load libHSbase again and explicitly (I think)
theorbtwo: have you traced the example in DynamicLoader 10:25
?
clkao: mm?
clkao some nasty rebless / overload crap
castaway decides its lunchbreak time
theorbtwo Hm, I haven't, autrijus.
ingy hi theorbtwo 11:35
theorbtwo IT WORKS! 11:36
ingy ?
theorbtwo I HAVE MADE FIRE^W^W LOADED ["sha1"]!
ingy yow!
gaal theorbtwo++
ingy theorbtwo++ 11:37
theorbtwo Of course, I still need a few bits and pieces, because I did this in no small part by commenting out most of SHA1__0_0_1.hs.
castaway wow, grats love!
theorbtwo But I think I know what many of them are.
ingy, saw you added the method I wanted to P6::Bible, thanks, I haven't yet fixed categorize_tests to use it, though. 11:38
ingy theorbtwo: I made your bible request patch fyi
ah
shapr w00h00 12:11
kungfuftr ingy: is there an faq somewhere about uploading to freepan, etc?
castaway looks at shapr. 12:12
mugwump kungfuftr: point SVN at tpe.freepan.org/repos/kungfuftr/ 12:20
kungfuftr ah... i sees 12:21
Limbic_Region not sure if gugod mentioned this previously wagner.elixus.org/~gugod/pugs.png but I saw the link in use.perl and found it interesting 12:23
shapr churn == rate of change?
ah I see 12:24
that would be neat to include with the smoke-test.
mugwump big spikes by the .po file updates
shapr Is there a smoke-test churn graph?
Limbic_Region shapr, in the absense of gugod - you can read the journal entry use.perl.org/~gugod/journal/24047 12:25
castaway hawo nm 12:26
shapr castaway: I thought that was mandarin at first, nyehowma 12:27
nothingmuch HOLA
caps, aah!
shapr ¿como est as? 12:28
mugwump ¿utf-8?
castaway na, just written-as-spoken 12:29
kungfuftr theorbtwo: hhhmmm... you know if anyones made any motions towards modularising the testgraph/smoke/harness stuff? 12:32
nothingmuch kungfuftr: i've pondered about it 12:37
and given some free time i will
nothingmuch is too busy to do anything but lurk
kungfuftr is tempted 12:38
nothingmuch please do
what we need to do:
mugwump 1. take action, not plan 12:39
nothingmuch std TAP to allow always reporting of CALLER::CALLER etc
mugwump: in theory
mugwump :-)
2. write tests
nothingmuch 2. make the backlinking a bit less hardcody
3. factor yaml harness format: 12:40
{ smoker => { id stuff }, report => { structure } }
with a more solid, hierarchal namespace
kungfuftr is currently seperating out his churn melarky 12:41
nothingmuch kungfuftr: try to generalize so that yaml harness could make sense for replacing Test::Reporter's guts 12:42
kungfuftr nothingmuch: huh? 12:43
ah right... hhhmmm...
nothingmuch it needs to be reusablwe
it should be valid for modules too
right now it's too pugs specific
and "real" releases as well as multi VCS revision info
even wacky stuff such as darcs context 12:44
ofcourse, you don't have to do all that
i'm just listing things I foresaw as troubler
testgraph needs a bit of a cleanup IMHO
syntax wise
it's a bit too scriptish for what it does, methinks
kungfuftr nothingmuch: shush! 12:45
=0P
nothingmuch another thing is it needs to have a generic way of doing stuff 12:46
like:
linking to doc: sprintf like format? callback?
consolidating results: should it know a consolidated result format?
what does consolidating mean?
i'd like to be able to merge together reports per platform, per revision, per ....
and see yellow where things don't exactly match 12:47
castaway ,)
kungfuftr nothingmuch: yeah, that'd need a central repository though
nothingmuch kungfuftr: nono
make the input valid as consolidated
or standalone
kungfuftr ah, i see
kungfuftr shall have a looky later 12:48
nothingmuch that is, the yaml harness format should know to represent both cases
and then we can whip up a script to merge the results
gaal reappears 12:58
the yaml harness isn't *too* pugs specific
we might consider making Test.pm use a more programmatic-friendly output format instead of TAP when it knows the yaml harness is running it 13:00
13:00 Sebastian is now known as caveman
theorbtwo I don't think any of it is terribly pugs-specific. 13:00
The most pugs-specific bit I can think of is the bit that gets the POD, given the first half of a link.
gaal it needs to factor out the source control revision detection stuff 13:01
13:06 caveman is now known as jsbach
kungfuftr wonders if there is a nice way to build up a nice UUID on a "per host-per user" way 13:07
theorbtwo kungfuftr: Generate a UUID, and save it. 13:24
mugwump ]][[gg\\hh]]hh 13:29
theorbtwo Hm, I now have SHA1 fully loading from the External.Haskell point of view, at least, but I'm not sure it's working. 13:34
Also, the way I did it is horribly horribly hackish.
...and, finally, I'm not sure it's working within Externals. 13:35
mugwump loses another 10 minutes of his life because Linux sux 13:38
kungfuftr theorbtwo: yeah, would be nice to be able to generate the same ID if you lost it, etc. 13:40
kungfuftr also thinks about serialising with Storable as an option (at least for local usage) 13:52
mugwump yaml! 13:54
mugwump apologises for that possibly redundant outburst, and blames turett's syndrome
stevan pugscode.org <<Overview Journal Logs>> | You have safely opened the door to many Perl 6 hackers. | pugs.kwiki.org | smoke: xrl.us/fmw2 | Mac OS X r1622 (451/4159) | Win2k r1600 (437/4135) 13:55
stevan so mugwump have you read much William S. Burroughs? or just Naked Lunch? 13:56
mugwump I haven't read his life works or anything like that 13:58
In fact only two of his books :) 13:59
stevan neither have I, I prefer the earlier stuff
which books?
mugwump Junky and TNL
stevan good ones :)
mugwump I really like his spoken word albums though
stevan I have this great book of his art,.. some really cool stuff "painted" with a shotgun 14:00
mugwump And I once saw a play called "Word Virus", by a local group of students ... covered the whole 50s counterculture scene
stevan I tried reading The Lost Boys, but it gets a little too much into his fetishes
mugwump heh, him and boys huh
stevan LOL, yes
theorbtwo is unsure where to go from here with the loading stuff, other then to move to hsplugins. 14:01
The problem is that the .o files we're trying to load don't carry their own dependency information.
stevan I actually read a really good biography of him as well,.. I borrowed it from a friend so I dont recall the title offhand
he was a very interesting character
mugwump Red cover? fairly thick, 400-=600 pages or so? 14:02
stevan mugwump: honestly I dont recall the cover, but it was about that many pages
mugwump Perhaps that was "Word Virus" :). I've got that, but have so far been too lazy to read it
Pretty sad, really, considering my nick 14:03
mugwump shrugs and sips some coloured translucent fluid from a tall glass cup with a straw
stevan I always enjoy reading bios of authors, it gives such insight into their work
mugwump Yeah ... I'm really enjoying Hunter S. Thompson's final work
stevan spreads some almond-flavored hash butter onto his toast 14:04
kungfuftr *cough* Storable!
stevan mugwump: Curse of Lono is one of my all time favorite books
theorbtwo *cough* Ramen!
stevan HST is a friggin nutcase :)
Fear and Loathing is a classic as well
kungfuftr hhhmmm... ramen... *drool*#
stevan the movie was excellent
kungfuftr misses his yakiniku
mugwump I watched a documentary on Johnny Depp, and it didn't even mention FaLiLV 14:06
stevan mugwump: odd, that was one of his best roles (IMHO of course)
mugwump nor the time Johnny spent in Grenada or wherever it was
stevan mugwump: you ever read Phillip K Dick? 14:07
mugwump not much
stevan his later work can get as strange as Burroughs at times (but with less *ahem* boys) 14:08
rgs valis ? 14:09
stevan rgs: valis is a great book
rgs I've read "divine invasion" on sunday
stevan rgs: nice
Flow my Tears the Policeman Said, and Three Stigmata of Palmer Eldridtch are my other two favorite PKD 14:10
he is good to help un-twist your brain after a day of brain-twisting programming 14:11
rgs I concur 14:12
stevan considers a READINGLIST file to go along with the VICTUALS
autrijus go ahead and add it :D 14:13
stevan :)
obra wonders if these should be kwidfiles that get installed 14:14
perldoc Pugs::ReadingList
perldoc Pugs::Victuals
autrijus argggh no :p 14:18
stevan I think we got rid of the Pugs:: namespace too
theorbtwo Hm? Pugs:: sounds like a good namespace for things that are Pugs-specific, and not p6-general, like the PA series. 14:19
obra whatever namespace.
I just meant "Are these things useful enough to be part of the installed distribution?"
autrijus not thinking so. certainly not Victuals. 14:20
probably not readlinglist :)
castaway "use"ful? :) 14:21
theorbtwo Well, readinglist is what's suggested to read if you want to learn how to hack pugs, no? In that case, people who are reading it would want the source anyway.
obra that was my assumption. but if it's paired with victuals, perhaps it's a list of "what should I read to have my mind warped?" 14:22
stevan theorbtwo: not really, I was thinking a more general list, not just programming books
jabbot pugs - 1623 - adding documentation and more @INC paths
stevan obra: yes, thats what I am thinking,.. mind warping and such
obra ah. then yeah, I withdraw the suggestion
stevan however I think those who are so inclined will dig through and find it
theorbtwo Oh, shouldn't be installed if it's got nothing to do with what people are installing pugs for. 14:23
stevan autrijus: what is the current Haskell book recommendation? 14:24
autrijus stevan: Algorithms, School of Expressions, Craft, in that order
stevan ok
theorbtwo What would be a good intro to Haskell for mathematicians with a little programming experince? 14:25
kungfuftr hhhmmm... a more sane @INC... now that would be lovely
mugwump theorbtwo: I'm really liking _Two Dozen Short lessons in Haskell_
I was a mathematician before a programmer 14:26
castaway wonders which mathematician theorbtwo is referring to 14:27
theorbtwo My sister.
osfameron castaway++
castaway hmm, osfameron ?
ah
stevan I can't find the Algorithms book? What is the full name? 14:28
autrijus Algorithms : A Functional Programming Approach (International Computer Science Series) [Paperback]
By: Fethi A. Rabhi, Guy Lapalme
stevan autrijus: thanks
I think Higher Order Perl should go on the list (although my copy has not yet arrived) 14:30
autrijus well, if anything, it will teach you on how bad perl5 is to write complex code ;)
stevan :)
Khisanth stevan: where did you order it from? I trying to find a site that doesn't require you to create an account with them just to order a book :/ 14:31
stevan Khisanth: bookpool, but you have to create an account there
castaway how else do you want to order it? *wonder* 14:32
Khisanth gah, I hate online retailers
mugwump will be ordering a copy of the brick when he returns to NZ
Khisanth castaway: I don't know, they take my money and send me the book? :)
autrijus I wonder if the file should be called READTHEM
castaway yeah, but how do they know where to? and how did you want to check on your order, without an accont?
anyway, just find one that you can also call, and voila, problem solved 14:33
Khisanth bleh they would probably want me to register over the phone
castaway would be odd
autrijus mugwump: I heard you've added a passage from Dao De Jing to your module docs. what is that? :) 14:34
Khisanth so is requiring an account in the first place! :p
mugwump my, word travels fast, I haven't even saved the buffer yet 14:35
Khisanth castaway: checking your order is one of those nice to have things, not a MUST have
mugwump It's a very tongue-in-cheek bit for Perldoc::Parser::XML
castaway all this paranoia 14:36
stevan autrijus: The Silmarillion of course 14:37
shapr thinks castaway is stalking him
castaway meeee?
theorbtwo No, shapr, she's stalking me.
autrijus stevan: *blink*
shapr Oh, you get to be paranoid then. Lucky you!
theorbtwo No, I just trust her.
stevan autrijus: I assume one blink is yes, two blinks is no 14:38
castaway smiles
shapr mugwump: Is that the Brick Book?
mugwump none other! 14:40
autrijus and Brick part II as well 14:41
I hope
shapr I still haven't got that one.
castaway smooches theorbtwo. :) 14:42
shapr I really want to read about region allocation. Did you see the recent connection between pi-calculus and region allocation on LtU? Fascinating stuff.
Makes me wonder if there's a more general computational shape theory floating around out there.
autrijus I've read about it. sadly, without practical application, my understanding stays shallow
<- must connect theory to practice to understand anything
shapr It's vaguely possible that connection could allow formal descriptions of an algorithm to be automatically parallelized. 14:43
theorbtwo wonders -- brick book?
I know the BOLTs, but BRICK?
autrijus theorbtwo: www.cis.upenn.edu/~bcpierce/tapl/ -- the book that inspired pugs 14:44
or, in other words, pugs wouldn't have been started if not for this book (and lots of caffeine)
shapr heh
theorbtwo Ah.
stevan autrijus: that is on the list too :)
theorbtwo I was trying to think of a Tolkien book that'd be called the brick book, and failing. 14:45
autrijus "Beowulf" 14:46
cognominal autrijus: the followup on TAPL is just a compilation of papers available from the net or is it more than that?
autrijus cognominal: it is blessed with pierce's superior editorship.
some chapters are available on the net, some are not. 14:47
the point though is that the chapters flowed with a coherent narrative.
cognominal it is [retty expensive to I wanted to know if it was worth to buy it
autrijus well, if you can only buy one, buy TaPL of course
and scour the net for ATTaPL chapters
cognominal I got TAPL 14:48
and got some papers from ATTaPL
I have to read for months
shapr grins
cognominal I am happy that pugs translated that in something concrete for me 14:49
shapr At least you'll be able to buy the T-Shirt - www.cafepress.com/skicalc
cognominal and the knowledge will suffuse into me? 14:50
autrijus mmm, JRR's beowulf is 461 pages
so probably worth calling it a brick book too: www.amazon.com/exec/obidos/tg/detai...866982906/
Limbic_Region autrijus - did you see wagner.elixus.org/~gugod/pugs.png ? 14:51
autrijus oh hrm. no, that's the critics
the translation of JRR's is still in the works
why yes, I did; in fact I asked him to write SVK::Churn :)
so I can put graphs into my talk.
I did that... I think an hour before my talk
shapr Here's the back of the shirt in greater detail: www.scannedinavian.org/~shae/HM-t-shirt.png
autrijus shapr: ah. wonderful. 14:52
jabbot pugs - 1624 - adding READTHEM file (everybody add your
autrijus I believe that's the first page I found about haskell wiki
shapr QuotesPage?
autrijus by googling for ken shan (an old friend of mine) 14:53
shapr ah neat
theorbtwo sniffs as he finds that he isn't quoted on the QuotesPage at all. 14:54
shapr theorbtwo: you're on my personal QuotesPage though. 14:55
castaway and thats better! :) 14:58
shapr right!
Is there a collection of perl quotes somewhere? 14:59
theorbtwo fortune -m '\bperl\b' | wc # 226 1483 9039 15:00
cognominal shapr: www.cmpe.boun.edu.tr/~kosar/other/lwall.html
en.wikiquote.org/wiki/Larry_Wall too 15:01
jabbot pugs - 1625 - whoops, didnt mean to commit that yet 15:02
Limbic_Region autrijus - you about? 15:34
mugwump Right - I'd better get to bed, it's 3:35am in my home time zone 15:35
Unamuno hey... just an innocent question and I don't mean to start a flamewar... what's the current status of perl6? I read on the site that 6.0.0 is supposed to be done third quarter 2005. does anyone believe that? 15:36
Limbic_Region Unamuno - that statement had a disclaimer - that funding was raised 15:37
in reality, Pugs has been delivering p6 since the 2005-02-01 15:38
of course it isn't feature complete and in all likely hood, won't be the final implementation - but it is real right now
Unamuno hmmm I see
so it's more functional than other implementations at the moment? 15:39
Limbic_Region there are no other implementations ATM
PGE and Larry's work on a p5 to p6 translator are probably the only other work being done
Unamuno hrm. ok. I thought work on parrot was ongoing 15:40
Limbic_Region do keep in mind though that Parrot is the VM that p6 will run on and it is quite advanced
Parrot ne Perl6
mugwump (and pugs can compile to now, with limitations)
Unamuno well, I know that
Limbic_Region unlike p5, the VM will be a completely part of the language
and unlike p5, the regex engine won't be 15:41
Unamuno but I'm not clear on how much work will be necessary on perl6 itself beyond parrot
Limbic_Region s/part/apart/
kungfuftr yay
Limbic_Region well - there are more subject matter experts around then you can shake a stick at - I am not one of them
Unamuno and pugs has no connection with parrot, then?
mugwump pugs can compile to parrot bytecode
Limbic_Region Unamuno - Parrot is one of the backends that Pugs targets
when Parrot isn't available it can still be run through Haskell 15:42
on that note, me needs food
Unamuno hmm ok.
well thanks for the info :)
mugwump & # really going this time :)
PerlJam Unamuno: pmichaud is working on PGE which will be used to bootstrap "the" perl6 compiler.
Unamuno: pmichaud is the perl6-compiler pumpking and the one who said 3Q 2005 15:43
I believe he can do it.
And if he can't, well, there's always pugs :-)
Unamuno hmm I see, ok.
PerlJam (personally I think that there's a better chance of succeeding faster if we get pugs to grok perl6 rules and then use pugs to bootstrap) 15:44
Unamuno you know... I saw a link to ruby on rails the other day and it made me start pondering ruby. but it's been extremely frustrating -- very poor documentation, and basically no utf8 or unicode support whatsoever, with no signs of it arriving anytime soon. 15:54
PerlJam Unamuno: yep, that's the same impression of ruby I've had since I first tried it a few years ago. 15:55
Khisanth and it's STILL the same impression people are getting?
PerlJam apparently.
Khisanth so why are there complaining about the progress of perl6 again? :)
integral Perl isn't much better, it seems every time SVK is mentioned people say "but it's written in Perl" 15:56
Unamuno and unicode is essential for what I'm doing. no room for quibbling. I need stuff like Unicode::Normalization, as well as the ability to split by characters, and convert chars to codepoint numbers and back. none of that is possible in ruby -- almost boggles the imagination
PerlJam I'll never forget my first ruby experience. Every single error was "syntax error". That's it. No other useful information.
Khisanth people keep pointing to things as being better but in end they aren't really better
ninereasons autrijus, ping 15:57
Unamuno what mostly frustrates me about perl lately is two things: reference/dereference is rather ugly and can be hard to write cleanly, and "real" classes/structs don't exist. of course I'm not sure _how_ "real" I want them to be, but a bit realer than they are now ;)
Odin- PerlJam: Didn't they change the 'official' plan to use pugs?
PerlJam Unamuno: perl6 fixes those, so get on board and start helping! ;-) 15:58
Unamuno PerlJam: I know, that's why I was asking about it :P
PerlJam Odin-: Not to my knowledge.
Odin- Hm. Pugs directed at Parrot, using PGE ... so as not to duplicate (too much) effort?
Khisanth but do you really have to be stuck with an image of what a "real" struct or class is?
Odin- PerlJam: Hm. Okay. :/ 15:59
PerlJam Odin-: It wouldn't surprise me though. Once PGE is functional, we could just hook it into pugs and bootstrap that way without coming up with anything fancier.
Odin- Exactly...
Unamuno Khisanth: all I want, really, is to be able to separate public from private, and to explicitly declare a sub as belonging to a class (so it knows its $self automatically)
PerlJam The key thing is being able to parser grammars and "execute" then and be able to modify the grammar on-the-fly 16:00
s/parser/parse/
Khisanth basically something to save the programmer from him/herself :)
and other people they are working with 16:01
Unamuno ruby has very good and interesting support for objects -- but it takes it a bit too far, in my opinion.
for example, I find it very strange that you can do things like 3.plus(2)
PerlJam Unamuno: heh ... perl6 will do that too
Unamuno noooo :P 16:02
PerlJam well, something similar, not that exact thing.
integral it's more like (3, 2).plus in p6
PerlJam Unamuno: what problem do you have with that? It seems like a natural thing to do in a first-class object oriented language. 16:03
Unamuno and python is just.... ummm.... I dunno what. I find the community to be very centrally controlled and the opposite of perl (there's only one way to do it)
PerlJam (python and ruby and perl are all headed towards being first-class)
shapr Haskell is a lovely scripting language =) 16:04
autrijus (3,2).infix:<+>;
Unamuno PerlJam: well, of course. the question is whether such languages really make sense in the end. I'm not totally sure on that yet. in many respects I don't mind "everything is an object", but some attention needs to be given to problems that are solved best other ways, and it's annoying when it becomes impossible to do so.
Odin- I've said it before, but to me it looks like Perl is really getting to be like a sort of "lisp with syntax"...
integral shapr: ugg. so *slow* to compile though ;-)
autrijus Odin-: Ruby is totally that already 16:05
integral: hugs :D
ninereasons autrijus, what have you done that's caused the compile times to drop so radically?
PerlJam Odin-: Isn't that haskell? ;)
shapr integral: I use hs-plugins for compilation.
autrijus ninereasons: I don't know; dropping 6.2 support and taking -O off Parser.hs
integral oooh, more things to read =)
Odin- PerlJam: Dunno. Maybe. Don't have time to get into it, I'm already waaaaay behind in my school stuff. :(
ninereasons I've gone from 188m to 25m on my slowest machine. 16:06
shapr 25minutes!
autrijus that sounds like a good improvement.
ninereasons 8m on my fastest machine - optimized!
PerlJam Hmm. Maybe I should try compiling pugs on my laptop again.
theorbtwo autrijus: The more I look at it, the more I'm thinking that DynamicLoader isn't going to work and hs-plugins is the way to go.
Khisanth Unamuno: just because a language allows you to do something doesn't mean you have to ... 16:07
Unamuno Khisanth: yes, but if a language doesn't allow you to do something, that means you can't :P
theorbtwo DynamicLoader doesn't use the .hi info, so we have to do all the dependency tracking yourself.
autrijus theorbtwo: okay. sorry for sending you on the chase
theorbtwo It's OK; I learned from it.
autrijus hrm, does NamedLoader not do .hi parsing?
Khisanth Unamuno: no it just means you have to implement it yourself
autrijus I vaguely remembered it did. 16:08
but if it does not, then totally drop it
theorbtwo The paper talks about it, but I don't see any code for it.
autrijus and switch to hs-plugins
right. I havn't been reading the code.
PerlJam Hmm. pugs compiles in 48s on my box. /me trys to gauge how slow ninereasons boxes are.
Unamuno Khisanth: it depends what you're talking about. say I want to write a short ruby script where I call a function that's defined later in the document. how do I "implement that myself"?
Khisanth Unamuno: and you were complaining that Ruby allows you to use 3.plus(2), not that it doesn't allow you to use it :)
integral memory seems a big factor too, PJ 16:09
autrijus oh. right. I dropped -H200m.
so if your machine has <200m of memory
then that's going to help a lot.
PerlJam < 200m free you mean?
autrijus right. well if you have <200m physical, then that would made pugs impossile to compile 16:10
or at least a certain adam preble reported such
integral s/physical/physical + swap/
PerlJam I think my laptop only has 256
that would explain why it took so long when I tried it before
autrijus nod.
sorry for that. :)
Unamuno Khisanth: well, no, my point was that the main focus in ruby is on "everything is an object" and so in reality 3+2 is just syntactic sugar. now, that one is so obvious that they made a way to do 3+2 as well as 3.plus(2), but I just get the feeling many other cases have been ignored, where doing things in a more function-oriented way make sense
autrijus but I did not know -H0 at that time.
now it's -H0, and it seems everyone is happy 16:11
ninereasons PerlJam, they're older machines; but they compile in comparable times to yours if I don't say "make optimized"
integral there's also ST-72's fun of 3 + 2 ==> (3).send('+', 2);
Unamuno Khisanth: it's not my fault that you interpreted my unclear ramblings incorrectly :P 16:12
autrijus <math:plus><arguments><argument><math:literal>3</math:literal></argument><argument><math:literal>2</math:literal></argument></arguments></math:plus>
shapr I didn't get much use from the second cpu - "make -j3 83.93s user 2.81s system 90% cpu 1:35.86 total"
Unamuno autrijus: ahhh bliss, where would we be without xml ;)
autrijus shapr: dons suggested ghc -M before
shapr ah, good idea
integral <tag name="plus" ns="math"><tag name="arguments"><tag name="argument">...
PerlJam -M ?
Khisanth Unamuno: a much better place! isn't that obvious? :p 16:13
shapr dumps dependencies into the makefile, I think
autrijus shapr: however I'm out of tuits to do that as it's not a showstopper
so, helps welcome, but otherwise won't make to 6.2.0 16:14
shapr yeah, I dig
I doubt there are a bunch of SMP pugs users.
Btw, when's 6.2.0 going out?
Unamuno Khisanth: hehe... I was looking up xml shorthand formats the other day. I guess I don't understand tabs are evil, linebreaks are evil, etc... I suppose it depends on what you're encoding the xml for. but for my purposes it's just way too verbose. 16:15
and for many others, as evidenced by the dozens of shorthand proposals
shapr If 6.2.0 is going out in the next day or two, I think I saw a bunch of simple builtins I could write in a few hours.
obra casey! 16:16
have you ported all of Email:: to p6 yet?
Khisanth Unamuno: it's ok as long as you aren't writing it :)
shapr kungfuftr: css is gone on smoke.
autrijus shapr: go go go :)
6.2.0 is due this sunday. 16:17
or monday morning.
but that's my time, so it's sunday for almost all of you :)
shapr Cool, I'll be polite and let others have a crack at simple builtins for a another few hours, after that it's open season. 16:18
PerlJam shapr: which simple builtins are you talking about specificly?
Corion was working on sleep yesterday for instance. 16:19
autrijus don't be polite. polite is not a virtue around here. just speak your mind on the channel and feel the anarchy :D
obra giggles
theorbtwo shapr: Mind giving me a hand with hs-plugins? It doesn't want to build for me. 16:20
Unamuno sorry, this is 6.2.0 of what?
cognominal discovers where autrijus got his pugs graphic roadmap
PerlJam Unamuno: pugs
theorbtwo pugs.
Unamuno ahh ok. kinda confusing considering it's not even perl 6 yet :)
shapr theorbtwo: for the best of hs-plugins you need the latest nightly from ftp
autrijus cognominal: yup. I added a link at first on haskell.org/hawiki/AutrijusTang
i.e. research.microsoft.com/Users/simonp...tive-2.pdf
theorbtwo It is perl6. It's not completed perl6. 16:21
autrijus but that link is a bit long, so I didn't add it to the png itself
shapr It's a brilliant and complimentary parody of the hairshirt retrospective.
autrijus Unamuno: yeah, but imagine the confusion if I called it 5.0.0
Unamuno theorbtwo: right, but not completed versions of a project generally don't have version numbers higher than completed ones. that's all I meant.
shapr PerlJam: last I looked, several of the missing builtins were simple wrappers around the equivalent Haskell functions.
autrijus now I can also call it 0.0.1 ... but I was high on caffeine and that looked no fun. sorry for my indulgence :) 16:22
Unamuno hehe
autrijus: you should have successive versions converge on 6.
autrijus starting from 5.9.0?
even more confusion 16:23
Unamuno no, from 6.2.0, now that it's already there
autrijus say what?
6.2.0, 6.1.9, 6.1.8... ?
PerlJam Unamuno: that's going quite the wrong way
Unamuno yeah :)
autrijus makes no sense :p
Unamuno well, he's the one who started it by calling it 6.2
PerlJam Unamuno: pugs is converging on 2*pi
Unamuno and there's a tradition of obfuscated tendencies in the unix and open source communities, so... 16:24
oooh ok.
well, I think my idea shows more creativity ;)
shapr PerlJam: what about trig.t ?
theorbtwo Thanks, shapr. 16:25
Khisanth creative to the point of confusion is no good :p
Unamuno it's only confusion until people catch on to it :P 16:26
heh, I can just imagine the flow of questions on the channel here though
theorbtwo Unamuno: Didn't this start because you thought the existing convention was already too confusing?
Unamuno yes, but right now it's confusing with no apparent logic. my way would at least have a logic to it, although a strange one.
PerlJam no apparent logic? 16:27
Unamuno um... well is there a reason an implementation of perl6 that's not complete should have a version number of 6.2?
Khisanth 2*PI is quite logical whereas a decreasing version number will mess up just about every version tracking system in existence if not all of them 16:28
theorbtwo The logic is perfectly obvious, if you remember that it's perl6, and pugs 6.2.
PerlJam Unamuno: you mean 6.28318530717959
theorbtwo And in any case, 6.2 == 6 (when you remember significant digits).
Unamuno well 2*pi is perfectly easy to calculate, but I don't see what pi has to do with perl, or why it's being multiplied by two. 16:29
whereas with tex, converging on pi fits a lot better.
theorbtwo: and yes, I realize that. there essentially is no "real" perl6 outside of a description, right? 16:30
I guess the problem is that there's basic only ever been one reference perl implementation before.
s/basic//
theorbtwo See dev.perl.org/perl6/doc/design/apo/A01.html, under RFC 141.
(See also RFC 141 itself.) 16:31
kungfuftr shapr: fixed! need to really sort that out 16:33
autrijus Unamuno: basically I've been attracted to P6RFC 141 ever since it was written 16:34
and there's also this "need to mark pugs as not the official perl 6 compiler" thing. 16:35
when I started versioning pugs, that is
Unamuno autrijus: incidentally, why did you choose haskell? 16:36
autrijus Unamuno: pugscode.org, "Overview"
grep for "Haskell"
Unamuno ah ok
PerlJam Unamuno: obviously it was the right choice. Look how much work has been accomplished in such a short time :-) 16:37
mj my new toy, see 16:38
wiki.kn.vutbr.cz/mj/attach/pugs/tut...ld.p6.html
and next>> links
autrijus oh wow. wow. 16:39
is this what I think it is?
a pugs-powered html tutorial generator? 16:40
PerlJam it's magic!
autrijus majic!
Unamuno PerlJam: well, it does seem to be quite a lot for a short time :) 16:41
autrijus: it doesn't say on that page when exactly you started... just a couple months ago?
autrijus Unamuno: 1st Feb, 2005
PerlJam Unamuno: Feb 1
Unamuno I'm curious to try it out now. time to compile ghc... sigh ;) 16:42
autrijus there should be binary installers for your platform... 16:43
PerlJam Unamuno: make sure you have ghc 6.4
Unamuno PerlJam: I know... I'm on OS X and fink has binaries for 6.2, which I already had for something else. but the easiest way for me to install 6.4 is through darwinports which compiles everything
gaal autrijus, where can i find the slides to your yapc talk? they seem to have gone off the main site
the main yapc::taipei site that is 16:44
mj thanks, but now i need to do real work :-(
autrijus gaal: wagner.elixus.org/~autrijus/p6tu/
gaal thanks!
autrijus Unamuno: the easiest way is thru .dmg :D
Unamuno I might be able to find binaries for 6.4 but it doesn't seem work it, I prefer proper package management... I'm not in a hurry anyway :)
s/work/worth/
autrijus or rather, for 10.3, haskell.org/ghc/dist/6.4/MacOSX/GHC-6.4.pkg.zip 16:45
Unamuno autrijus: where does it install?
autrijus /usr/local/bin/ghc
I think
Unamuno yeah, that's what I don't like about pkgs. they ask for your password and don't tell you what they're doing.
autrijus uh. they are just tarballs.
you can cd into them and use tar to inspect its content.
Unamuno errr... not usually. 16:46
mj html tutorial generator, but writteln in perl5 with system( 'pugs.exe', ...)
autrijus *shrug* that's what I do anyway :)
PerlJam mj:you should write it in perl6
autrijus mj: I wonder if you can port it to pugs
Unamuno autrijus: hmm well I'll try that in the future, but I haven't been able to the couple times I tried 16:47
incidentally, I find it a positive trend that people on freenode don't automatically deride people who use OS X ;)
mj maybe, but not today perhaps, source code wiki.kn.vutbr.cz/mj/attach/pugs/tut-root/tut/ 16:49
autrijus Unamuno: SubEthaEdit powers a large chunk of pugs development.
PerlJam Unamuno: yep. I've never been a Mac fan, but when a coworker gave me and old G3 with OSX on it, I've never been happier.
s/and/an/
I don't think that would have happened prior to OSX though 16:50
anyway ... back to work
Unamuno autrijus: really? hmm that surprises me somehow. I used SubEthaEdit for a while but my work recently bought BBEdit, which is still the best, imo. 16:51
autrijus Unamuno: BBEdit doesn't get you teamwork
Unamuno PerlJam: I've always been a mac user. long time on unix too though and ran linux for a while until OS X made it mostly not so necessary. but I have never, ever used windows regularly :)
although I've had to use it enough that I know how to bend it into shape 16:52
autrijus unless of course BBEdit somehow gained teamwork capability I did not notice 16:53
Unamuno autrijus: nope. I don't need that feature personally, though. not at the moment anyway. 16:54
autrijus ah. have you tried it at all? it's highly addictive.
Unamuno autrijus: oooh, I misread your first comment -- I thought you said SubEthaEdit *funds* pugs development.
autrijus uh, no. "fuels" maybe
BBEdit is good to us, too -- they agreed to add a Perl6 syntax mode on next release 16:55
Unamuno does it recognize haskell?
autrijus SEE? I don't know, prolly not by default
but I think setting the syntax to "SQL" gets you a good haskell mode.
;)
Unamuno hehe. how does the collaborative stuff work anyway? I suppose it works over the net as well as local networks... and you see what people are doing in real time I guess?
16:57 computer is now known as errr
Unamuno BBEdit has a surprisingly good perl syntax mode. very good at catching missed quotes for example and it even recognizes here documents. 16:59
Limbic_Region chip about? 17:02
www.perlmonks.org/index.pl?node_id=445661 # Is Perl6 going to release memory back to the OS
chip *bamf* 17:03
autrijus Unamuno: right, it's basically multiplayer notepad
Unamuno chips about
Limbic_Region www.perlmonks.org/index.pl?node_id=445785 # What compilers are targetting Parrot
autrijus L_R, our resident perlmonk channeler
Limbic_Region well - use.perl journals as well, but you seem to read them before I get a chance to tell you about them autrijus 17:04
autrijus yup.
the volume is... more managable than perlmonks
Limbic_Region hmm - /me thinks by chip's *bamf* reply, he really isn't here and that is an automated response
autrijus do not be fooled! 17:05
autrijus tries the summoning again
chip about?
Limbic_Region seen chip
jabbot Limbic_Region: chip was seen on Fri Apr 8 01:03:06 2005
autrijus see? not an automated response
Limbic_Region well, I /msg'd him the links anyway 17:06
autrijus - are you familiar with Dan's "What the heck is X" series from his blog? 17:07
I was thinking that "What the heck is bootstrapping and what does it have to do with Perl6" would be a good thing to give an explanation of 17:08
Unamuno autrijus: the ghc binary is not a tarball, it's a paxball ;)
Khisanth hmm 17:09
Limbic_Region: maybe the memory releasing will be the same as perl5? 17:10
oh wait that still depends on the OS
Limbic_Region Khisanth - on the memory releasing issue - it is a complicated answer
If an OS doesn't support, the language can't do much about that
on the platforms where it is supported, the language is going to have to work to some extend to make the OS happy about releasing the memory back 17:11
in p5 it almost never makes sense to do that because it doesn't use contiguous blocks (AFAIK)
in p6/parrot, it may be more realistic, but in p5 we often make optimizations by trading resources 17:12
if I asked for memory once - I will likely ask for it again, so let's keep it around
Limbic_Region is by far not a Subject Matter Expert though - which is why I pointed it out to chip 17:13
s/extend/extent/
bootstrapping is a much more straightforward matter
autrijus Limbic_Region: yes, quite 17:14
wrt dan
Limbic_Region: uhm, hm, yeah, write one?
(or scheme chip into writing one)
Limbic_Region well, I was thinking about that style, but have a section at www.pugscode.org
Khisanth well hopefully perl6 will have slightly less overhead :)
when storing things
Limbic_Region I would be willing to framework some - but not being very smart, someone else would have to fill in details 17:15
autrijus sure. PA01 has some initial details
Limbic_Region well - I see it as being a more in depth FAQ
but I have a lot of grand ideas sitting back not actually doing anything
I will come up with an outline for bootstrapping with some basic content I know is accurate and let you decide if it is worth it 17:16
s/you/the channel/
autrijus cool!
Limbic_Region++ # in advance
Limbic_Region er um, now the format - POD would be easiest for me 17:17
autrijus go ahead and use POD.
if POD<=>Kwid converter is not there when you're done, it's not your fault :)
there is a Kwid=>Html converter though, and a Html=>Pod one 17:18
but that's lossy as hell.
theorbtwo All converters loose. Some just loose less. 17:19
autrijus (Spork::Formatter::Kwid and Pod::HTML2Pod respectively)
Juerd How lossy is hell?
autrijus total entropy, man
theorbtwo But total entropy increases over time! 17:20
autrijus until it reaches hell 17:21
then it freezes over and starts anew
stevan autrijus: I dont wanna be a pain, but any idea when multidims will land?
Khisanth so hell is not hell until the end of hell? :)
autrijus stevan: you are not a pain. before 6.2.0. I started doing designs but was distracted by business stuff 17:22
stevan: so most likely tomorrow morning
stevan autrijus: sounds good to me (your morning is usually my late evening, so thats perfect)
autrijus Khisanth: right, see Asimov, "The Last Question"
also # interconnected.org/notes/2003/05/last-query.html 17:23
stevan: the prerequisite -- IntMap change -- has landed, but I need to tweak Eval.hs some more 17:24
kungfuftr rightio... time to overhaul the testsuite stuff 17:25
stevan kungfuftr: why?
kungfuftr stevan: modularise it and make it a little more generic
ie: not have 5 different scripts run across it 17:26
stevan kungfuftr++
it would be a very nice thing to have for all my (soon to be antiquated) perl5 projects
kungfuftr i should have a peek at yacsmoke again too 17:28
Limbic_Region autrijus - isn't there a pugs wiki? 17:29
maybe it would make more sense to add a "What the heck is" section there?
stevan Limbic_Region: pugs.kwiki.org/
Limbic_Region thanks stevan
stevan Limbic_Region: always a pleasure :) 17:31
kungfuftr autrijus: btw, had a chat with merlyn... as far as he's aware, there's nothing to stop someone doing a "Learning Perl" companion guide thing for Perl 6 17:32
stevan kungfuftr: did you see the idea for Perl6 Cookbook?
autrijus pleac! pleac!
kungfuftr: I think pleac will rock. 17:33
chip Professor Backwards drowned after yelling "PLEH! PLEH!"
autrijus and a certain publisher may even consider publishing it.
stevan thought autrijus was chooking on his Diet Coke there for a second
autrijus # pleac.sourceforge.net/pleac_perl.html
eh. ;)
I'm officially off Coke for now, though. 17:34
kungfuftr stevan: yup... learning perl's quite a nice step though... especially for the very basics that have changed
autrijus tea for me on usual days.
Limbic_Region after re-reading what has been written about bootstrapping, I don't see it as a good candidate for the first "what the heck is" - I will just point people there and wait to see what the next thing is most initiates are puzzled about
autrijus Limbic_Region: that's cool too.
stevan finds the $work server is still unresponsive, so back to Pugs hacking it is :) 17:35
ninereasons how convenient for us, stevan
stevan kungfuftr: pleac may be a better place to start though. It will help crystalize the differences
kungfuftr can't believe that's no good generic smoke packages on CPAN
stevan ninereasons: :) 17:36
kungfuftr stevan: true... but a learning perl companion guide thing would be so sweet to have when pugs/perl6 gets production ready
Unamuno smoke?
autrijus kungfuftr: is learning perl's code availalbe on the web? 17:37
hi rgs! 17:38
rgs hi autrijus
stevan kungfuftr: either way is good :)
ninereasons mj, your sample is nice to show to perl5 people, at wiki.kn.vutbr.cz/mj/attach/pugs/tut...pleiter.p6
things like that make perl5 people say "yay"
kungfuftr checks 17:39
Juerd string.say still looks weird.
autrijus for (%mates.kv) -> $girl, $guy {
this can lose the parens
chip are the parens required in the simpleiter example?
jinx
autrijus kinx
linx too.
kungfuftr autrijus: examples.oreilly.com/lperl3/ 17:40
autrijus that reminds me. anyone finds the trailing comma thing irritating enough? 17:41
i.e. considered 6.2.0 blocker?
(1,2,3,) # currently illegal in pugs
kungfuftr autrijus: www.oreilly.com/pub/a/oreilly/ask_t...olicy.html
ninereasons more irritating, is that my ($s,@a) = <1 2 3> doesn't work 17:42
kungfuftr ninereasons: ($s,*@a) = <1 2 3>; ?
chip flattens kungfuftr 17:43
kungfuftr mwuahaha
autrijus no, = should flat by default.
i.e. that's a bug.
although I was seriously confused when I first coded it.
so needs a rewrite :)
metaperl_ does that term "flatten" make sense to everyone but me?
autrijus (of 10 lines of code -- line 333 of Eval) 17:44
kungfuftr autrijus: a trailing comma might denote an undef value perhaps... but probably better to be explicit about it than not
Juerd simpleiter.p6 is not homosexual compatible.
autrijus kungfuftr: the trailing comma is ignored in p5
and same in p6
kungfuftr k
autrijus larry also said that trailing semicolon is ignored too
metaperl_ were those angle brackets valid syntax?
chip <> is qw() <<>> is qqw() [or would be]
autrijus angle brackets are the new qw<> 17:45
they are also the new {''}
Juerd autrijus: {qw//}, actually.
autrijus well, yeah, sure
chip Hm. I wonder what $a{qw/1/} does
it ... works. hm 17:46
Juerd chip: That's the same as $a{1}
chip so it is. I expected qw{} to react poorly to scalar context.
Juerd chip: Which is %$a{1}
chip: It's not scalar context afaik.
metaperl_ chip: perldoc perlop discusses qw and qq but now qqw .... whatis qqw?
Khisanth like qw but with interpolation? 17:47
Juerd metaperl_: q : qq :: qw : qqw
metaperl_: Apply logic.
chip Oh wow. Perl 5:
perl -MData::Dumper -le '$a{qw/1 2/} = 3; print Dumper(\%a)'
$VAR1 = {
'12' => 3
};
Juerd hadn't expected that
ninereasons hm. surprising
chip Even better 17:48
rjbs chip: You made that up!!
chip perl -MData::Dumper -le '$a{qw/1 2/} = 3; print Dumper(\%a)' | cat -v
$VAR1 = {
'1^\2' => 3
};
It's doing the fake multidimensional thing
Juerd chip: Ahh, $;
rjbs ahhh
Juerd always forgets about that until he needs it
chip forsees a feature of ObfuPerl
Juerd Well, not need, but want.
rjbs never wants it.
Juerd rjbs: When I want something to be keyed by multiple things that only in relation to eachother make sense.
chip anyway, I'm noodling on the wrong channel. AWAY! 17:49
Juerd rjbs: It's a shame to waste lots of hashes on that: $foo{$bar}{$baz}{$quux}, when %{ $foo{$bar} } makes no sense
Then $foo{$bar, $baz, $quux} or $foo{"$bar/$baz/$quux"} (I use the latter more often because I think it's clearer) fixes that. 17:50
kungfuftr bah, i so need a unix box at home
Juerd kungfuftr: Then get one.
kungfuftr Juerd: no as easy as it seem unfortunately 17:51
Juerd Is too.
kungfuftr: What box are you using now?
(at home)
crysflame win 3 17:53
wow
sorry
if an op sets mode +c i believe that strips colors and bold from the channel
shapr yes 17:54
crysflame wow, 98 of us
kungfuftr Juerd: win xp box... development done on a freebsd remote box
Juerd kungfuftr: Then run cooperative linux. 17:55
kungfuftr linux-- # sorry
Juerd zealot.
kungfuftr would move to OSX if he had the money 17:56
rjbs kungfuftr: What's your roadblock to having a unix box at home? 17:57
Juerd rjbs: hating linux, apparently.
rjbs Clearly he could just run freebsd at home, too.
So presumably the problem is with hardware acquisition. 17:58
Juerd Not at the same time as Windows XP
On the same box
(If you can afford vmware, you can also afford a separate box, as the latter is cheaper even.)
kungfuftr rjbs: wireless only and my wireless card is not bsd compatible... however forgot about NDIS, so might do a 5.3 install
shapr yay, pugs on lwn.net
lwn.net/Articles/130401/ 17:59
ninereasons cook seems to be the designated pugs reporter 18:00
shapr ?
ninereasons hm. I guess I don't know what I'm reading there: looking at Search results 18:01
on lwn.net
shapr who's cook?
ninereasons ([Development] Posted Apr 5, 2005 21:40 UTC (Tue) by cook) 18:02
I don't know - it doesn't show up on the post. but "by cook" is in the search result 18:03
shapr interesting
autrijus mm 5 people hacked on hangman.p6 aleready 18:05
making it the most popular example
ninereasons or, considering that I'm one of those five, making it the most hacked up sample. 18:06
Line 333 ff of Eval.hs is still gibberish to me. I need to read more haskell books. 18:07
castaway .. but vmware can be any number of machines, so offsetting its price against one is not fair ;) 18:10
shapr ninereasons: you want an explanation? 18:14
ninereasons yes. I'd like to understand why it doesn't do what it's supposed to do. 18:15
shapr This? "[lhsExp@(Syn "," lhs), exp] -> do"
ninereasons yes, and following
shapr You understand case?
autrijus line 336 is the entire bug, actually, but I'm journaling, so shapr++ :) 18:16
shapr !
ninereasons assume I know nothing, and you'll be very close to the truth.
shapr fair enough
castaway (wtf is that, haskellish?)
autrijus castaway: yes, src/Eval.hs line 332/333
hey alin :) 18:17
castaway still hasnt persuaded gentoo to install ghc 6.4
shapr let's say you have some value, you can switch on the value with case. case somestring of "foo" -> putStrLn "bar" ; "bar" -> putStrLn "baz"
that sort of thing.
ninereasons ok
alinbsp good evening
ninereasons case, switch, that kind of thing 18:18
autrijus actually line 335 is a lie. "slurpy" should read "non-slurpy" ;)
but it really should be slurpy. 18:19
shapr the first case happens if you get a two item list where the first item uses a Syn constructor with a string "," as its first value
Khisanth autrijus: that is very confusing
shapr it also assigns the whole value inside that Syn constructor to the name lhsExp
autrijus Khisanth: indeed. I was very confused when I coded it up.
shapr just from the names, I'd guess lhsExp means left hand side of the expression 18:20
and exp probably means expression
autrijus my ($a, @b) = (1,2,3);
castaway as long as it works, autrijus
autrijus here $a and @b are the lhs
and lhsExp is ($a, @b)
exp is (1,2,3)
ninereasons omg, i think I actually understand
shapr mapM_ is just like map with the addition that it's monadic, (mapM) and it doesn't care about the result (the trailing underscore) 18:21
autrijus ...and you see a GHCism, parallel comprehension
shapr so evalExp gets applied to each part of the list that is its argument
ninereasons one item at a time 18:22
shapr that funny looking list critter is a list comprehension
@plugs [(x,y) | x <- [1,2,3] , y <- [4,5,6] ]
that's your basic list comp 18:23
ninereasons nice trick
autrijus @plugs [(x,y) | x <- [1,2,3] | y <- [4,5,6] ]
argh.
shapr grins
autrijus why can't lambdabot enable -fg-exts?
think of the children! :)
wilx :)
shapr because a few people found ways to execute shell commands via -fg-exts
autrijus oh wow. wow.
ok. good enough
shapr yeah, I hope no one on #haskell ever turns black hat 18:24
> [(x,y) | x <- [1,2,3] | y <- [4,5,6] ]
[(1,4),(2,5),(3,6)]
since I have both irc and ghci in emacs, I can just cut'n'paste
autrijus nice trick 18:25
shapr I'd use zip for this normally
@plugs zip [1,2,3] [4,5,6]
shapr blinks
autrijus sure, but I find parallel comprehension easier on my brain :) 18:26
shapr hey if it works for you :-)
ninereasons autrijus has an interesting brain
autrijus basically I can't remember zip2, zip3, zip4
shapr Also, standard zip only goes up to zip3 I think 18:27
oh, there's zip4?
autrijus since in my mind it's the same thing
shapr anyway, it's limited.
I doubt parallel list comps are limited.
autrijus they are not
but we digress.
anyway, that parallel comprehension takes lhs, zips it with the evaluated list of rhs, cutting the longer one to fit the shorter one 18:28
shapr so, for each piece on the left side and each piece on the right side, they get combined like Syn "=" [l, Val v]
autrijus ...nice, we actually complement each other
shapr bea says hi to autrijus 18:29
autrijus leaves the rest to shapr and goes on journaling
shapr (my significant other)
18:29 Aankh|Clone is now known as Aankhen``
shapr ninereasons: was that everything? 18:29
ninereasons I followed your explanation, 18:30
autrijus bea: hi!
ninereasons I'm digesting.
shapr ninereasons: is the rest of that case understandable?
autrijus ninereasons: so, line 337 evals the lvalue again
and return it
that is so that
$a = $b = 3
works.
shapr runs off to take a post-unicycling shower. 18:31
ninereasons ok. yes
autrijus the "enterLValue" means that the retval is good as lval
(($a = 3) = 4) works
wilx Hmm, P.K.Dick? I almost bought the book today. But then I decided for something easier to read :) 18:35
ninereasons (($a = 3) = 4) works ?
autrijus I just checked and it doesn't ;)
wilx just READTHEM.
autrijus something else is blocking it.
file a bug... I mean test? 18:36
ninereasons I would but I don't see why it should work 18:37
autrijus well, it works in perl5
that's a reason good enough.
ninereasons ok :)
so it should evaluate to "4" ? 18:39
autrijus yes.
ninereasons is there some wonderful use for that syntax, autrijus ? 18:43
autrijus chomp($_ = <STDIN>); 18:44
ok, not wonderful (actually terrible)
but common.
ninereasons ok. i see that.
autrijus yay I get to use the word "scintillating" too! 18:47
I've always wanted to use that after I see it repeatedly on perlmonks.
chip ($a = 1) =~ s/1/2/ 18:52
autrijus chip: I read somewhere that you thought 18:54
for my $item (@array) { ... }
is a regrettable hack of yours
is that true? do you consider
for @array -> $item { ... }
to be significantly better?
chip autrijus: I regret the double-keyword effect; it moves the variable even farther from the left.
ninereasons weird, chip. why doesn't strict complain? 18:55
autrijus ninereasons: because p5 is evil and must die.
(sorry. $a and $b special casing being one of my pet peeves)
chip autrijus: The p6 version is better in context; in p6, what follows the for is always the list. if you want the variable, you hunt for the { and look to the left.
autrijus: That said, I don't regret doing it; given p5 as the base, it was probably the best bad hack available. 18:56
autrijus ninereasons: $c will raise exception
ninereasons weirder yet.
then, why should similar behavior be reproduced in perl6, autrijus
autrijus ninereasons: $a special casing? definitely not! 18:57
($a = Foo.new).process
Limbic_Region ninereasons - you do realize what special casing autrijus is referring to right?
autrijus is a valid syntax though.
Limbic_Region s/realize/know/
ninereasons only from the example, Limbic_Region 18:58
Limbic_Region $a and $b get special treatment because of sort blocks in p5
they pass strictness tests without being declared as a result
and a few other checks as well
autrijus which leads to subtly buggy code.
ninereasons I see. thank you for the context; Limbic_Region
chip with $a and $b, Larry inched toward $^a $^b, which I like a lot more
autrijus verily. 18:59
kungfuftr $^<FOO> denotes a placeholder variable, right? 19:02
theorbtwo s/which leads/which sometimes leads/
autrijus kungfuftr: right. or implicit var.
theorbtwo It's more like an undeclared argument.
autrijus aye. true.
kungfuftr autrijus: any idea if perl6 is going to have a way programmable way of identifying a specific object? 19:03
autrijus kungfuftr: elaborate?
kungfuftr autrijus: say for example for easy use with persistence
autrijus you have =:= 19:04
but I don't know what exactly are you looking for
kungfuftr autrijus: yar, was thinking more like primary key melarky
autrijus StableName? UUID?
it's all negotiatable (on p6l) :) 19:05
kungfuftr yar, that sort of thing
autrijus if you talk to mugwump, it's likely that he will design it in
kungfuftr mugwump: oi ye bastard!
autrijus since his design is the .hs pugs actually runs... :)
shapr sneaks half of prolog into pugs
autrijus ooh I want the other half too 19:06
shapr laughs
I'm just being a smartass, as usual.
theorbtwo IIRC, there was supposed to be a .ident method.
autrijus interesting. reference?
theorbtwo Now you're asking hard questions.
autrijus guess that's my job :)
oh btw, see "First Set" svn.perl.org/perl6/doc/trunk/plans/...n_docs.pod 19:07
Limbic_Region theorbtwo - do you remember the context at least surrounding the .ident method?
autrijus also see "Candiates for delegation" and that means You!
Limbic_Region I might be able to help recall the reference if you have more context - it is fuzzily in my mind as well ATM 19:08
theorbtwo I think it was the conversation that the =:= operator came from. 19:09
Limbic_Region hmmm - nope, not ringing a bell
I was thinking it had more to do with a mutiple inheritence discussion
theorbtwo Conversations on p6l are always so twisted -- it may have been both. 19:10
Limbic_Region well, I tend not to follow p6.l so it is likely that I heard a different conversation concerning the same method
kungfuftr autrijus: it was just me thinking about the fact that a fully typed perl6 class could be automagically serialised to postgresql 19:12
autrijus mugwump of the Tangram clan will likely agree with you. 19:13
kungfuftr autrijus: yar, sam and i used to work together... lots of tangram voodoo 19:15
autrijus mmm voodoo
what do you think of the opposite clan, the Prevaylor voodoo? 19:16
kungfuftr prevaylor?
autrijus www.prevaylor.org
kungfuftr seems dead
autrijus wrong link 19:17
www.prevayler.org/wiki.jsp
kungfuftr autrijus: ah yes, it's nice, but not always practical, especially in production 19:18
plus, you don't have as much control in general of things like transactions, etc. 19:19
autrijus yeah.
kungfuftr and also, SQL can be a lot more powerful in certain areas 19:20
autrijus right. and with sqlite the speed is acceptable 19:21
to me at least.
I've still yet to find a practical use case for prevayler.
kungfuftr autrijus: cost for resiliant hardware would be my initial concern 19:22
tangram, cdbi, etc. are all fine enough for most uses 19:23
autrijus aye. 19:24
ok, journal up. need to be sleeping :)
see you tomorrow *wave* &
kungfuftr nn!
ninereasons thanks and bye
autrijus YAPC::NA Hackathon Update: Breaking news: Larry Wall and Luke Palmer are also coming. Wow. 19:26
zzz &
kungfuftr oh... sweet 19:28
Limbic_Region autrijus for sharing the great news, autrijus-- for falling asleep without fielding questions 19:29
chip Larry? Neato!
Is that cottage going to hold everyone?
Limbic_Region hopes the journal has more details
chip - you are going to right?
chip 0.9 probability. I sure want to!
0.98 probability of attending YAPC proper 19:30
Limbic_Region my wife can't yet leave the country, so this year will be out :-(
kungfuftr only gets to go to YAPC::EU this year 19:32
Limbic_Region autrijus - if you see this when reviewing the logs, read use.perl.org/~milardj/journal/24057 and don't be modest about the praise being heaped on
kungfuftr goes off to install freebsd 19:36
stevan cool, larry and Luke at the hackathon 19:38
rgs woot 19:39
stevan it will certainly make it interesting :)
stevan really needs to figure out his travel plans so he can go
rgs which hackathon ? yapc::na ?
stevan rgs: yup
lemme find the mail 19:40
rgs: here is one www.nntp.perl.org/group/perl.perl6.compiler/514
rgs ok
stevan and the original invite
www.nntp.perl.org/group/perl.perl6.compiler/487
oh no - I just saw where autrijus used "scintillating" 19:41
theorbtwo I certianly have to go now. 19:42
stevan theorbtwo: we are expecting you for sure
theorbtwo Perhaps I'll be able to talk some sense into Larry re =:= and trying to paper over the difference between references and values.
rgs I can't go to yapc::na 19:43
stevan rgs: why not?
rgs too far. too expensive. not the time.
stevan rgs: where are you?
rgs paris, france
stevan ahh, yes,... 19:44
although if you can make it to germany maybe you can sneak into theorbtwo's carry on bag
Qiang-work hm, compare to autrijus, it isnt that far tho :)
Limbic_Region www.perlmonks.org/index.pl?node_id=445851 # Announcement of the YAPC::NA 19:49
theorbtwo wonders if he's actually RSVP'd to this thing. 19:54
Limbic_Region the hackathon or YAPC::NA? 19:57
Jean can't leave the country yet, so I won't be going to either :-(
theorbtwo The hackathon. 19:58
I'm not RSVP'd to YAPC::NA, but AFAIK, they haven't started taking money yet. 19:59
Limbic_Region is a tad bit jealous 20:00
of the hackathon - not so much the conference
though, it will be nice to have a house with spare room(s) so when you and Jess are visiting PA you can stay with Jean and I for a day or so
theorbtwo I think we're going to be with my parents and possibly sister -- so no, but thanks for asking... 20:01
And now that I realize you live around the DC area, I have to figure out more people to visit.
Limbic_Region maybe you misparsed 20:02
I wasn't saying this specific time, but in general
and I only indicated a day or so because I know how hard it is to compress visiting friends when family wants to monopolize your time (I haven't lived in Maine for a decade) 20:03
chip for mandating scheduled releases of Parrot 20:04
theorbtwo grins.
Limbic_Region had whined about that for some time to Dan and Leo WRT community involvement/excitement
theorbtwo At some point, I'll figure this out.
Limbic_Region hmmm - chip plus plus (/me really hates this client) 20:05
theorbtwo Possibly after my father dies and my mother moves, so they all live closer together.
Limbic_Region theorbtwo - the secret is to take longer vacations 20:06
I actually sent Jean up to Maine to visit my relatives for a couple of weeks without me just so she could get some quality time visiting 20:07
fly before Jess and return after her
theorbtwo It's an idea. 20:08
But I'm not sure I want to be sepperated from her for that long.
Limbic_Region ahh - how sweet 20:09
absence makes the heart grow fonder - Jean ended up coming home early (home sick) - but don't tell her I said that - being homesick at 25 was a bit embarrasing for her 20:10
castaway :)
theorbtwo OTOH, her family is having some sort of shindig for her mother's birthday and I'm not invited for the week before. 20:11
kungfuftr yay! freebsd installed and wireless card working smoothly even without native drivers
ninereasons wonders how long he'll put off upgrading from freebsd 4.11, to 5.x 20:41
theorbtwo G'night, #p6ers. 20:45
stevan gnite orb 20:53
chip debianizing Devel::ebug looks pretty easy, except for the FIVE NEW DEPENDENCIES 21:11
ww
ninereasons is String-Koremutake a toy, or is it useful (an ebug dependency) 21:15
kungfuftr ninereasons: useful enough 21:16
ninereasons for mangling numbers?
i imagine a group of number nerds sitting around saying "nugrohigo", "drano" and cracking up 21:18
Corion Pugs porn updated at datenzoo.de/pugs/win2k.html (for Win2k, r1631) 21:34
Pugs will need to have a strict! rule against Makefile.PL being interactive and entering an endless loop requiring user feedback (when there is no user) :((( 21:42
integral ooh, down to 4 minutes for a make, nice :-) 21:43
ninereasons .oO *sigh* it would be nice to get things right the first (or ... fifth) time. 21:46
Corion sleep is open for anyone, not just me!
(backlogging :) )
ninereasons yes, very nice integral.
whatever was done is a big improvement.
Corion so I won't mind if sharpr or anyone else implements it. I wrote the test for it, that's enough. And I didn't write it, I merely stole it from p5
integral ninereasons: -H0 so ghc doesn't suck a lot more RAM than you actually have - swapping is slow
ninereasons it was becoming a huge thing to make pugs .. not so, anymore 21:47
Corion pugscode.org <<Overview Journal Logs>> | You have safely opened the door to many Perl 6 hackers. | pugs.kwiki.org | smoke: xrl.us/fmw2 | Mac OS X r1622 (451/4159) | Win2k r1631 (457/4163) 21:49
Corion Wheee. cpansmoking is driving my machine to swapping ... (1GB RAM, single user :) ) 21:49
autrijus - fix the trailing comma issue! I find code ugly that writes 21:50
, item
instead of
item,
ninereasons integral, does this faster compile time mean that pugs will be slower than before? 21:55
integral ninereasons: maybe. I'm not sure. AFAIK you can do 'make optimised' to get a faster pugs :-) 21:56
Corion good night 21:57
Corion fades
ninereasons 'make optimized' now is 8 min, compared to 68m before -H0 was changed, on this machine. 21:58
Limbic_Region ninereasons - that statement may have more meaning if "this machine" were described in better detail (cpu(s), memory, OS) 21:59
ninereasons freebsd 4.11, 1cpu, 800MHz, 1gb RAM, 22:01
m.. no, that's 666.45-MHz ... 262144K bytes real mem. 22:11
anyway, real 2m43.930s for "make" compared to real 8m43.868s for 'make optimized'.
I'm not reluctant to compile anymore. not such a big deal. 22:12
Limbic_Region chip you about? 22:44
seen chip?
jabbot Limbic_Region: chip was seen on Fri Apr 8 05:11:04 2005
Limbic_Region www.perlmonks.org/index.pl?node_id=445880 # Parrot vs. C cagematch (Just for Fun)
crysflame seen bot :) 22:57
jabbot crysflame: I havn't seen bot , crysflame

Powered by Cro and the Raku® Programming Language.

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

Thank you!