Pugs t-shirts www.cafepress.com/pugscode | Pugs 6.2.9 released | pugscode.org | pugs.kwiki.org | paste: sial.org/pbot/perl6 | www.geeksunite.net
Set by stevan on 15 August 2005.
buu You are insulting the noble art! 00:00
lscd c'mon, every noble art has its base forms ;) 00:01
putter sociopaths, or bots? 00:03
lscd both? 00:04
QMario Quiet...... 00:26
Iscd, are you still here? 00:29
What about "you" buu?
lscd yes
QMario I have a question. 00:30
Can I assign a key more than more value in a hash?
lscd erm, you should be able to assign a compound data type like a list to it, which could store multiple values
if you try to write two different values to a key in an a hashtable, you'll only have whichever one you wrote later
QMario lscd, so then an array is my best choice? 00:31
lscd it sounds like a decent choice to me, but i don't know what kind of data you're working with, etc
QMario sial.org/pbot/12929 00:32
Use that.
I am trying to assign 1 number to several different letters.
lscd? 00:33
lscd just got back
i was grabbing some water
QMario Drink up. ;) 00:34
lscd ahh, yeah, sounds like an array is a good idea
though you'll of course need some way to get the right chunk of it at some point 00:35
QMario True. This source code is going to "eat" me. 00:36
lscd well, what are you wanting to do?
print out what a phone number spells or something? 00:37
QMario Yes. I think it is annoying when companies or other type of organizations give letters in their phone numbers. Don't you?
Especially if it is urgent buisness. 00:38
lscd it's useful for actually remembering them, i suppose; aside from that, i haven't seen those in years; people don't really do that here
QMario s/buisness/Business./
lscd, are you in Canada or Europe?
lscd but, erm, isn't getting letters in phone numbers the opposite problem?
europe
you?
putter sigh. for an interpreter tree-walker regex engine, noting all the failed captures (on failed alternation branches) actually requires work... blech. ;) 00:42
lscd putter: hmm? why? [i'm not seeing why it would be more than adding them to a list? and why do you want to note them?] 00:43
QMario Do people still use 5.25" floppy drives? 00:58
lscd presumably some, but i've not seen any in use for many years 00:59
old systems/third world/embargoed countries? maybe 01:00
QMario So how would I tell my "for" loop to take each element in the array and make its output 2. Same source code previously mentioned.
lscd btw, you didn't answer the canada/europe question
QMario lscd, confidential. 01:01
People want to track me down in #perl.
lscd QMario: well, you'd have 3 possible answers for each number
QMario for ($phone=~/$_2[0]/){$phone=~s/$_2[0]/2/;}
I want that loop to access each element in the array and make the output two. 01:02
Hello jroc!!! ;)
jroc hi
lscd erm, for a 7-digit number, you have 3^7 possibilities
QMario for ($phone=~/$_2[0..2]/){$phone=~s/$_2[0..2]/2/;}
Would that work?
What does the term "too dense" mean? 01:03
And "meh"?
jroc usually applied to me
lscd hehe, i'm sure i didn't mean it in the context of intelligence 01:08
jroc Juerd: are you still offering accounts on feather?
lscd hm.. i'm not seeing where it was used; mind pasting the sentence, QMario? 01:09
putter dinner & 01:18
QMario I have "perfected" my telephone program. 02:16
Yay.
Thank God1
s/1/!
Is lscd still here? 02:17
Hello ajs_home!! :)
lscd yep 02:18
QMario Do you know buu well? 02:27
elmex will the CGI.pm for perl6 be less awkward than the one for perl5 ?
QMario Dbrock.
dbrock what? 02:28
lscd i don't
QMario You are here? 02:29
I just received an auto-reply.
When does Perl6 come out?
elmex right, when will it come out?
QMario: in 5-7 years 02:30
QMario: when parrot is finished, and perl6 specced out, and perl6 parser is implemented
they already took quite some years
QMario What is parrot? 02:31
elmex the VM which will run perl6
perl6 will be compiled to actual bytecode, which will be run by a generic VM.
QMario: that parrot vm isn't finished yet, but it already runs python pretty well. - not that many python programmers are concerned :) 02:32
QMario Virtual Machine? Is this the first for Perl? Does Perl want to "copy" Java now?
lscd there are a lot of good reasons to use vms; java's far from their inventor
elmex QMario: well, yes, a vritual machine. and no, the java vm is for static languages
QMario: the new VM will be very dynamic 02:33
lscd wonders about it and smalltalk
elmex QMario: they also say: 'parrot will be fast, because we make it fast' or: 'parrot will be fast because it will be fast'
QMario Are there any prototypes out yet?
elmex QMario: yes, the parrot vm is already quite something 02:34
QMario: just recently there was a completly new garbage collector designed for it
QMario: and it already runs many toy-languages
QMario What are toy languages?
elmex QMario: and also not-so-toy languages
QMario: like brainfuck
lscd hq9++! 02:35
elmex QMario: the brainfucklanguage is just a toy
nothing serious, but parrot can run it :)
QMario Please refrain with the language. :(
lscd elmex: hm.. i wouldn't quite call brainfuck a toy, though it's not for real apps
elmex QMario: the programing language really is named 'brainfuck'
lscd QMario: that's the name of the language
'brainfuck' is an english term, which is not an expletiative
elmex i'm sorry if it insults you :)
lscd no more than 'asset' is
or 'associate' 02:36
QMario Why would someone name a programming language with that name?
elmex QMario: and the parrot VM already runs the python languages.
lscd QMario: because it f*'s with your brain ;)
elmex QMario: because the language is ugly like hell
QMario They should translate the name into 133t.
Elmex, Hell is much worse than anything you are experiencing right now. 02:37
elmex QMario: i don't think so :)
QMario Or at least use "***" signs. 02:38
What would the name be in 133t anyway?
elmex QMario: well, afterall, if you know perl5 and want to learn perl6: FORGET EVERYTHING YOU KNEW ABOUT Perl :) Perl6 isn't much like perl5 anymore
QMario Why not?
elmex QMario: because they throw away perl5 completly, and designed a completly new language from scratch
QMario I am still learning Perl5.
elmex me too, for 5 years now 02:39
QMario Well, is it going to be OOP?
elmex perl5 is pretty OOP
just get around learning about packages and bless
it's really the same
QMario What is "bless"?
elmex but perl6 will have class { }
QMario: perldoc perltoot
QMario: or: perldoc perlboot 02:40
QMario: that are tutorials about perl5' OOP
QMario Why don't they teach Perl in Computer Science? 02:41
elmex QMario: but perl6 will have a class {} deklaration and everything, and more .... and it will be easier in perl6 to write code noone will understand
QMario: i have no idea why they don't
QMario What about Python?
elmex QMario: i also have to learn Java in CS ... but it really isn't much of a pretty language, it actually sucks IMHO.
lscd QMario: probably some people do; but from a cs point of view, there's not much point to perl
elmex i don't know Python myself enough, i just know i don't like it
QMario Why not?
lscd it's not got as much industry support as, say, java; nor is it particularly nice from a language-designer point of view 02:42
elmex i learned haskell in the last year of my studies for CS
Java is just hyped
lscd elmex: eh, python's fairly decent
elmex: sure, i hate java; but a lot of cs courses teach it
elmex lscd: i believe so, yes
lscd: yes ;(
lscd: here too
lscd including mine, alas
elmex lscd: but thats not a reason for java 02:43
lscd: also many people use windows :)
and it's not good
i really hope Perl6 will be better than Java
and 'better' than Perl5 - note: i haven't defined 'being a better language' ye 02:44
t
i really don't know anyone who is into programming as a hobby or a free software programme, who likes to program in Java... 02:45
i think it's a buissiness language
hyped like windows
broken by design like windows 02:46
lscd elmex: yes, exactly
elmex and kept alive by hype and monopoly
Khisanth the next cobol ...
elmex hehe :)
lscd lots and lots of advertising dollars to keep it in the eyes of 'decision makers' too
elmex there are many lines of cobol in the world still :)
lscd Khisanth: yep, i've been calling it that for ages
elmex can you reach www.c2.com/ ? 02:47
Khisanth there are still many lines of everything in the world still
elmex i have difficuties to connect to port 80 there and get a response
svnbot6 r6696 | fglock++ | * TODO update 02:51
Khisanth same here ...
elmex interesting 02:52
*sob*
svnbot6 r6697 | Darren_Duncan++ | /ext/Algorithm-TokenBucket : converted all 3 tabs in TokenBucket.pm to spaces, matching the indent style of the rest of the file 03:52
r6698 | Darren_Duncan++ | /ext/Benchmark : in Benchmark.pm, replaced 10 tabs with 8 spaces each (Algorithm-TokenBucket was also 8), matching the space-indent style in the rest of the file; likewise replaced 31 tabs in basic.t 04:04
r6699 | Darren_Duncan++ | /ext/File-Spec : in Win32.pm, replaced 1 tab with 4 spaces, matching the space-indent style in the rest of the file
r6700 | Darren_Duncan++ | /ext/Kwid-Event-Parser : in Parser.pm, replaced 14 tabs with 4 spaces each, matching the space-indent style the rest of the file 04:10
r6701 | Darren_Duncan++ | /ext/libwww-perl : in Status.pm, replaced 50 tabs with 8 spaces each, matching the space-indent style of the rest of the file; likewise replaced 19 tabs in Headers.pm with 8 spaces each, then reindenting 4 lines to be 2 spaces per level like the rest of the file; in Cookies.pm, replaced 4 tabs with a variable number of spaces each, to line up 2 columns of values 04:22
r6702 | Darren_Duncan++ | /ext/MIME-Base64 : in Base64.pm, replaced 32 tabs with 8 spaces each, matching the space-indent style in most of the file, and further cleaned up the indents in almost 20 other lines 04:34
r6703 | Darren_Duncan++ | /ext/Net-IRC : in IRC.pm, replaced 122 tabs with 8 spaces each, matching the space-indent style in most of the file; in IRC-OO.pm, likewise replaced 59 tabs; in SeenExt.pm, likewise replaced 2 tabs 04:46
r6704 | Darren_Duncan++ | /ext/Perl-Compiler : in Node_pm_gen.pl, replaced 6 tabs with 4 spaces each, matching the space-indent style of most of the file
coral oh. 04:58
dduncan oh, yes 05:25
svnbot6 r6705 | Darren_Duncan++ | /ext/Perldoc : in Node.pm, replaced 3 tabs with 8 spaces each, matching the space-indent style of most of the file; in Text.pm, 1 such replacement; in WS.pm, 6 such replacements; in 01-dom.t, 6 such replacements 05:28
r6706 | Darren_Duncan++ | /ext/Pod-Event-Parser : in Parser.pm, replaced 13 tabs with 4 spaces each, matching the space-indent style of most of the file (the tabs were in nearly identical places to the Parser.pm file in Kwid-Event-Parser) 05:34
Aankhen`` darr: Where were the Tabs in Status.pm? 05:37
Er.
dduncan even.
I haven't used hard Tabs anywhere... 05:38
'Least, not knowingly. :-P
dduncan oh, they *were* there ... run a diff and you'll see them 05:39
svnbot6 r6707 | Darren_Duncan++ | /ext/Set : in Hash.pm, replaced 9 tabs with 8 spaces each, matching the space-indent style of most of the file; replaced 15 tabs in Set.pm likewise
dduncan the thing is, with every file I've fixed today, each file was mostly spaces, and a few lines with tabs
Aankhen`` Hmm.
Aankhen`` tries to recall.
Oh, right.
I copy-pasted that hash. 05:40
dduncan in most cases, its like someone was working on them in an editor that had tab stops at 8 characters and ...
to be lazy, whenever they were in 4 indent levels (they used 2 chars per indent) they typed a tab rather than 8 spaces
and their editor wasn't configured to enter 8 spaces when a tab is pressed
Aankhen`` Well, tip o' the hat to you. ;-)
dduncan deductive reasoning 05:41
that's the only way that things lined up properly; and on my editor where indents are 4 chars, those tabs looked messed up
Aankhen`` My Tabs are set to 4 (spaces) too.
dduncan you use 2 chars per indent level usually, right? 05:42
Aankhen`` Not me, iblech.
I use a single Tab, i.e. 4 spaces.
dduncan well, a lot of the files I'm working with are configured that way
Aankhen`` Headers.pm was iblech's work.
dduncan s/working with/fixing/
Aankhen`` crosses his fingers and tries to build Pugs. 05:43
dduncan right now I'm only fixing the /ext, not the core
Aankhen`` Bleh, so much for that.
dduncan Pugs 'make' worked fine last I tried 05:44
Aankhen`` It hasn't been working for me for about a week and a half.
Hmm, a little less than that.
dduncan while I prefer 4 chars per indent level, I'm not going that far as to reformat all the files I see ... I'm just doing the minimum of replacing tabs with spaces, to make it look to everyone else what it looks like to the main author 05:45
Aankhen`` nods.
OK.
I need to exercise, shower, eat brunch... not necessarily in that order.
BBL. &
dduncan also, in every case I've hit so far, its obvious that the author didn't intend to use tabs-only, since they aren't ubiquitous, but those got in by accident or laziness 05:46
back in the day, I used tabs on purpose for all my indenting, and consistently so, but this week I converted all those to spaces, consistently
I haven't hit anyone else using all-tabs for indenting yet 05:47
svnbot6 r6708 | Darren_Duncan++ | /ext/Test : in Test.pm, replaced 17 tabs with either 4 or 8 spaces each, matching the space-indent style of most of the file 05:57
r6709 | Darren_Duncan++ | /ext/Test-Builder : in 022_Test_Builder_Test_Fail.t, replaced 8 tabs with 4 spaces each, matching the space-indent style of most of the file; replaced 2 tabs likewise in 030_Test_Builder_Output.t 06:03
r6709 | Darren_Duncan++ | ; replaced 6 tabs likewise in 040_Test_Builder_TestPlan.t
r6710 | Darren_Duncan++ | /ext/Tree : in Tree.pm, replaced 119 tabs with 4 spaces each, matching the space-indent style of most of the file; likewise replaced 1 tab in FindByPath.pm
r6711 | Darren_Duncan++ | /ext/Tree : in extended_tests.t, replaced 345 tabs with 4 spaces each 06:09
dduncan fyi, and with that, there are zero files left in /ext containing tab characters 06:10
as for the rest of the Pugs distro, we shall see ...
luqui dduncan, what made you decide to go on this textual ethnic cleansing?
dduncan inspiration from "Perl Best Practices" for one thing ... also, everyone who used tabs besides me in /ext used a mixture of tabs and spaces in the same file, and that really messes up how code looks to anyone but the author 06:11
this is all in an effort to increase maintainability of code 06:12
luqui dduncan, are you also adding the appropriate vim and emacs modelines?
so that in the future when people press "tab", they are consistent with the file?
dduncan I don't use those programs ... don't know how
I know how to do that in BBEdit
coral # :ts=4 mumble something or other for vim 06:13
i think
dduncan I know those programs can do it though ... I think Best Practices even provides the config specs
for vim and emacs
luqui well, that is another cleansing for another day
also, you might find you have trouble normalizing with the haskell code 06:14
because it's much more "alignment"-ly indented
dduncan meanwhile, lib/pugs/hack.pod was updated a few days ago to specify that we use spaces exclusively for indents, so no new tabs should appear 06:15
luqui yeah, but nobody really reads that
dduncan I'm leaving the haskell code until last
as for that file, its where autrijus said that instruction should go
luqui sure
nobody really reads anything before hacking
or at least.. I don't
dduncan afaik, autrijus uses spaces already, as do most others, especially those working on the haskell 06:16
well, the bottom line is to be self-consistent
luqui but I think hack.pod is the best place for it
dduncan at least on a per-file basis, all indents should use the same characters and have the same n-chars per level
but having all the files the same also helps people that copy and paste between them 06:17
or otherwise work with multiple files, so they don't have to change habits within a project
luqui dduncan, except for haskell 06:18
I mean, sortof
there is a "normal indent"
that's used after, say, a "do"
dduncan is haskell whitespace-sensitive?
luqui but there's also a lot of indenting up to where the expression started
yes
dduncan well, I think I'll leave the haskell alone entirely, since I don't hack haskell 06:19
luqui that might be a good idea...
luqui has been surprised a few times by the indentation rules
dduncan I'll focus on the perl 6 files and pod files, and perl 5 files, which I understand
and that aren't whitespace sensitive
luqui thinks he will start another normalization project 06:20
normalizing all the internally-referenced operators into the global namespace
dduncan in Perl 5 land, I'm making several changes to my modules to align with Best Practices ... whitespace is the first one
the next CPAN releases will use space indents instead of tabs ... that change is already in my repository 06:21
svnbot6 r6712 | Darren_Duncan++ | /t/ : replaced indenting tabs with spaces in these files: kv.t, pairs.t, goto.t, try.t, io.t, readline_chomped.t, time.t, sort.t, zip.t, rounders.t, substr.t, trans.t, 02-strings-with-spaces.t.disabled, create_files.pl, system_tests, showav.c (they are all under 'builtins') 06:45
r6713 | Darren_Duncan++ | /t/* : replaced indenting tabs with spaces in t/builtins/undef.t (and that's the end of the 'builtins' subdirs)
dduncan fyi, I will be skipping over the 3790 tabs in /t/rules/rules.t because that is a generated file 07:04
svnbot6 r6714 | Darren_Duncan++ | /t/* : replaced indenting tabs with spaces in these files under [data_types, junction, oo, operators]: enum.t, hash.t, pair.t, junction_any_pick.t, misc_junctions.t, instance.t, mutators.t, indirect_notation.t, chaining.t, overload.t, anonymous.t, conflicts.t, super.t, adverbial_modifiers.t, operator_overloading.t, pipe.t, precedence.t, quoting.t, smartmatch.t, splat.t
dduncan otoh, the script that generated it could be fixed later
svnbot6 r6715 | Darren_Duncan++ | /t/* : replaced indenting tabs with spaces in the README file and these files under [packages, pugsbugs, pugsrun, rules, statements]: autoload.t, Test.pm, parse-fail.t, parsing_if.t, rule_used_in_Str.t, 07-dash-uppercase-v.t, der_grammar.t, grammar.t, control_blocks.t, gather.t, given.t, if.t, loop.t, next.t, redo.t, unless.t, while.t 07:22
GeJ wonders if dduncan does a perl oneliner or a pugs' one to replace the tabs with spaces 07:25
dduncan no, I just do a "search and replace + save" in BBEdit
also, BBEdit has a nifty multi-file search function
so I can just point it at a folder and say, "show me all files having tabs"; it gives a list, complete with count 07:26
then I double-click each list item, hit replace all (or other custom step depending on circumstances) and save
GeJ the GUI is the path to the dark side 07:27
dduncan that's where I got the count of tabs replaced in the /ext stuff earlier ... I didn't count them by hand
I find it a lot easier to use than a command line text editor
under the circumstances, it seemed too much trouble to write a script, also
I wanted to see the situation manually for each file to ensure the replacement is done right per circumstances 07:28
and BBEdit has so much script-like goodness built-in, I'm not doing *that* much work
GeJ Was just kidding anyway. I guess everybody has his own tools of choice. As long as the job is done, that's not a problem :) 07:31
dduncan you got it, buddy 07:32
and with r6717, all tabs in /t/* are gone also, save the generated rules.t 07:46
svnbot6 r6716 | Darren_Duncan++ | /t/* : replaced indenting tabs with spaces in these files under [subroutines, syntax, unspecced, var]: bare_block_with_dollar_underscore.t, currying.t, defaults.t, lvalue_subroutines.t, multi_sub.t, nested_named.t, sub_return_values.t, subroutine.t, unhashify.t, wrap.t, parsing_parens.t, subscript_parse_bugs.t, symbol_table.t, chop.t, config.t, coro.t, method.t, roundtrip.t, default_scalar.t, is_constant.t, state.t, temp.t
r6717 | Darren_Duncan++ | /t/* : replaced tabs with spaces in t/unspecced/coro.t
r6718 | Darren_Duncan++ | /t_disabled/* : replaced tabs with spaces in these files under rules/from_perl6_rules: pos_modifier.t, hashkv.t, exhaustive.t, overlapping.t, propcharset.t, properties.t, counted.t 07:52
dduncan there are approximately 19,724 tab chars left in pugs distro files, of which about 3,750 are in 2 skipped /t/* files 07:54
I will skip the haskell files and anything just copied from outside 07:55
mainly just the actively worked on non-haskell files will be looked at
luqui evalbot is now on an auto-update loop 07:56
somebody should commit so I can see if it is working
luqui has a commit coming, so it's no biggie
dduncan I'll also add to the Changes file what I've been doing, so thats a commit 07:57
otherwise, I'll postpone any other detabbing changes for a few days, to let the large number I did today shakedown
ensure the smoke etc continues to work 07:58
mugwump why the sudden whitespace changes? 08:02
dduncan to improve maintainability of pugs 08:07
a good 100 files had a mixture of tabs and spaces in them for indenting, which don't look right in most editors
having all spaces, as recommending in Perl Best Practices, makes the code look right to everyone
svnbot6 r6719 | Darren_Duncan++ | Added new ChangeLog section for revision 6.2.10, added comment about my recent tab expanding
dduncan I also note that, while its a month since 6.2.9 went out, either no one has been updating the ChangeLog about it in that time, or they've accidentally added to 6.2.9, as I almost did ... will check commit log to know for sure what has been happening ... 08:09
mugwump personally I haven't even touched anything pugs-ish in the last month. Hopefully I'll get back into it soonish 08:10
dduncan fyi, in perl 5 land, I converted Rosetta all to space indenting yesterday ... in the repository, but plop nonfunctional then 08:11
luqui ?eval 42
evalbot_6718 42
evalbot_6719 42
luqui well... um... it almost worked
dduncan dood
oh, and ossf looks different today 08:12
luqui should rtfm for "kill" :-)
dduncan according to ossf, there was indeed no changes to ChangeLog between Aug 3 and today, so no problem re what I mentioned 08:13
anyway, I'm off to bed (now 1:14am here) ; good night all 08:14
mugwump nn dduncan 08:15
svnbot6 r6720 | luqui++ | In heredoc support, instead of giving the parser (string "hopefully this won't 08:19
r6720 | luqui++ | match qz87jh##") for the beggining string, I just gave it the parser (fail),
r6720 | luqui++ | which will never match!
hcchien well, pugs make failed.src/Pugs/PIL1.hs-drift failed at util/drift.pl line 8
luqui huh, weird
hcchien, what platform?
hcchien luqui: OS 10.4 08:20
luqui I'm not sure what drift is 08:21
but it looks like a hugs thing
dduncan did pugs used to work for you under 10.4?
I heard there may still be outstanding issues with that system on various software fronts ... I'm sticking to 10.3.9 for now just in case 08:22
hcchien dduncan: yes. I got the 6.2.9 compiled before.
luqui ?eval 'test eval'
evalbot_6719 'test eval'
dduncan evalbot seems fast now, unlike a few months back ... maybe its the prelude at work? 08:23
there used to be a few seconds delay in replies
hcchien 10.4 uses gcc 4.0 for default. It makes some troubles.
luqui dduncan, well, I'm compiling it optimized
and the prelude is precompiled
so it's probably a combination of the two 08:24
dduncan supposedly ghc 6.4.1 works with Tiger, but 6.4.0 had issues
luqui wooo! dynamic updating!
dduncan that is so ... totally ... kewl! 08:26
although not much more technically complex than the smoker, I imagine
luqui uh, it was a lot less complex than the smoker
it's a 20 line script :-) 08:27
dduncan I just meant that the smoker's been auto-updating for awhile
luqui oh, yeah
and evalbot used to
but then iblech stopped running evalbot and I started
and I didn't ask him for his script
plus I added the revision number name magic 08:28
pretty high achievement, don't you think %-)
dduncan the rev number is a plus ... lets us know who's talking
hcchien does anybody run the pugs on Mac OS 10.4? 08:29
dduncan I will by christmas, probably, but not meanwhile
luqui oops, forgot to waitpid in evalbot 08:32
Juerd 03:11 < jroc> Juerd: are you still offering accounts on feather? 08:51
Yes, and if you hadn't left, you'd have read this. 08:52
godsmack1r What are accounts of feather, btw?
Huh, why isn't there a MANIFEST?
08:52 godsmack1r is now known as rafl
Juerd feather is a machine 08:52
A feather account is an account on that machine
integral no MANIFEST in svn, only in distributions, I thought 08:53
rafl Sounds plausible. How is it generated? Is there a script? 08:54
Juerd At random
integral "make manifest" iirc 08:55
dduncan that's right ... no MANIFEST is checked in to svn, only generated for distros 09:00
its much simpler, since people have kept forgetting to update it, back in the older days of pugs when the MANI was in svn
anyway, as my last comment before I truly go to bed, r6720 goes all the way through 'make' without problems ... I'll try 'make test' tomorrow 09:01
svnbot6 r6721 | Stevan++ | Perl6::MetaModel 2.0 - 09:07
r6721 | Stevan++ | * added more tests to t/09_subclassing_Class.t
r6721 | Stevan++ | - you can easily create $::Class subclasses through the
r6721 | Stevan++ | Perl6::MetaModel macros
r6721 | Stevan++ | - added new field for Perl6::MetaModel macro "metaclass"
r6721 | Stevan++ | to allow for easy composition using classes other than
r6721 | Stevan++ | $::Class
r6722 | rafl++ | Debian: 09:31
r6722 | rafl++ | * Build optimized by default.
r6722 | rafl++ | * smoke-pugs instead of smoke as default test target.
kgftr|konobi moo 11:14
joao moo is cool :-) 11:18
(i mean, moo.di.uminho.pt:7777 ) :-)
kgftr|konobi castaway is in the building 11:24
=0)
vkon given PIL mention in .\docs\notes\plan, does it already fully-working? Also, file \src\Pugs\PIL1.hs-drift looks good; will be there PIL2.hs? 14:06
?eval $?PUGS_VERSION 14:08
evalbot_6723 \'Perl6 User\'s Golfing System, version 6.2.9, August 3, 2005 (r6723)'
svnbot6 r6723 | dankogai++ | My first contribution to Pugs. Enjoy! -- dankogai
nothingmuch wtf wrt tabs vs spaces? why not just keep things as they are? I mean, what does it really matter? 14:21
luqui knows
it's dduncan... 14:22
nothingmuch that much I know
i also know that dduncan did this to his own code
luqui no, I mean, it's dduncan, that's why
nothingmuch beh
this is crummy since it ruins the ability to properly diff before and after tab conversion,
luqui thinks he made a similar point
nothingmuch oh welll... too late now. =(
luqui that you kill anybody's pending patches 14:23
vkon "diff -b " will help
nothingmuch yeah, that too
vkon: it's besides the point, the conversion is pointless
there is no need for consistency
except in the haskell code where it's already enforced
luqui well, damian is really to blame, right?
nothingmuch oh shit, that new book said so?
luqui yeah
vkon nothingmuch: sure!
nothingmuch well, whatever 14:24
nothingmuch finished reading mail and goes back to reading
well, at least I made some perlmonks XP off damian's book ;-) 14:25
perlmonks.org/?node_id=488288
oh, i'd like to make another complaint 14:27
higher order functions are yucky in forth
integral
.oO( Joy )
14:29
svnbot6 r6724 | Stevan++ | Perl6::MetaModel 2.0 -> 14:32
r6724 | Stevan++ | * added t/15_submethods.t which is converted from
r6724 | Stevan++ | the Pugs t/oo/ test suite
r6724 | Stevan++ | - fixed bug in submethod handling revealed by
r6724 | Stevan++ | this test (TDD++)
r6724 | Stevan++ | * some work on Synopsis 12.5 (a.k.a The Perl 6 Meta Object Protocol)
renormalist Hi all! I started an Emacs cperl6-mode by deriving it from cperl-mode and evolutionary adding new perl6 syntax to it to make indenting and highlighing work again. It already works for me and would like to contribute that to Pugs. Yesterday, putter tried to find someone who could give me commit permissions but unsuccessful. Now I try it again today. So is someone interested and if yes, can give me permissions to commit? 14:41
putter renormalist: putter awoke to find himself an admin. there should be an invite in your inbox. 14:44
luqui putter, congratulations. I know it is a high honor to be an admin on a project like this that has such tight code security 14:45
:-)
putter lol ;) 14:47
renormalist: and please a yourself to AUTHORS. tnx.
s/a yourself/add yourself/ sigh 14:49
putter pictures standing on a street corner with a placard - "Contribute to PUGS \ Free commit bits!" 14:50
hmm. maybe at the next local pm meeting
luqui hehe
renormalist ok, I'm registered now. 14:55
I now read docs to see how do I edit in pugs svn. 14:56
putter prolog rx engine is now passing 85% of perl5's re_tests file-o-tests. (which is very incomplete, and I've been coding-to-tests rather than coding-and-testing, so actual coverage is way less than 85%) long caveat aside... so it's almost time for care-free refactoring and the start of p6 integration. (its a combined parser) 14:58
luqui renormalist, are you familiar with svn? 15:00
putter, what is the scope of the rx engine?
that is, in what subportion of pugs will it be used? 15:01
putter It _should_ (unimplemented) be posible to be reentrant, and to fully support embedded code (source filters++). but near term objective is for modules/Grammars/Perl6.pm to start working, so someone can push on a p6-rules-based p6 parser.
vkon putter: you mean tests in t/rules.t? I understood a couple of bits there, but mostly those tests are based on parrot, which is currently not fully complete.
renormalist luqui: yes, regular svn user. I already have an anonymous pugs checkout, but don't know whether I re-checkout with username/pw or whether I can svn switch or the like or whether it works different at all. 15:02
luqui renormalist, just commit
it will ask you for username and password then
renormalist ah, ok.
have to try it later, currently at the wrong machine. 15:03
luqui putter, so...
main pugs?
is it perl 6 code
luqui is very confused
putter vkon: the perl 5.9.something distribution has an t/op/re_tests (?) , driven by a t/op/regex.t (?). one-line-per-test data file, pattern, string, expectation. probably what our rules.t and perl5.t were ultimately derived from.
svnbot6 r6725 | fglock++ | * perl5/List - map() uses uses Code.arity() information 15:06
putter luqui: re so..., just an update to 1- keep me motivated, and 2- create interest&expectation of delivery so it doesnt become yet another unreleased neat thing sitting on my disk.
vkon putter: mostly, in t/rules/rules.t there should be replaced \d to <[:digits:]> and similar to this... But not all classes are implemented in parrot
putter: to say, there is a class [:alpha:] but it is not what should correspond to \w, because of underscore... 15:07
putter re what is, prolog, with a perl wrapper. prolog running on swi via search.cpan.org/~salva/Language-Pro...aswi-0.08/ .
integral putter: it doesn't work in AI::Prolog? 15:08
putter vkon: err, why replace \d ? \d has to work too...
vkon putter: indeed, \d should work, but it must be replaced (IMHO) within character classes, say, in [\d] 15:09
putter: actually a script producing t/rules/rules.t was buggy, and so many places were hsnd-edited afterwards 15:10
putter integral: I'm not familiar with how "real" AI::Prolog is. My fuzzy recollection is it is still rather a toy system. Regardless, aside from DCG's, most of it should be "portable" prolog (there's an oxymoron). 15:11
integral hehe. 15:12
putter vkon: re "must be replaced" - does the spec say so? P5 certainly permits [\d].
t/rules/rules.t has had a long and rich life supporting multiple projects, and an evolving p6 rules syntax. feel free to grab a recent re_tests from perl 5.9.mumble and spit out a new one. 15:14
vkon putter: well, I must admit you know this better. I walked through files few days ago, and begun to vaguely understand which files stand for. I didn't knew at which side to fix. 15:15
putter: but which worker script will spit out a new one? (I tried to ask this here :) ) 15:16
svnbot6 r6726 | fglock++ | * perl5/Value - new Class 'Rat'
putter np. putter wishes yet again for a vcs which permits "doesnt clutter up the code" code annotation. 15:17
s/vcs/development environment/
vkon putter: I knew that Perl5 allows [\d], but does Perl6 allows this??
putter I'm not sure. will have to know in another day or few. ;) 15:19
vkon putter: to summarize, t/rules/rules.t is *not* regenerated anymore, because 1. generator script is not found anywhere, and 2. many hand-edits are done. (however, once that script will be found, its better to improve it, and not t/rules/rules.t) 15:23
putter <[abc]+<digit>> vs <[abc\d]> hmm... i didnt find anything better / more recent in Bible than <[,\h]> in A05. havent checked p6l. 15:28
vkon ?eval 'a' ~~ /<[abc]+<digit>>/ 15:31
evalbot_6726 Error: cannot cast from VUndef to Pugs.AST.Internals.VCode (VCode)
putter vkon: the end-game is perhaps perl5-like automated testing. I started to say I suspect pugs isnt fast enough for that yet, but with parser backtracking and memory size issues, I'm not sure.
The script which generated rules.t is long gone (was on another project (damian's p5 rules implementation), and the regex language has changed since then). 15:32
So any scripted recreation of rules.t will be a new script.
fglock putter: I need some help making anonymous blocks work in PIL-Run - map() compiles fine, but there is an error in p5 15:33
svnbot6 r6727 | fglock++ | * PIL-Run - implemented map()
putter Until such a script is written, further editing of rules.t is fine. it's what we've got. 15:34
vkon putter: agreed... if there's no objections, I'll see that big file within next 4-5 hours and will commit my vision of changes there... with explanatory comments.... any objections? 15:35
putter marvels again at autrijus conversing in n simultaneous threads on m channels 15:36
vkon: go for it.
fglock ... 15:37
vkon ok then. 15:38
autrijus putter: me? what? :) 15:40
putter: btw, you are a metacommitter now, so feel free to committerify renormalist 15:41
fglock is pick() supposed to remove the selected element from hash/array?
autrijus putter: rt.openfoundry.org/Foundry/Project/...Queue=pugs is the admin interface after you log in
putter luqui: re what is, perl5 starts up embedded prolog, loads it with prolog code defining an engine, creates "rx" objects with hand off to prolog for evaluation, and postprocess result into Match objects. eventually should be able to integrate this with a runtime (js-w-perl5 or pilrun), but immediate goal is just to make possible a p5 script, with a heredoc of p6 rules, which parses things (to permit grammar creation).
autrijus: done. :)
autrijus cool! 15:42
putter s/creation/development/
autrijus come to think about it, I'll spread the metacommittership a bit more
putter ok, fglock, sorry for the delay. ok, can you give me a test string to pursue? 15:43
autrijus luqui and fglock: you two are now metacommitters as well. I think we should make all channel regulars metacommitters to avoid the renormalist situation :)
fglock $a={1}
autrijus: :)
autrijus bbl... :) 15:46
fglock putter: when map() works, uniq() and grep() can be implemented in PrimP6 :)
putter ! :) 15:47
hmm... I get a pilrun prelude parse error...
the p6 prelude that is 15:48
nothingmuch seen Alias 15:49
putter fglock: map should get a mm Code object? 15:50
fglock yes 15:51
bbiab & 15:54
putter oh. putter was about to ask fglock to fill in the ###... in 6728. ah well. 15:55
svnbot6 r6728 | putter++ | ApiX.pm: minor tweak. 15:57
putter I am so gone. fglock: p6 prelude wasnt working, so not even a little regression testing. so I may have broken things. I'll be back in 10 hours or so. 16:14
svnbot6 r6729 | putter++ | PIL-Run - quick and dirty first cut at Code support. The Code branch of p6_new needs to be filled in. Untested. Not even regression testing done. fglock++
putter bye all & 16:15
16:35 typester is now known as typeout
svnbot6 r6730 | fglock++ | * PIL-Run - map() works! putter++ 16:50
rafl Juerd: What distribution runs feather? 17:03
17:04 typeout is now known as typester
svnbot6 r6731 | autrijus++ | * Thoroughly fail+todo this test as on OSX it triggers the nasty 'EVACUATED object entered!' bug. Need to retest once GHC 6.4.1 is released. 17:08
rafl Are all those .hs, .lhs, .c and .o files really needed to be installed? 17:19
Juerd rafl: Debian
rafl: of course. 17:20
rafl Juerd: Great. May I do some builds on that machine? Building pugs + smoking sucks a but on my 800MHz machine with 512Mb RAM. 17:21
Juerd rafl: You may request an account by email
rafl: I need your real name
rafl Juerd: It's in my /whois. my email address is [email@hidden.address] Or do you need an email because of a gpg signature or something? 17:23
Juerd I need an email for my own administration :)
And because that is accepted in .nl as an official document
rafl Juerd: OK, where may I send it to? 17:24
Juerd This is just in case you become a mad h4x0r one day :)
Any of my email addresses, except the gmail one.
svnbot6 r6732 | autrijus++ | * add Lola Lee to AUTHORS for the report on the previous t/oo/destruction.t change 17:26
rafl Juerd: Sent. 17:27
svnbot6 r6733 | fglock++ | * implemented grep() in PrimP6, but it doesn't work yet 17:31
GeJ poor evalbot, you're making it nuts 17:35
17:37 typester is now known as sleepster
Juerd rafl: Username rafl, florian or flo? 17:39
rafl: Or different even?
rafl Juerd: rafl is fine. 17:40
Juerd rafl: see query 17:49
coral is evalbot self-upgrading? 17:50
geoffb coral, yep
Added this morning
coral neat :) 17:51
rafl Juerd: No query received. 17:53
geoffb coral, luqui++'s magic :-)
svnbot6 r6734 | fglock++ | * PIL-Run - id(), clone() 17:55
Juerd rafl: Weird. 17:57
rafl: I did privmsg you 17:58
coral luqui++
it is definitely a bit odd to watch
y'all should attach evalbot to workbench
Juerd rafl: I'll do it in public then, trusting the people here
coral so that it doesn't have to visibly exit/enter every time
Juerd rafl: Your password is florra
coral or some equivalent bot framework
Juerd rafl: On feather.perl6.nl
rafl: Change it asap 17:59
coral otoh it's neat watching it leave/rejoin newer
rafl Juerd: Done, thanks. 18:07
svnbot6 r6735 | fglock++ | * PIL-Run - implemented reverse(), but the string is not returned ???
Juerd rafl: yw
rafl: How are you going to test this, by the way? debootstrap?
obra seen autrijus 18:09
rafl Juerd: I assume all pugs build-deps are installed? If yes I won't need deboostrap. If not I would probably need to use pbuilder (which I should, because it assures that I've got the build-Dep line right). That would require you to allow me to execute pbuilder as root using sudo. 18:10
Juerd rafl: They are
rafl: pugs is automatically built every 15 minutes :) 18:11
svnbot6 r6736 | fglock++ | * PIL-Run - implemented uniq() in PrimP6, but doesn't work yet 18:12
rafl Juerd: pbuilder would be better anyway. But I would understand if you don't want to setup sudo. 18:13
obra 'evening, lambdacamels
rafl Juerd: Is there an svk/svn mirror on that box? 18:21
Juerd rafl: Oh, if you want full sudo access, just ask for it
rafl: Yes, read ~autrijus/config
rafl: feather is a community box. Nothing on it is strict. 18:22
svnbot6 r6737 | fglock++ | * PIL-Run - supports Code.name(), Code.body(), Code.arity()
r6738 | iblech++ | * Usual svn props.
r6738 | iblech++ | * All examples/ which are tested in t/examples/examples.t:
r6738 | iblech++ | Added remaining reminders to update examples.t and examples/output.
r6738 | iblech++ | * New examples/output/algorithms/lambda-calculus:
r6738 | iblech++ | example.t tests the output of dankogai++'s lambda-calculus.p6 now.
rafl Juerd: I don't need to be root. Only pbuilder needs to run as root.
svnbot6 r6738 | iblech++ | * pugs::hack: Added the settings needed to make <Tab> indent by four spaces in
r6738 | iblech++ | Vim, dduncan++, can somebody add the appropriate settings for emacs?
Juerd rafl: I'm not giving out partial root access :)
That sounds and tastes like maintenance
rafl Juerd: rafl ALL=(ALL) /usr/sbin/pbuilder in /etc/sudoers is enough, I guess. 18:25
(and installing pbuilder and cdebootstrap, of course. 18:26
s/$/)/
Juerd no, that'll be ALL then.
rafl: You got it 18:27
rafl Juerd: That's not all. Sorry, user resi is not allowed to execute '/usr/bin/apt-get moo' as root on localhost.localdomain. when having resi ALL=(ALL) /usr/sbin/pbuilder in the sudoers file. 18:28
svnbot6 r6739 | fglock++ | * PIL-Run - exec()
Juerd rafl: WHo's resi? 18:29
rafl Juerd: A user on the box I'm working at currently.
Juerd rafl: What can I do about that box? 18:30
rafl Juerd: Nothing. It just shows that the line above in /etc/sudoers doesn't allow to execute everything as root. 18:31
Juerd rafl: No, what you get will be ALL 18:33
Because nobody gets custom config - that's too much work
rafl :-( 18:34
Juerd: Wow.. that machine is fast! 18:39
svnbot6 r6740 | iblech++ | * t/builtins/control_flow/eval.t: skip in PIL2JS and PIL-Run.
r6740 | iblech++ | * PIL2JS: pugs-smokejs.pl: Hack:
r6740 | iblech++ | my $a; sub b { $a++ }; b(); # does not work; gets converted to...
r6740 | iblech++ | my $a; our sub b { $a++ }; b(); # ...which does work.
r6740 | iblech++ | See comment in pugs-smokejs.pl for a more detailled explanation.
r6741 | fglock++ | * fixed map slurpiness 18:45
geoffb OK, there's something freakish about evalbot updating before svnbot mentions the commit 19:01
But since svnbot has long been a bit sluggish, I'd say that's the place to go fix
svnbot6 r6742 | fglock++ | * PIL-Run - '=:=' and Hash index use the same 'id' 19:03
vkon when 'ab' ~~ /(a|b)*/ evaluated, $0 appears to be an array containing all its matches, and not last value as in Perl5, thus causing many P5-derived tests to fail. Is it intentional? Looks like this is not hard to fix, but am I thinking in right direction? 19:12
nevermind, I'll get my big editor and see if I can fix this... 19:17
?eval say "yep, this is so :)" 19:18
evalbot_6742 yep, this is so :) bool::true
svnbot6 r6743 | rafl++ | * Added libyaml-perl to Build-Depends. 19:27
r6743 | rafl++ | * Added pugs-modules.dirs to install the directory for lintian overrides.
iblech vkon: This might actually be the correct behaviour. Ping p6l? 19:33
svnbot6 r6744 | fglock++ | * perl5/List - map, grep no longer tested in t/list-operations.t
r6744 | fglock++ | because they are higher level functions
r6745 | iblech++ | examples/network/:
r6745 | iblech++ | * svnbot.p6: Added documentation, including the anser to
r6745 | iblech++ | "why does svnbot lag?".
r6745 | iblech++ | * README: Referred to svnbot.p6's POD.
iblech geoffb: Indeed :) (BTW, evalbot doesn't need to be restarted (as evalhelper.p5 spawns a new pugs upon each ?eval), but I like it :)) 19:36
svnbot6 r6746 | fglock++ | * perl5/List - refactored map() 19:44
Juerd rafl: No, it's just not very slow 19:46
rafl Juerd: Am I allowed to install some stuff? Some debian development tools mainly. 19:47
Juerd rafl: Sure, go ahead. Just stick to debian packages. 19:51
If you go compile things yourself, or forcefully install otherwise blocked packages, I'll have to kill you 19:52
But in other circumstances, have fun :)
rafl Juerd: Why didn't you btw? ghc isn't installed via dpkg, for example.
Juerd Is too
ghc-cvs
rafl Cool, never seen that. 19:53
Juerd The only things not installed via dpkg are pugs and parrot
And they're neatly put in my homedir
ghc-cvs was needed because the default ghc wasn't recent enough
rafl What's not recent enough with ghc6 in sid? 19:54
Juerd At the time, it was too old for pugs to be compiled
I don't know if it still is
But this works well too 19:55
rafl Juerd: I usually compile it with sids ghc6 and it works fine. 20:00
Juerd Then it is 6.4 now 20:01
rafl Yes.
svnbot6 r6747 | fglock++ | * PIL-Run - Array.delete(slice), Hash.delete(key) 20:09
rafl Hm. does make manifypods create proper manpages for Perl6::Pugs and pugs::run? For me they are just empty. 20:34
svnbot6 r6748 | fglock++ | * PIL-Run - fixed Array.exists(), does() 20:39
20:48 nnunley3 is now known as nnunley, mml is now known as mml_
svnbot6 r6749 | fglock++ | * PIL-Run - capitalize() 21:03
r6749 | fglock++ | - PrimP6 is emitting warnings about grep() ???
21:18 mml_ is now known as mml, mml is now known as mml_
svnbot6 r6750 | iblech++ | * Usual svn props. 21:33
r6750 | iblech++ | * STATUS: Update (P5 regexes in PIL2JS, laziness in PIL-Run, more tests)
r6750 | iblech++ | * pugs::run: Fixed information about -Bbackend and pointed to PIL2JS's
r6750 | iblech++ | jspugs.pl and PIL-Run's crude_repl.pl. Also updated the entries about
r6750 | iblech++ | -Cbackend and Pugs's safemode.
r6751 | Darren_Duncan++ | /ext/Locale-KeyedText : replaced all occurances of 'is constant' with 'is readonly', the new way to do the same thing
r6752 | fglock++ | * PIL-Run - the %ENV hash works
r6753 | fglock++ | * PIL-Run - p6 %ENV is passed on to exec() 21:45
r6754 | fglock++ | * PIL-Run - system() 21:51
r6755 | iblech++ | * Removed t/var/is_constant.t, as there is no "is constant" any longer. 22:09
r6755 | iblech++ | * New t/var/is_readonly.t: Added tests for "my $foo is readonly", including
r6755 | iblech++ | tests for "my $foo is readonly; exists($foo) # should be false".
r6755 | iblech++ | * New t/var/constant.t: Added tests for "constant pi = 3", etc. Also pinged p6l
r6755 | iblech++ | to see whether lexically scoped constants are ok and added appropriate
r6755 | iblech++ | =for discussion tests:
r6755 | iblech++ | www.nntp.perl.org/group/perl.perl6....uage/23000
r6756 | vkon++ | improve and clean t/rules/rules.t; still many things to do
dduncan ping vkon 22:13
vkon I'm here... 22:14
dduncan regarding t/rules/rules.t, that is the one t/* file that I didn't detab, because it was generated ... is this file regularly going to be resynced with the perl 5 version, or is it just manual updates from here on in? 22:15
did your cleaning involve rerunning the generator script, or was it a manual action?
vkon I searched for that generator script and it was decided here that it wass lost; so manual editing is used.
dduncan in that case, unless you already did so, I'll detab it now 22:16
that said, all the tabs I saw were in comments, not code
vkon go ahead... I'll continue working with that file tomorroww, and, of course, I'll re-download it before usage. 22:19
dduncan I shouldn't take more than 20 minutes
vkon have a nice day
dduncan you too 22:20
svnbot6 r6757 | Darren_Duncan++ | went ahead and detabbed t/rules/rules.t, since it probably won't be regenerated any time soon (script was lost supposedly); it had 3790 tabs, all in the comments, each was replaced with 4 spaces 22:33
r6758 | Darren_Duncan++ | converted tabs to 8 spaces each in inc/Test/Harness.pm and inc/Test/Harness/Straps.pm 22:45
r6759 | Darren_Duncan++ | converted tabs to spaces in these 3 /inc/Module/Install/* files: Makefile.pm, Pugs.pm, WriteAll.pm 22:51
dduncan that takes care of all the /inc files 22:54
svnbot6 r6760 | Darren_Duncan++ | updated lib/Perl6/MakeMaker.pm so that it generates Makefiles having lines indented with spaces (4 per indent level) rather than tabs; there were 3 tabs replaced in total 22:57
dduncan and that takes care of all the /lib files 22:58
rafl Don't hurt me for that question, but what is the benefit from using spaces? 23:00
dduncan better maintainability of code ... the code looks the same to everyone, no matter how their editor or viewer is configured re how wide a tab displays 23:02
the worst situation is having both tabs and spaces for indenting in the same file 23:03
having pure tabs isn't as bad, but I'm flipping those too
rafl Hm, I always liked the the fact that the file looks like everyone wants if you use tabs and configure your editor properly.
dduncan however, you run into a lot of problems if some code lines are indented with tabs, and others spaces 23:04
rafl For sure, yes.
dduncan there's also the advantage of knowing consistently that your code can display in the same width of display no matter what people's tab settings are 23:05
eg, if everyone tries to keep their code within a 78 or 80 char width to avoid wrapping in terminals, someone using narrow tabs won't have it look good to them, but have the code wrap for someone else using wider tabs 23:06
rafl Agreed. 23:07
dduncan a more thorough explanation of the issue is in Damian's "Perl Best Practices" 23:08
I will now be away from comp for about 20 mins
mugwump sounds like you've been reading the GNU coding standards documents, dduncan :) 23:22
dduncan actually I haven't, and in fact Perl Best Practices cites a few divergences from there, such as with where opening and closing braces should go 23:26
I'm sure they have a lot in common, though
actually, I think I looked over the GNU standards page a few months back, but I'm not sure 23:36
but it doesn't really matter, since I'm citing PBP anyway
fyi, I'm now proceeding through the projects in /perl5; each project is its own commit, so it is easier to isolate 23:38
unlike what I've gone through in other directories, the /perl5 stuff seems to be consistently all tabs
I assume Java is not whitespace sensitive, from what I remember in using it 23:39
and yes, it does seem kind of odd to have 2 java projects under /perl5
lscd it's not, beyond the fact that you use it to separate tokens, ie, you can't say intblah instead of int blah
dduncan yes, of course
but it doesn't care between different kinds of whitespace 23:40
svnbot6 r6761 | Darren_Duncan++ | /perl5/Code-Perl : updated all files to replace tabs with 4 spaces each 23:52
dduncan okay, apparently some of these have mixes of tabs and spaces too 23:57