pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:02 justatheory joined 00:10 hanson122 joined 00:12 cognominal_ left
hanson122 join #perl 00:12
00:12 hanson122 left 00:18 cognominal_ joined, polettix left 00:19 weinig left 00:27 ofer joined 00:36 justatheory left 00:39 vel left
Limbic_Region hrm - if dmake broke then like nmake will too 00:42
:-(
oh wait, it appears all that is needed is a realclean
diakopter - you around 00:43
I am going to email maxcdo 00:44
00:47 Schwern joined
diakopter I'm around 00:47
Limbic_Region salutations Schwern
diakopter - what were you going to ask maxcdo? 00:48
diakopter nothing worth anything
you figured it out I see
Limbic_Region oh, was wondering if you had some advice
well, it hasn't finished building yet
but it didn't bomb out with unsatisfied dependency this time
so where did the 6.6.1 dependency come from? Was there a major checkin or something? 00:51
diakopter audreyt bumped it
Limbic_Region I ask because I worked 53 hours this week even though I took a day off
Limbic_Region was a bit busy
diakopter I don't recall the exact reason
it's in the logs
Limbic_Region oh, wonder if the pad refactor went back in
diakopter I don't think so 00:52
Limbic_Region hrm 00:53
The recent commits link of pugscode.org and the RSS are both not working
diakopter urgh
Limbic_Region is there a way to get just the commit message for like the last 20 checkins from svn? 00:54
my svn fu is weak
oh, pugs did build after the realclean 00:55
diakopter dev.pugscode.org/timeline
Limbic_Region that's not loading here 00:56
diakopter it is here...
dns problem?
Limbic_Region no, it just took like 20 seconds
diakopter 8.61s here :)
Limbic_Region thanks btw 00:57
diakopter if you haven't sent that email to that person, tell him/her to join us here! :)
er, ask, not tell. 00:58
Limbic_Region oh, already sent 00:59
but yeah
I got the email from the ticket
oh, the 6.6.1 dependency was for date/time package apparently 01:01
but lots of other code cleanup as the result of a utility
audreyt++
01:02 Aankhen`` left, ewr left 01:03 ewr joined
diakopter Limbic_Region: TimToady added a TODO earlier, informally 01:06
spec the roles that will be assumed by the built-in types, immutable and mutable, container and value. 01:08
which I guess will largely be a reorg/fleshing of S29
Limbic_Region I'm a bit too tired to look now but perhaps tomorrow 01:10
01:10 Schwern left, jisom_ left
Limbic_Region worked 38 hours straight and then after a night's sleep just pulled another full day (on Saturday) 01:10
diakopter icky
overtime++ 01:11
Limbic_Region yeah - especially after I told them from the beginning that no amount of testing was going to get the build blessed for production
actually, overtime--
I work for the government where the overtime is capped at a certain rate - so if you already make more than that rate, they pay you less per hour than regular time
comptime++
01:19 Schwern joined 01:34 rashakil joined 01:37 Limbic_Region left
diakopter requests another request for comment request, for fun. 01:38
01:43 Schwern left 01:59 ruz joined 02:01 kanru joined 02:22 mako132_ joined 02:29 Schwern joined 02:30 thorat left 02:33 Schwern left, jisom joined 02:34 bonesss left, Schwern joined 02:36 mako132_ left 02:50 Squern joined, Schwern left 02:52 Squern left 02:54 Aankhen`` joined 02:56 Squern joined 02:59 Squern left 03:23 nperez joined 03:24 jisom left 03:26 ewr left, ewr joined 04:02 jisom joined 04:08 b00t joined 04:11 dactyli joined 04:12 ChanServ sets mode: +o diakopter, diakopter sets mode: +b *!*@one.hand.on.mouse.and.one.hand.on.my-dick.biz, dactyli was kicked by diakopter (on popular demand, you are now leaving the channel)), diakopter sets mode: -o diakopter
spinclad diakopter++ 04:13
diakopter 'course, now the banlist is ****flooded.
can only ops see the banlist? 04:14
spinclad dunno... sure, can't print a banlist like that without *'s, it weren't right 04:15
04:16 pro21ms4 left
spinclad no, i can see it too 04:17
diakopter <sigh>
04:18 ChanServ sets mode: +o diakopter, diakopter sets mode: -b *!*@one.hand.on.mouse.and.one.hand.on.my-dick.biz, diakopter sets mode: +b *!*@one.hand.on.mouse.*
spinclad but at least the entity is not itself present here 04:18
04:19 diakopter sets mode: -o diakopter
diakopter I think that works. 04:19
spinclad that looks good ... or *.biz even
diakopter lol
diakopter checks the irclogs for fun
sure enough, since the beginning of my logs... all .biz PTRs have been part-spammers or PTR-obsceners. 04:22
04:29 araujo joined
spinclad compugoat and this guy 04:29
anyway, /me to bed. & 04:33
04:38 jisom_ joined, jisom left 04:48 jisom joined
Aankhen`` diakopter: PTR? 04:48
Pointy Throated Rodent?
04:50 jisom_ left 04:52 xinming joined 05:04 nperez left, jrockway joined 05:08 amnesiac joined 05:09 nperez joined
diakopter Aankhen``: I may be wrong in assuming that's where ircd gets those hostnames, but I was referring to the "reverse DNS lookup" or PTR ("pointer") record for a given IP address. 05:11
05:15 jisom left
Aankhen`` diakopter: Ah. 05:18
05:20 rlb3_ left 05:41 BoogieMan joined 05:48 ewr left, ewr joined 05:59 hanson122 joined
hanson122 join #perl 06:00
06:00 hanson122 left 06:18 Aankhen`` left 06:19 bsb joined 06:24 Alias_ joined 06:25 pro21ms4 joined 06:28 ewr left 06:35 RFox joined
pugs_svnbot r17063 | Darren_Duncan++ | trunk: 06:38
: ext/Muldis-DB/ : updated Core.pod to add the (large) initial complement of tuple and relation operators
diff: dev.pugscode.org/changeset/17063
lambdabot Title: Changeset 17063 - Pugs - Trac
06:44 weinig joined 06:48 weinig left 06:51 Aankh|Clone joined 06:52 funktio joined
devbot6 dev: Ticket #26 (defect closed): [BUG] 'dmake' broken on Win32 with GHC 6.6.1 <dev.pugscode.org/ticket/26#comment:1> 07:01
07:02 Jmax left
devbot6 dev: Ticket #26 (defect updated): [BUG] 'dmake' broken on Win32 with GHC 6.6.1 <dev.pugscode.org/ticket/26#comment:2> 07:03
07:25 amnesiac left 07:56 protoscript joined
protoscript Hey guys 07:56
jrockway hi protoscript
07:56 elmex joined
protoscript I was wondering, why did perl6 decide to make operators special things instead of just methods... that prevents using the hyperoperators syntax as a map 07:56
jrockway what do you mean? 07:58
(maybe nopaste what you'd like to write) 07:59
dduncan perl operators are just a special syntax for methods, for the most part 08:00
protoscript ok
it's just one line
@bad_employees >>fire<<
lambdabot Unknown command, try @list
protoscript it complains that fire should be an operator 08:01
dduncan try spelling it .fire
protoscript ah ok
dduncan ?eval @bad_employees >>.fire<<
evalbot_r17041 Error: ā¤Unexpected ">>."ā¤expecting operator
dduncan may need work
protoscript < 1 2 3 > >>.say<<
doesn't work either 08:02
dduncan at least you can use the dot form in other places that take operators
protoscript I was really curious regarding the philosophy behind the decision
dduncan eg, $foo .bar $baz;
er
protoscript since I thought one of the greatest achievements of ruby, (smalltalk too?) was to make "operators" just funny named methods
well that is cool too 08:03
dduncan I was trying for the analogy for $foo += $bar, but replacing + with .meth
$foo .meth= $bar
maybe?
protoscript I could never stomach perl < 6 but perl6 seems really great
dduncan perhaps reduction syntax works .... $foo = [.meth] @bar
audreyt ?eval [1, 2, 3]>>.sqrt
evalbot_r17041 (1.0, 1.4142135623730951, 1.7320508075688772)
08:04 rho joined
audreyt you only need one side of >> 08:04
because .method is not binary op
protoscript I'm silly regarding aesthetics and -> syntax and the oo stuff in perl5 just grated on my nerves
08:04 pro21ms4 left
dduncan audreyt, since you're here, I'll say that everything ran to completion for me yesterday ... build and smoke completed with success ... perhaps a bit faster than my last try also 08:04
audreyt ?eval <1 2 3>>.sqrt
protoscript ?eval [1,2,3] >>.say
evalbot_r17041 Bool::True
Error: ā¤Unexpected ">>."ā¤expecting operator
audreyt ?eval <1 2 3>>>.sqrt
evalbot_r17041 (1.0, 1.4142135623730951, 1.7320508075688772)
audreyt ?eval <1 2 3>>>.say
evalbot_r17041 OUTPUT[2ā¤3ā¤1ā¤] (Bool::True, Bool::True, Bool::True) 08:05
audreyt dduncan: cool!
protoscript: note that it says "2\n3\n1\n"
because hyperops runs, conceptually, in parallel
dduncan audreyt, what do you see as the main preconditions now for a 6.28.0 release?
audreyt so the order of side effects, such as input/output, is not guaranteed
dduncan: why, mergeback of my pugs-mo local branch
protoscript yeah !
dduncan so that isn't what you did yesterday
any eta for that? 08:06
protoscript very interesting audreyt
audreyt no it isn't. yesterday was just excising dead code
playing with HPC, etc
I think my condition is improving quite a bit this week
so maybe I'll try hacking for real next week... but no promises
protoscript where is the best place online to get docs rearding pugs perl6?
audreyt mostly it's incessant sleep
dduncan try pugscode.org
protoscript hope you feel better audreyt :-) 08:07
dduncan yes, be well
audreyt protoscript: dev.pugscode.org perhaps?
www.perlfoundation.org/perl6/ too
protoscript thank you
audreyt np :) 08:08
audreyt fades, unsurprisingly, back into sleep &
dduncan and I"m about to enter sleep myself
1am
08:08 dduncan left
protoscript are there any cute web frameworks written in perl6? 08:08
jrockway i'm sure there is some jifty6 floating around somewhere (at least on slides) 08:10
audreyt would be the person to ask though
jrockway likes his perl5 web framework ;)
08:10 ruz left
protoscript ok thanks jrockway 08:10
what framework is that?
jrockway Catalyst
protoscript Do you think the codebase could be improved porting it to 6? 08:11
jrockway definitely
if perl6 didn't exist, we could get by with perl5
protoscript (I mean theoretically, since its probably a giant undertaking)
jrockway but there are some perl5-isms that hurt us
like slow method calls
protoscript I've heard from the experts at canonical that catalyst is the only dynamic lang web framework period that gets orm rights 08:12
jrockway we are going to port it to Moose soon, which should at least make the code look nicer
protoscript -s
protoscript googles moose
jrockway Catalyst doesn't force any ORM on you, but a lot of people use DBIx::Class
and I agree that it gets ORM-ing done right
protoscript I do python dev for money 08:13
jrockway the nice feature is that you can say my $something = $rs->search({ some conditions }); ...; $something->search( { more conditions } )
everything is pretty lazy, so you'll get an efficent query that way (and readable code of course)
protoscript: moose is at search.cpan.org/dist/Moose/lib/Moose.pm 08:15
lambdabot Title: Moose - A complete modern object system for Perl 5 - search.cpan.org
08:15 protoskript joined
protoskript grumblez 08:15
jrockway i have a few catalyst talks at www.jrock.us if you're interested 08:17
(the YAPC::NA::2007 ones are better than the YAPC::Asia::2007 ones IMHO)
protoskript absolutely thanks
jrockway jifty is another perl5 framework that is more integrated and magical 08:18
it's pretty cool if you're into that kind of thing ;)
protoskript yeah it looked super-cute 08:19
just have so much trouble with perl5 oo and some of the syntax
protoskript is a shameful languae snob
jrockway fair enough
the object system kind of takes a while to grasp 08:20
once you are familiar with the conventions (and Class::C3), it's quite usable
protoskript do u program in perl professionally?
jrockway but if you come from a (say) java background where everything is ultra-specific ("private int foo") it might not be your cup of tea
protoscript: yeah
protoskript My background is javascript 08:21
Which made me the subject of much derision in 1999 when I tried to explain that I could write great server-side apps in javascript 08:22
"Why not just use vbscript?" o_O
jrockway javascript is a fine programming language 08:23
the "rhino on rails" might give it some additonal credibility
protoskript Agreed. No one really understood that back in the 90s
But AJAX has given it much more cred
jrockway exactly
what's funny is that AJAX was technically possible long before it became popular 08:24
protoskript for career reasons i had to move to C#, then ruby then python which I use today
I know I used to do ajax-ish stuff in 2000
I had a chat room that integrated with a dialer system for a call center
just took google to have everyone say ':ok i guess it is reasonable" 08:25
jrockway yeah
same for python too :)
protoskript indeed. ever see prothon back in the day? 08:26
that was an incredibly awesome language
so sad that it was ditched and now i cant even find material on it
jrockway no, didn't see it 08:29
protoskript It was a prototype-based python that fixed some of the goofy python decisions
aaaanyways sorrt for the languages tangent 08:30
jrockway now called mochikit ;)
protoskript haha
08:32 protoscript left 08:34 protoskript left
jrockway gnite :) 08:34
08:35 protoskript joined 08:38 ruz joined 08:39 Squern joined 08:41 Squern is now known as Schwern
protoskript hmmm catalyst looks really nice 08:47
08:55 monomorph joined 09:06 iblechbot joined 09:10 rindolf joined 09:31 riffraff joined 09:33 masak joined 09:37 bsb left 09:40 DarkWolf84 joined 09:45 protoskript left, iblechbot left 09:56 Aankh|Clone left 10:04 buetow joined 10:07 Blwood joined 10:09 BoogieMan left, ruoso joined 10:13 perlMa joined 10:14 Blwood left 10:27 Southen joined 10:34 RFox left 10:36 ludan joined
ludan hola 10:36
10:40 riffraff left, chris2 joined 10:46 |perlMa| joined 10:48 Nadine joined
Nadine hello 10:49
10:49 Nadine left 11:04 asksh joined 11:06 Blwood joined 11:13 perlMa left 11:14 ask__ joined 11:21 larsen joined 11:23 asksh left 11:36 rindolf left 11:37 masak left 11:38 perlMa joined 11:39 perlMa left 11:43 kanru left 11:51 pro21ms4 joined 11:53 ofer left 11:54 penk left 11:57 |perlMa| left
TimToady well, got to .is safely if anyone is listening. or even if not... 11:59
Alias_ .is?
Iceland? 12:00
Isreal?
Israel
12:00 ofer joined
TimToady actually, Iceland 12:01
12:01 ofer left, ofer joined 12:13 riffraff joined 12:17 ofer left
araujo doesn't even know where it is that 12:18
Alias_ araujo: half way between canada and finland 12:21
12:26 rho left, hanson122 joined 12:27 hanson122 left
araujo oh.. mm 12:29
12:30 vel joined 12:54 riffraff left 12:55 JarJarBinks joined 12:58 vel left 13:04 vel joined 13:06 hakobe_ is now known as hakobe 13:17 buetow left, buetow joined 13:24 pro21ms4 left
moritz TimToady: how do you like Iceland? I always wanted to visit it ;) 13:27
TimToady there's a lot of lava 13:28
13:29 vel_ joined
moritz ;) 13:29
13:30 rho joined 13:31 vel_ left, Alias__ joined, vel_ joined 13:37 buetow left, buetow joined 13:38 chris2 left, masak joined 13:43 Alias_ left 13:44 vel left 13:56 Limbic_Region joined
moritz tries to compile ghc 6.6.1 from Debian testing for sarge 14:06
do we _really_ need 6.6.1? ;)
masak ;)
I'd complain as well, but my compilation worked :P 14:07
moritz let's see if mine does
at least all build-dependencies are in etch
or at least as debian/control lists them 14:08
masak I did it from source, that usually works fine on CentOS
moritz masak: did you have to update supplementary packages like libghc6-network-dev or the like?
masak moritz: nope
moritz masak: that's good news ;) 14:09
14:11 b00t left
moritz compiling ghc takes quite a while - I wonder if it bootstraps itself 14:13
14:15 vel joined
masak moritz: don't remember, but it wouldn't surprise me 14:16
moritz needs more RAM and a faster maschine ;) 14:21
14:35 larsen left 14:36 vel_ left
devbot6 dev: Ticket #27 (defect updated): make realclean rebuilds the makefile instead of make-ing realclean <dev.pugscode.org/ticket/27#comment:1> 14:39
diakopter moritz: it does bootstrap itself with a downloaded binary on macintel (using [Mac]port[s]) 14:40
14:41 jedai joined 14:43 ludan left
diakopter moritz: but it still took an hour on dicore2Ghz/2GBram. :( or 2 hours. 14:44
moritz diakopter: thanks, I'll just wait until it's finished - no hurry to rebuild pugs ;) 14:45
diakopter moritz: which platform/architecture have you? 14:46
moritz diakopter: pentium M (i386) with debian 14:47
14:50 phileas_fox joined
diakopter moritz: ah. Lenny has 6.6.1... 14:50
moritz diakopter: yes, and I'm recompiling its source packages 14:53
diakopter ah :D
moritz diakopter: because the binary package in lenny has dependencies on a new libc6 (which conflicts with a "base" package) 14:54
but that is an automatically pulled in dependency, not a necessary one
diakopter moritz: I see. would you mind posting a tiny howto here for how to do that for a stock Etch system? 14:55
(I mean, I hope it's tiny)
14:56 daxim_ joined, ChanServ sets mode: +o diakopter
moritz diakopter: sure, just a sec... 15:02
echo 'deb-src ftp://ftp.debian.org/debian/ lenny main contrib non-free' >> /etc/apt/sources.list
apt-get update
apt-get build-dep ghc6
apt-get source --compile ghc6 15:03
dpkg -i $resulting_debian_package # I'm not there yet :(
diakopter: that's about it 15:04
it looks like I've entered another phase of the bootstrapping process 15:05
diakopter moritz: thanks :) 15:07
moritz diakopter: you're welcome
diakopter starts that on an Etch vps
15:08 aindilis left, aindilis joined
moritz btw I didn't use the --compile option (I forgot about it), but stepped into that directory manually, and started dpkg-buildpackage -rfakeroot -b -uc 15:08
diakopter did apt-get build-dep ghc6=6.6.1-2 just in case 15:10
15:10 daxim left
moritz diakopter: not necessary unless you use some evil pinning or the like 15:10
15:10 Alias_ joined
diakopter moritz: :D rafb.net/p/SnrJal80.html 15:12
lambdabot Title: Nopaste - /etc/apt/*
moritz diakopter: seeing those pins I don't think you need my advise ;) 15:13
diakopter yes I did, I haven't before built from a deb-src... 15:14
moritz ok
15:15 Alias__ left
moritz pinning scares me... when I read the corresponding manpages and then edit /etc/apt/preferences it never does what I want... 15:15
diakopter I learn quickly, but I forget even more quickly. But I usually don't forget the event of learning (and then usually forgetting) something. and I don't remember learning build-dep --compile :)
moritz and then I change something completely stupid, and it works. hopefully
build-dep is *really* handy ;) 15:16
diakopter It took probably 1.5-2 hours of reading various todo posts to come up with that one. I left out parts of the /etc/apt/preferences... here's the whole thing so you can see how I used volatile... rafb.net/p/FCP3tV73.html 15:17
lambdabot Title: Nopaste - /etc/apt/*
15:17 Limbic_Region left
diakopter unfortunately you can't glob the Package: line for multiple packages. :( :( (clamav*, e.g.) 15:18
oh, I forgot - cat /etc/apt/apt.conf 15:22
APT::Default-Release "etch";
with those 3 files set up like that, the minimum required version of a package provided by sources.list is always installed, no matter where you apt-get install from. (e.g. apt-get install spamassassin/unstable) 15:26
I tried all kinds of other configurations, but it wasn't until that arrangement that it stopped wanting to unnecessarily upgrade stuff. 15:27
diakopter tries to stop being offtopic 15:28
moritz diakopter: it's _all_ about getting pugs to compile again, isn't it? *g* 15:30
diakopter sure!
moritz: oh yeah. that's what I was doing. :/
masak :)
15:32 justatheory joined 15:36 dmpk2k|z joined
diakopter wow...... now that's a long command-line. 152 columns by 55 rows, nearly full. 15:38
15:39 justatheory left
diakopter read through 'Once and Future Perl' yesterday 15:42
masak diakopter: book? article? 15:46
diakopter LAW ppt slidedeck
masak LAW?
diakopter oops 15:49
masak I guess my question is: does it have an URL? 15:50
15:51 joao joined
diakopter looks 15:53
15:56 iblechbot joined 16:24 pmurias joined, pmurias left 16:26 masak left 16:27 Blwood left 16:40 ashleyb joined, nperez left 16:50 ashleyb left 17:00 aindilis left 17:01 aindilis joined 17:08 fridim joined, amnesiac joined 17:16 Kattana joined
rhr so I'm trying to build a new ghc according to the directions moritz++ gave and have run into a problem... 17:21
ghc 6.6.1 built and installed, but it looks like I have to rebiuld libghc6-filepath-dev and friends. when I go to install the build-deps for libghc6-filepath-dev it wants to remove the old version but fails: 17:22
ghc-pkg: cannot find package FilePath-0.11
and apt is now stuck not able to do anything because of this 17:23
diakopter rhr: you're on etch? 17:26
rhr yup
diakopter rhr: if I understand it correctly, I think you should remove libghc6-filepath-dev 17:28
rhr the above error is what I get with the old ghc6 installed. with the new one I get: /var/lib/dpkg/info/libghc6-filepath-dev.prerm: line 13: /usr/lib/ghc-6.6/bin/ghc-pkg: No such file or directory
diakopter then reinstall the .deb you built
rhr diakopter: I can't
the error prevents it from being removed
diakopter hrm
I guess you'll have to force or manually remove it 17:29
rhr how does one do that?
diakopter gappy index. consults master index. 17:30
(my build keeps failing b/c that machine is running out of ram continually... 17:31
rhr I've never encountered the situation where dpkg --purge doesn't work
diakopter hrm; I didn't even have libghc6-filepath-dev installed. 17:32
rhr even dpkg dpkg --purge --force all doesn't work 17:36
diakopter same error message? 17:37
rhr yup
diakopter rhr: replace the libghc6-filepath-dev section in /var/lib/dpkg/status with: Package: libghc6-filepath-dev 17:38
Status: purge ok not-installed
Priority: extra
Section: devel
then try apt-get install -f maybe? 17:39
rhr ok, that got apt unstuck. thanks! 17:40
17:40 iblechbot left
diakopter yw 17:41
rhr now I'll try to build the other stuff
diakopter tries building ghc6=6.6.1-2 with 600 free MB this time :/ 17:43
rhr so pugs wants libghc6-mtl-dev but the build deps can't be satisfied on etch :( 17:44
I'll try to build haddock, maybe that will get it... 17:45
17:46 Psyche^ joined, kanru joined 17:58 Patterner left, Psyche^ is now known as Patterner
moritz rhr: I'm trying that as well 17:59
rhr so far I've had to build haddock haskell-utils libghc6-time-dev libghc6-mtl-dev libghc6-network-dev 18:00
pugs is compiling now...
18:01 devogon joined
moritz I already built pugs, but I have to rebuild ghc6-mtl 18:01
and haddock as a build dep
18:03 DarkWolf84 left, DarkWolf84 joined 18:04 DarkWolf84 left 18:09 rindolf joined
diakopter 's ghc6.6.1 build seems to have made it to rts 18:11
moritz dispairs 18:13
rhr in case anyone else wants to avoid compiles, I put all the .debs I built at airen.bcm.umontreal.ca/ghc-pugs-etch/ (amd64 only) 18:14
lambdabot Title: Index of /ghc-pugs-etch
rhr not apt-gettable
moritz rhr++
18:15 funktio left
moritz don't you want to do the same with i386? 18:18
diakopter is doing that..
(we might see about adding them to backports.org?)
moritz diakopter: that would be great 18:19
perlbot: nopaste 18:20
perlbot Paste your code at sial.org/pbot/perl and #perl will be able to view it
lambdabot Title: sial.org Pastebot - pasteling
18:21 BoogieMan joined
pasteling "moritz" at 84.148.51.222 pasted "install problems with mtl" (16 lines, 841B) at sial.org/pbot/26317 18:21
18:21 kanru left
moritz any ideas concerning regarding my paste? 18:21
rhr a backport would be very cool. I can put debs for other archs up if someone supplies them.
moritz: that looks very similar to what happened to me 18:22
moritz rhr: so how did you fix it?
rhr if apt and dpkg won't remove the old version, you have to edit /var/lib/dpkg/status manually as diakopter++ describes above 18:23
change the status line for the package to Status: purge ok not-installed 18:24
worked for me anyway... 18:25
diakopter rhr: although... now that I think about it... since that package is provided by the ghc6... that might confuse things 18:27
rhr which package?
diakopter so maybe that section should be removed entirely. where's #debian when you need it?
moritz /var/lib/dpkg/status doesn't list libgh6-filepath-dev 18:28
and ghc6 provides it
rhr in my case it was the filepath package that was the trouble, and it turns out not to be necessary...
18:29 cmarcelo left
rhr moritz: you might try re-installing all of the old ghc stuff and then removing all of it before installing the new debs 18:30
diakopter seconds that suggestion; over and out
rhr I don't really understand the underlying problem though, so listen to me at your peril :) 18:31
moritz rhr: good idea ;)
18:49 rho left 18:51 baest joined 18:54 weinig joined, weinig left 18:55 weinig joined, joao left, joao joined 18:58 Blwood joined 19:16 r0bby left, r0bby joined 19:20 amnesiac left 19:21 fridim left 19:25 amnesiac joined 19:27 Daveman joined 19:28 frobnitz joined
moritz I had to overide the build deps for lighc-network-dev 19:36
but now Makefile.PL went through cleanly
rhr moritz: what was the missing build-dep? I had to build haddock and haskell-utils in addition to the ghc stuff, but that satisfied everything for me 19:46
moritz @tell audreyt pugs build requires time-any, but Makefile.PL doesn't check for it
lambdabot Consider it noted.
moritz rhr: dpkg-checkbuilddeps: Unmet build dependencies: ghc6-doc libghc6-base-doc libghc6-parsec-doc 19:47
rhr oh yeah, that comes from the ghc6-doc deb you get when you build ghc 19:48
I should write up a "how to install ghc 6.6.1 on etch" thing for INSTALL 19:49
moritz rhr: aye, but ghc6-doc doesn't install for me (dies with a haddoc error)
rhr did you build your owh haddock from testing?
I didn't have that problem 19:50
diakopter @tell audreyt see ticket 27 please 19:51
lambdabot Consider it noted.
moritz S05/"Nothing is illegal" # irc log test 20:23
Tene moritz++ # irc log hacking 20:26
20:29 Aankhen`` joined 20:49 jisom joined, joao left 20:54 ashleyb joined
pugs_svnbot r17064 | rhr++ | trunk: 20:55
: [INSTALL] Instructions for installing new GHC on Debian etch
diff: dev.pugscode.org/changeset/17064
lambdabot Title: Changeset 17064 - Pugs - Trac
20:57 jrockway left
rhr if anyone tries the ghc install procedure for etch, please report your success or failure 20:58
20:58 jrockway joined
Aankhen`` rhr: I just used aptitude. 20:59
rhr does it do all of that for you? 21:00
or did you install libc6 etc. from testing? 21:02
Aankhen`` One sec, I haven't looked at the instructions yet.
21:02 polettix joined
Aankhen`` Is dev.pugscode.org down? 21:03
rhr not here...
Aankhen`` Hmm, looks like I can't access any sites at all.
Well, not quite true. I can't access anything but Google.
21:04 Aankhen`` left 21:06 Aankhen`` joined
pugs_svnbot r17065 | moritz++ | trunk: 21:10
: [irclog]
: * substituted spaces with &nbsp; in actions in the nick column
: (prevents line breaks)
: * fixed OSCON in abbr.dat
: * added synposis links of the form S05/"Nothing is illegal" as suggested by
: [particle]++
diff: dev.pugscode.org/changeset/17065
lambdabot Title: Changeset 17065 - Pugs - Trac
21:10 jedai left
moritz when I do a 'svn log --limit 5' just after a commit, why can't I see my own last log message? 21:11
ashleyb TimToady: I would love to hear your take on this conference you're at, (sort of a wrap up report), or if you have a pointer to slides that'd be cool too.
Aankhen`` rhr: I was wrong, looks like I only have GHC 6.6. 21:12
rhr Aankhen``: ok, you can be a guinea pig :) 21:20
pugs_svnbot r17066 | rhr++ | trunk:
: [INSTALL] corrections to etch ghc install instructions
: moritz++ diakopter++
diff: dev.pugscode.org/changeset/17066
lambdabot Title: Changeset 17066 - Pugs - Trac
Aankhen`` rhr: I'm not sure I understand. Do the instructions only work on AMD64? 21:24
diakopter Aankhen``: can you get to dev.pugscode.org yet? 21:25
Aankhen`` diakopter: Yup, it was a DNS problem on my end.
rhr no, just that there are prebuilt debs for amd64
Aankhen`` rhr: Oh, okay.
diakopter and i386
rhr diakopter: where are the i386 debs?
diakopter sec 21:27
pugs_svnbot r17067 | moritz++ | trunk:
: [irclog] disabled xhtml header for spam.pl
diff: dev.pugscode.org/changeset/17067
lambdabot Title: Changeset 17067 - Pugs - Trac
diakopter feather.perl6.nl/~diakopter/
lambdabot Title: Index of /~diakopter
diakopter oops; gotta re-add the /icons alias 21:28
Aankhen`` is following the instructions in INSTALL.
Tene submits a patch to INSTALL adding the line "Mail $20 to Tene" 21:29
Aankhen`` adds to that line: , c/o Aankhen 21:30
rhr Aankhen``: thanks for testing 21:31
pugs_svnbot r17068 | rhr++ | trunk:
: [INSTALL] diakopter++ has i386 ghc debs
diff: dev.pugscode.org/changeset/17068
lambdabot Title: Changeset 17068 - Pugs - Trac
Aankhen`` rhr: NP.
rhr: What's your address, just in case I need to send you nasty packages? :-D
rhr [email@hidden.address] 21:32
Aankhen`` Hmm, I guess that'll do.
BTW, it's generally a good idea to say: [off] [email@hidden.address] 21:33
So it's not logged.
rhr ah, didn't know about that. I think all the spammers have my address already though...
21:34 lisppaste3 joined
Aankhen`` Heh. 21:34
21:35 amnesiac left
Aankhen`` I used to get four times more spam on my private address compared to my public address. 21:35
rhr up until a few years ago I got almost no spam at all, then somehow the floodgates opened. bogofilter has done a pretty good job with it though. 21:37
moritz wonders how much spam [email@hidden.address] and [email@hidden.address] receive ;)
Aankhen`` LOL..
21:38 bonesss joined
moritz Aankhen``: have you ever visited exampl.com? *g* 21:38
s/exampl/example/
rhr wow, it used to not resolve
moritz if the domain was available i'd buy it and put up code examples 21:39
and get thounds of backlenks for free ;) 21:40
*thousands
21:41 mj41 joined
SamB man that HTML is old school 21:42
you can actually read it!
moritz SamB: the HTML of all my homepages is very readable 21:45
21:45 rindolf left
moritz SamB: perhaps with the exception of the irc logs ;) 21:45
which are overtuned and just too much data
SamB I'm betting there is no MX record for example.{com,net,org} 21:46
and no smtpd either
moritz Aye 21:47
21:49 ruoso left
moritz bed & 21:50
22:10 monomorph left 22:16 mr_ank left 22:21 aindilis left, aindilis joined 22:28 marmic joined 22:29 Aankhen`` left 22:30 moritz left, moritz joined 22:31 moritz left 22:35 Aankhen`` joined, phileas_fox left, moritz_ joined 22:38 vel left
Aankhen`` Yeesh, GHC still compiling. 22:38
diakopter Aankhen``: what architecture you on? 22:42
Aankhen`` IA-32?
diakopter uname -a 22:45
you didn't want to use the .debs I posted?
22:46 aindilis` joined
Aankhen`` Oh, i686. 22:46
diakopter :D
k
oh, actually... 22:51
hrm. I don't understand why uname -a says i686 for me as well, but the .debs that got built say i386
Aankhen`` makes threats in GHC's direction to try to get it to build faster. 22:52
diakopter silly VPSes. 22:53
22:54 miyagawa joined
SamB diakopter: i386 is another name for IA-32 22:56
i686 is a kind of IA-32 as well
22:59 SubStack joined, aindilis left 23:01 mr_ank joined 23:08 BoogieMan left 23:09 dmpk2k|z left
Aankhen`` rhr: Ping. 23:15
rhr: At step 3 ("Install new GHC", "dpkg -i ghc6_6.6.1-*.deb"), I get a warning about not being able to delete the old GHC directory because it's not empty, and errors about dependencies not being satisfied. Is that alright? 23:16
(I'm guessing the dependencies thing is okay since the next step is to install them, but I'm worried about not being able to delete the old directory.)
rhr Aankhen``: I got the warning about the directory also 23:17
what is the dependancy problem? 23:18
Aankhen`` One sec, I'll put it on pastebin.
pasteling "Aankhen" at 24.80.73.85 pasted "GHC install error" (12 lines, 563B) at sial.org/pbot/26323
rhr I'm not sure what the deal is with the directory, but its not being deleted didn't seem to harm anything
23:19 bsb joined
rhr hmm, I guess you should reinstall haskell-utils and then install the new ghc again 23:20
Aankhen`` apt-get install haskell-utils?
rhr the etch haskell-utils
yes
Aankhen`` OK.
23:21 rlb3 joined
Daveman Aankhen`` :) 23:25
23:26 Eidolos left 23:27 BoogieMan joined 23:29 cognominal_ left 23:32 Eidolos joined
pasteling "Aankhen" at 24.80.73.85 pasted "Haddock error" (27 lines, 1.5K) at sial.org/pbot/26324 23:35
Aankhen`` rhr: Error with the doc package now.
rhr fascinating... 23:36
Aankhen`` Um, yeah, totally. 23:37
rhr you've installed the new haddock already?
Aankhen`` AFAIK, yes.
rhr then maybe try installing the old haddock, installing ghc-doc again, and then installing the new haddock 23:38
Aankhen`` OK. 23:39
Er, how
?
23:39 justatheory joined
rhr apt-get install haddock; dpkg -i ghc6-doc_6.6.1-*.deb; dpkg -i haddock*.deb 23:40
Aankhen`` `apt-get install haddock` seems to try to install ghc6-doc and fails with the same error as I pasted.
Same errors, rather.
rhr ok, can you apt-get remove ghc6-doc? 23:41
Aankhen`` Done.
rhr then retry those 3 commands
Aankhen`` OK.
haddock is already at the newest version, apparently.
23:42 mako132_ joined
rhr oh, are you looking at the latest rev of INSTALL? I changed the order of the ghc6-doc installation 23:42
Aankhen`` Nope, no good. apt-get install haddock insists that it's at the newest version already.
I believe so. One sec.
Yes, I am. 23:43
rhr in that case you should compile and install the new haddock before installing ghc6-doc, I think
Aankhen`` I followed all the instructions. :-S
rhr can you verify which version of haddock is installed?
maybe you need to apt-get remove haddock and then apt-get install haddock to get the old version 23:44
if it's the new version that's causing the problem 23:45
Aankhen`` Okay, did that.
rhr ok, now can you dpkg -i ghc6-doc_6.6.1-*.deb ? 23:46
Aankhen`` Same error. -_-
rhr hmm, apt-get remove haddock; dpkg -i ghc6-doc_6.6.1-*.deb ? 23:47
Aankhen`` Now it says haddock is not installed.
(When it reaches ghc6-doc.)
Wait, one sec. Maybe I'm reading this wrong. 23:48
I'll pastebin it.
pasteling Someone at 24.80.73.85 pasted "# apt-get remove haddock; dpkg" (23 lines, 978B) at sial.org/pbot/26325
Aankhen`` rhr: There you go.
rhr I wish I knew what those error messages actually meant...
Aankhen`` Maybe it's trying to read new haddock docs with the old haddock? 23:49
rhr but it didn't work with the new haddock, right?
Aankhen`` I think the new haddock just didn't install at all.
rhr oh, what was the problem there?
Aankhen`` It went through the motions but doesn't seem to have had any effect.
I'll try going through the steps for haddock again. 23:50
rhr huh, ok 23:51
Aankhen`` Nah, no luck. 23:55
BTW, I'm not sure if this is all I'm supposed to get... 23:56
pasteling Someone at 24.80.73.85 pasted "# dpkg -i haddock*.deb Selecti" (6 lines, 235B) at sial.org/pbot/26326
Aankhen`` It just drops me back at the prompt at that point.
rhr that looks normal to me 23:58
Aankhen`` Ah.
rhr let me try it...
Aankhen`` Well, whatever it is, ghc6-doc is still not happy.

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!