#musicbrainz-devel

!Help I'm debating whether or not to convert the vote display from a table into a dl, but the one thing that's stopping me is that if I don't use a table the name column get's all screwy with different width names :( http://xabbu.net/mb/edit15.png

Advice please!

navap: you could set the width of the left part of the dl (although I'm not sure what would happen with a name longer than the width)

Yeah, and short names end up with excess space on the left :/

I would probably just leave it as a table, it has rows and columns so it seems enough like tabular data to me :P

Hi all, trying to submit my application, does anyone have any ideas as to how to submit code samples?

Should I just copy/paste them right into the editor?

Is the code available elsewhere online? It might be better if you were able to provide a link to a CVS of some kind.

it's just local

hmmm

I could upload it to my webpage and link, that's probably easiest

or put them on pastebin/gist

and link to that

Ahh, the elusive haircut actually happens today

or so it thinks

nope, mom booked it! no going back now

hahaha

it must be getting /really/ bad...

it's over 5 weeks overdue

it's bad :P

aCiD2`: So will you merge in http://codereview.musicbrainz.org/r/636 please?

hello!

TOD!

hi warp

hello! :)

navap: yea, will do it shortly

aCiD2`: I'm messing with MOD_SET_RELEASE_DURATIONS

warp: Thanks for the seconding by the way :)

aCiD2`: the old edit has a cdtoc id, can I just use the same id in the new edit?

"INSERT INTO cdtoc SELECT * FROM public.cdtoc;" says yes :)

good :)

How does foreign_keys work for that?

you want to load a CDTOC object?

yes

Well, Data::CDTOC extends Data::Entity, which means it should have get_by_ids. So CDTOC => [ $self->data->{cdtoc_id} ] should work

I made it cdtoc_ids, something was complaining about it not being an ArrayRef with just an Int in data.

well, 5 is not the same as [ 5 ]

I really need to patch MooseX::Types::Structured to actually give half useful output. Truncating the "actual value" that's displayed is useless, and it's impossible to see which field actually failed to validate

I know :)

anyway, for now I just do cdtoc_ids => [ $cdtoc ], in upgrade which MooseX::Types::Structured is happy with.

You could do ArrayRef[Int] | Int, but that's just setting yourself up for more headaches

ah, wee. I get the actual discid visible in the template now.

http://en.wikipedia.org/wiki/Main_Page bwahahah, nice news today :D

I like the last one

:)

ok, now that I got it running I understood enough of it to turn that ArrayRef[Int] back into a regular Int

aCiD2`: oh, I didn't create a test for EditTrackLength yet. Your example ( t/edit-migration/edit-release-name.t ) uses Test::Mock::Context which I still cannot install.

bah, no worries

yay for doing 4 hours work in 10 minutes

:)

warp: are you happy with your icu extension?

Also, mbs-136 is about sorting on the website, but your review doesn't actually change any of the Perl code

r/612 is the first step for MBS-136. It's not enough to close the ticket, no.

aCiD2`: hm, I don't supposed the range operator (..) can step?

s/ed/e/

warp: I don't think so

/achieve? :)

achieve?

i meant what are you trying to achieve

I have a list of (id, name, id, name, ...) values, and want to pair them up into a list of hashes

map { { id => $foo[$_], name => $foo[$_+1] } } grep { $_ % 2 } (0..$#foo)

so you want (5 'name' 6 'othername') to be.. [{ 5 => 'name' }, { 6 => 'othername' }] ?

use List::AllUtils qw(natatime);

I was thinking use List::MoreUtils 'zip' because it sounds like you have a list of ids (the range) and a list of names, and want ((id name) (id name))

no, zip zips two sides of something together into one list

you could zip @ids, @names to get the kind of list warp already has

oh, I get the list now

then yea, natatime :)

rafl: do you think autobox is worth the syntax weirdness it causes?

I had a play yesterday bringing it in for some list processing heavy stuff... and just thought the whole $whatever->map(sub { }) was too different from map { } @$whatever

well, given that the syntax weirdness it causes is round about exactly zero.. yeah :)

then I considered PerlX::MethodCallWithBlock, but I don't trust that with my children

combine it with lambda or Function::Parameters or something

I've been considering new code to use MooseX::Types, MethodSignatures and Autobox, but I can't convince myself to use Autobox, MethodSignatures scares me with the whole slow down stuff - and Types we are half using and I have no problem with :)

i believe @list->map(f ($x) { $x ** $x }) to be rather pretty

ahh, Function::Parameters is what provides f

hey all

i was wondering wtf that was in the last snippet I read a few days ago

yes, I do this with Function::Parameters, that's quite a bit more readable than: @list->map(sub { my $x = shift; $x ** $x })

well, it provides 'fun' or something, but i changed it to be whatever you like at import time

well, you still have $_, so it could be ->map(sub { $_ ** $_ }) as well

I've got quite a distate for $_ after using it too much

which IDE can you suggest for perl development?

jensl: emacs :)

that's what me and warp use

weirdos

acid2: i knew this is coming up :)

right on

ok fine then i stick to emacs :)

anyway.. i don't even have to think about using autobox or not anymore. it's a lovely module with basically no drawbacks that you don't already have by using moose

MethodSignatures is a whole other thing

jensl: I have some customizations you might want that make perl coding a bit nicer

it really depends on which one you're refering to

Method::Signatures - do not use - ever

acid2: would be great

rafl: I've considered MX::MS and M::S::Simple

I figured if we're using Moose we might as well go with MX::M::S

(for the extra validation we'd get)

MooseX::Method::Signatures - use if you want to reduce the code you write greatly, in exchange for lots of slow validation

I might wait until we launch NGS and see how much space we have to slow it down

Method::Signatures::Simple - just use it dammit. it makes you write a little less code, which will be easier to read, and has about zero drawbacks

hah

ok, that's good to hear then

jensl: http://gist.github.com/351889 that's most of the stuff I couldn't live without

acid2: thanks

http://search.cpan.org/~ewilhelm/lambda-v0.0.1/... you have got to be fucking kidding me.

Um warp, are you working on MOD_REMOVE_RELEASES?

because... I've already done that :(

aCiD2`: hardly rocket science

(lambda)

getting Devel::declare to work together with utf8.pm might be, so we could make Function::Parameters register that nice symbol for us as well :)

rafl: i meant that in a bad way, not a good way

aCiD2`: hey, it wasn't in progress on jira

yea, it seems to be the only one that /didn't/ have a time clocked for it