The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2016-03-07

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:31 jontaylor joined #mojo
01:37 jontaylor_ joined #mojo
01:42 circ-user-INaqQ joined #mojo
02:04 asarch joined #mojo
03:18 zivester joined #mojo
03:46 noganex joined #mojo
04:22 PryMar56 joined #mojo
06:07 jwang joined #mojo
06:15 * sri yawns
07:10 Lee joined #mojo
07:11 McA joined #mojo
07:16 sri looks like perl is climbing again ;p http://www.tiobe.com/tiobe_index
07:16 sri almost bigger than javascript!
07:16 sri hahaha
07:17 dod joined #mojo
07:18 coolo sri: just because "perl" are actually 2 now ;)
07:19 sri praise perl6 for saving us all!
07:20 sri rofl
07:22 coolo but perl was 3rd in 1996. no idea what was 4th - it's dead now
07:23 dod joined #mojo
07:28 sri so, do we have any more ideas for making Mojo::Template more general purpose?
07:29 sri this is where we stand now http://mojolicious.org/perldoc/Mojo/Template#SYNOPSIS
07:30 sri maybe those examples should be using something else than html
07:36 Lee joined #mojo
07:37 good_news_everyon joined #mojo
07:37 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pe7
07:37 good_news_everyon mojo/master 16043ed Sebastian Riedel: use more minimalistic examples
07:37 good_news_everyon left #mojo
07:45 dod joined #mojo
07:46 good_news_everyon joined #mojo
07:46 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pvE
07:46 good_news_everyon mojo/master 625ee42 Sebastian Riedel: performance is a feature
07:46 good_news_everyon left #mojo
08:03 sri wonder if anyone actually uses a custom template syntax, or if we could remove all those attributes for some more performance
08:19 Vandal joined #mojo
08:27 trone joined #mojo
08:28 AndrewIsh joined #mojo
08:33 trone joined #mojo
08:49 berov joined #mojo
08:52 osfabibisi joined #mojo
09:02 kes joined #mojo
09:06 absolut_todd joined #mojo
09:26 good_news_everyon joined #mojo
09:26 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pWg
09:26 good_news_everyon mojo/master 486b22a Sebastian Riedel: show an array reference too
09:26 good_news_everyon left #mojo
09:29 CandyAngel I got the sewing machine out. The Mojodress project begins!
09:30 sri :)
09:42 good_news_everyon joined #mojo
09:42 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2p4g
09:42 good_news_everyon mojo/master 5009417 Sebastian Riedel: no need for extra tags
09:42 good_news_everyon left #mojo
09:48 good_news_everyon joined #mojo
09:48 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pBB
09:48 good_news_everyon mojo/master 69ae562 Sebastian Riedel: better description for process method
09:48 good_news_everyon left #mojo
09:50 good_news_everyon joined #mojo
09:50 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pBd
09:50 good_news_everyon mojo/master ddc627d Sebastian Riedel: mention that reusing templates is much better for performance
09:50 good_news_everyon left #mojo
09:54 sri been thinking that maybe $mt->() would be a fun shortcut for $mt->process
09:54 sri but how often do you really need that ;p
10:47 good_news_everyon joined #mojo
10:47 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2prP
10:47 good_news_everyon mojo/master 87c085d Sebastian Riedel: just use the template options
10:47 good_news_everyon left #mojo
11:17 matt_ joined #mojo
11:17 matt_ hello mojo community
11:17 matt_ I have a very strange problem that I'm trying to understand
11:17 matt_ <% if (0) { %>1<% } else { %>0<% } %>
11:18 matt_ this will not display the zero
11:18 matt_ it will display other stuff, but not a zero. any idea why?
11:22 salva joined #mojo
11:29 sri works for me
11:30 sri MOJO_TEMPLATE_DEBUG=1 perl -Mojo -E 'my $c = app->build_controller; say $c->render_to_string(inline => "<% if (0) { %>1<% } else { %>0<% } %>")'
11:31 salva joined #mojo
11:33 Jonis doesn't work on 3.97 mojo, so may be running an old version
11:34 sri lol
11:35 Jonis yeah :/
11:35 Jonis there's a plan in the works
11:40 Jonis matt_: <%= 0 ? '1' : '0' %>
11:45 kes sri: hi
11:47 kes I have proposal about rendering. Have you a minute to look at?
11:47 sri kes: sure
11:48 kes http://paste.scsys.co.uk/506889
11:48 dvinciguerra joined #mojo
11:48 matt_ I'm running 4.63 (ubuntu trusty package version)
11:48 sri didn't you post that here before?
11:48 sri i still don't understand it
11:48 sri matt_: that's ancient
11:49 matt_ sri: fair enough, I will upgrade
11:49 matt_ Jonis: that works, thanks
11:49 sri matt_: 6.54 is current :)
11:49 matt_ gosh, ubuntu is well behind
11:49 kes I have posted but I do not notice any answer. What is your question?
11:50 sri matt_: it's only 1174 lines ;p https://github.com/kraih/mojo/blob/master/Changes#L1174
11:50 matt_ :)
11:50 sri kes: i don't understand it, so i have no questions
11:51 sri kes: i'm afraid it's your responsibility to make sure you're understood
11:51 sri we've had this problem before, and you got passive aggressive with me
11:52 sri i'll gladly look at all proposals, but you have to make sure to present them in a way that people around here understand
11:54 kes Instead of direct manipulating '$content->{content} //= $output' better to use $c->content( $output )
11:56 kes at this line: https://github.com/kraih/mojo/blob/master/lib/Mojolicious/Renderer.pm#L133
12:00 kes sri: What this means "passive aggressive"? and how I should got?
12:02 sri kes: sorry, i don't have time, maybe you can find someone else to act as an intermediary
12:02 nic kes: Those do two different things
12:02 nic why is the second one better?
12:04 kes first worse because Renderer know internals of template. second just tells to get that new content
12:05 nic and what about when content already exists?
12:06 kes https://github.com/kraih/mojo/blob/master/lib/Mojolicious/Plugin/DefaultHelpers.pm#L63
12:07 kes same code is run. $content->{content} //= $output VS $hash->{$name} //= _block($content)
12:13 kaare joined #mojo
12:22 kes nic: --^
12:28 AndrewIsh Sorry, stupid question, can a Mojo template be embedded within another Mojo template?
12:29 sri http://mojolicious.org/perldoc/Mojolicious/Guides/Rendering#Include-templates
12:29 kes http://mojolicious.org/perldoc/Mojolicious/Guides/Rendering#Template-inheritance
12:29 jontaylor joined #mojo
12:32 nic kes: I think you're saying there are two ways of achieving the same result.  Which way were you able to figure out quicker?
12:35 kes yes, direct access is faster but the right way is to use method.
12:35 Farow joined #mojo
12:35 kes is not?
12:37 AndrewIsh sri / kes: Thanks very much, reading now
12:37 AndrewIsh joined #mojo
12:38 sri AndrewIsh: i'm rewriting and renaming that section currently :)
12:38 nic kes: I have never read about this thing called "the right way".  Do you have any links?
12:39 AndrewIsh sri: OK, thanks for the heads up ;-)
12:42 kes nic: maybe this one? https://en.wikipedia.org/wiki/Encapsulation_%28computer_programming%29
12:44 kes and yes. you may chose either way. Sorry, If 'the right way' the wrong combination of words
12:45 nic Last question from me: have you benchmarked the difference?
12:46 kes in any case the direct access is faster. But no, I do not benchmark the difference
12:47 asarch joined #mojo
12:48 mdom I need to run a single mojolicious::lite app with websocket support. Any good recommendation where to host it? It's just for playing so the service can be as low end as possible... :)
12:48 nic If you were leading this project and someone proposed "Let's make this slower by an unknown amount" I'm sure you'd have some questions of your own :)
12:49 mdom I'm looking at DigitalOcean in the moment, but maybe somebody has a good tip for Mojolicious... :)
12:49 * nic <3 CloudAtCost
12:49 good_news_everyon joined #mojo
12:49 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2pAn
12:49 good_news_everyon mojo/master 2b956ef Sebastian Riedel: better explanation for partial templates
12:49 good_news_everyon left #mojo
12:50 sri AndrewIsh: http://mojolicious.org/perldoc/Mojolicious/Guides/Rendering#Partial-templates
12:52 AndrewIsh sri: Thanks for that, looks good :) I'm just implementing it, but the template I'm including isn't appearing in the template that's including it. I've got the partial in the same directory as the main template, do I have to do anything else to ensure it can find it?
12:53 sri i'm sure the log has something to say
12:53 AndrewIsh sri: good point
12:59 good_news_everyon joined #mojo
12:59 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2ppL
12:59 good_news_everyon mojo/master 040dea4 Sebastian Riedel: use a directory for the example
12:59 good_news_everyon left #mojo
13:05 kes nic: ok. thank you for attention
13:15 kes I have just rename the path, but want the old path still exists and redirect to new. $page_route->get( '/old_path' )->redirect_to( 'new' ); May I achieve redirection without implementing controller?
13:20 anparker maybe like that? $r->get->(...)->to(cb => sub { shift->redirect_to(...) });
13:22 good_news_everyon joined #mojo
13:22 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hvC
13:22 good_news_everyon mojo/master c11b91d Sebastian Riedel: better explanation for content blocks
13:22 good_news_everyon left #mojo
13:22 sri http://mojolicious.org/perldoc/Mojolicious/Guides/Routing#Route-to-callback
13:26 kes sri, anparker: that is I am looking for. thank you.
13:51 neilhwatson joined #mojo
13:52 good_news_everyon joined #mojo
13:52 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hLB
13:52 good_news_everyon mojo/master d117929 Sebastian Riedel: the process method can handle all compilation steps
13:52 good_news_everyon left #mojo
13:56 good_news_everyon joined #mojo
13:56 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2htJ
13:56 good_news_everyon mojo/master d962856 Sebastian Riedel: fix return value
13:56 good_news_everyon left #mojo
13:57 ramortegui joined #mojo
14:37 denny joined #mojo
14:52 good_news_everyon joined #mojo
14:52 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hC2
14:52 good_news_everyon mojo/master cf16599 Sebastian Riedel: more consistent method names
14:52 good_news_everyon left #mojo
14:54 mcsnolte joined #mojo
15:01 good_news_everyon joined #mojo
15:01 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hlm
15:01 good_news_everyon mojo/master cb50ada Sebastian Riedel: link to Mojo::Template
15:01 good_news_everyon left #mojo
15:08 good_news_everyon joined #mojo
15:08 good_news_everyon [mojo] kraih tagged v6.54 at 6fe7d67: https://git.io/v2h4Z
15:08 good_news_everyon left #mojo
15:10 good_news_everyon joined #mojo
15:10 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2h4w
15:10 good_news_everyon mojo/master 2eb9d20 Sebastian Riedel: bump version
15:10 good_news_everyon left #mojo
15:16 good_news_everyon joined #mojo
15:16 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hRW
15:16 good_news_everyon mojo/master 33f2cdb Sebastian Riedel: that should have been one entry
15:16 good_news_everyon left #mojo
15:30 acamari joined #mojo
15:37 orev joined #mojo
16:01 zivester joined #mojo
16:12 lluad joined #mojo
16:35 jontaylor joined #mojo
16:43 kyshtynbai joined #mojo
16:43 gryphon joined #mojo
16:47 jontaylor joined #mojo
16:50 jontaylor joined #mojo
16:55 sri wonder if we should do something about this https://github.com/kraih/mojo/blob/master/lib/Mojolicious/Command.pm#L59
16:55 sri only supports args for now
16:56 jberger since that's in the Mojolicious:: namespace I don't see it as a general usability issue as much
16:57 jberger we could think about it, but I see it as a very different question
16:57 sri yea
16:57 sri it's not very important
16:58 jberger I'm curious if grep.cpan.me shows anyone else using that?
16:59 sri i'm sure there are users
16:59 sri all the generator commands
17:03 meshl joined #mojo
17:17 good_news_everyon joined #mojo
17:17 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2hhF
17:17 good_news_everyon mojo/master 8ca2611 Sebastian Riedel: more consistent examples
17:17 good_news_everyon left #mojo
17:56 dod joined #mojo
18:01 disputin joined #mojo
18:58 trone joined #mojo
19:04 disputin joined #mojo
19:19 acamari hi to all
19:20 jberger acamari: o/
19:20 acamari any ideas/comments about how to/why not replace Mojo::JSON with Mojo::JSON::XS globally?
19:22 PryMar56 joined #mojo
19:22 jberger Well that particular implementation is incredibly out of date
19:23 Grinnz_ and has serious issues and is based on JSON::XS
19:23 jberger Grinnz has a better one
19:23 jberger Take it away Grinnz_ :p
19:23 acamari JSON::XS is "bad"?
19:23 Grinnz_ https://metacpan.org/pod/Mojo::JSON::MaybeXS
19:24 acamari let me read
19:24 Grinnz_ https://metacpan.org/pod/Cpanel::JSON::XS#cPanel-fork are some reasons why Cpanel:: is better (which is what JSON::MaybeXS defaults to)
19:24 Grinnz_ you'll also note on the Mojo::JSON::MaybeXS docs, that Cpanel:: matches a lot better with Mojo::JSON
19:26 acamari \me reading
19:26 * acamari reading*
19:37 sri oh noes, don't deploy in dev mode! http://perlybook.org
19:37 Grinnz_ *facepalm*
19:37 sri hehe, version 3.95 deployed with psgi
19:41 jberger did we just DDOS it?
19:42 jberger or is the owner aware and fixing?
19:42 sri i can still open it
19:43 sri it's funny to see such old versions
19:43 sri the templates look so different now when you introspect them
19:45 acamari thanks Grinnz_, I'll try mojo::json::maybexs
19:49 jberger yeah I see it again
19:49 Adura How long do you think it'd take someone to write a Mojolicious book?
19:50 jberger I keep saying that if we write it collaboratively (an author per chapter/topic) it wouldn't take long at all
19:51 Adura Hmm, yeah...
19:52 acamari +1
20:00 csson acamari: note that the latest JSON::XS version pulls in Canary::Stability, and therefore does not support perl 5.22: https://metacpan.org/source/MLEHMANN/Canary-Stability-2006/Stability.pm#L106
20:02 disputin joined #mojo
20:04 Grinnz_ which doesn't mean it's broken, but yes, the other reason not to use JSON::XS is to avoid depending on mlehmann's whim
20:04 jberger yeah, I should move my FWBM to Cpanel too
20:06 jberger and if only my todo-stack weren't so deep I'd be trying to make a FFI::Platypus bindings to libuv again
20:07 jberger once uv bindings are as good as EV bindings then I'm MLEHMANN-free
20:07 sri +1
20:07 jberger actually I wonder if plicease works on commission?
20:08 jberger or maybe I can trick calid to do it :-P
20:29 meshl joined #mojo
20:37 dvinciguerra joined #mojo
20:58 jnbek joined #mojo
21:20 Zaid joined #mojo
21:22 Zaid Checking in for the first time folks
21:27 Zaid Quick question. Are there any ppm repositories that host Mojolicious plugins?
21:34 jberger Zaid: hi, not that I'm aware, but almost all Mojolicious plugins are pure perl so do you really need ppm?
21:36 Zaid @jberger Not really. I was actually trying to obtain ReplyTable :)
21:36 jberger ah well clearly you have discerning taste ;-)
21:36 Zaid I ended up downloading the PM file and pasting it in the relevant folder
21:37 jberger you can't use cpan(m)? what platform are you on?
21:37 Zaid Windows
21:37 jberger activestate?
21:37 jberger (I'm assuming because of ppm)
21:37 Zaid Yes. cpan cries a little
21:37 genio sri: You see this? https://blogs.microsoft.com/blog/2016/03/07/announcing-sql-server-on-linux/
21:38 Zaid It's not finding a C compiler
21:38 jberger Zaid: just be sure you have the appropriate renderer modules too
21:38 acamari csson: yeah, there seems to be no reason to use json::xs
21:38 jberger Zaid: you shouldn't need a C compiler
21:38 * CandyAngel feels that using CHI to store an integer in a file may be overkill
21:38 Grinnz_ cpanm != cpan
21:39 Grinnz_ i have no idea if the fatpacked cpanm script from https://cpanmin.us would "just work" in activeperl but it's worth a try
21:39 Grinnz_ the problem you might run into is even without the need for a C compiler, you might need make
21:40 Grinnz_ which you can install on windows, but I would sooner just use strawberry perl
21:40 Zaid Grinnz_: yes, that is what it complains about
21:40 genio you likely already have dmake
21:41 meredith i think more recent AS perl dists bundle cpanm
21:41 acamari :/ : It's a branch or fork of the official perl, by schmorp, to
21:41 acamari ***    improve stability and compatibility with existing modules.
21:41 meredith you can uh.. ppm install MinGW to get the "works like magic" c compiler, but they actually limit access to it
21:42 meredith if you are still on an older AS perl and haven't already installed the MinGW ppm, it gets locked behind a support contract paywall in order to install
21:42 Grinnz_ acamari: in other words, because he didn't like that his core hacks broke when people started improving perl again
21:42 Zaid meredith: trying it now
21:43 genio I'd suggest downloading Strawberry Perl of the same version as your activestate install and pretending the AS install no longer exists
21:44 meredith yeah if you aren't already tied to AS's perl dist, i'd prefer strawb.  pretty sure komodo and perl dev kit work with it just fine, too
21:44 meredith that said, _with_ the mingw package installed, AS perl has been way less painful to work with than it used to be
21:44 acamari Grinnz_: yeah, i guess so..
21:45 disputin joined #mojo
21:45 jberger Grinnz_: boy its really too bad that a perl build system needs make
21:46 jberger if only ... if only  ... there were some alternative
21:46 Grinnz_ hey, I agree with you on that part :P
21:47 Grinnz_ [ModuleBuildTiny::Fallback] is easy enough on any of my dists that require 5.10
21:50 Grinnz_ I almost feel like I should just [ModuleBuildTiny] my dists and anyone with outdated CPAN.pm will see something like "Can't locate Module/Build/Tiny.pm" and be smart enough to install that... but oh well
21:52 jberger shouldn't that end up in the configure requires?
21:53 genio I should really research the different build-a-module methods some day
21:54 Grinnz_ jberger: outdated CPAN.pm do not understand configure rqeuires, hence the whole issue
21:54 Grinnz_ the CPAN.pm built in to 5.10.1 does not, but it does have Module::Build in core at that juncture, so [ModuleBuildTiny::Fallback] alone is good enough for that
21:54 genio .oO( if every module on CPAN drops support for Perl < 5.10.1 )
21:55 Grinnz_ however, 5.8.x does not have either, so you also need [MakeMaker::Fallback], at which point I just use [MakeMaker] and forget about everything else
21:56 Grinnz_ if you require at least 5.12, you can depend on configure requires
21:57 Grinnz_ so obviously, someone just needs to invent time travel to fix 5.10
21:57 jberger but they can update CPAN.pm
21:57 jberger they just won't I guess
21:57 Grinnz_ you can't depend on them doing so
21:57 jberger I guess I'm glad that I mostly don't care
21:58 jberger \o/
21:58 jberger for my own modules I mean
21:58 jberger someone unfortunately does still have to care
21:58 Grinnz_ also, if you assume everyone will install your module with cpanm, you don't have any of these problems
21:58 jberger for a little while at least
22:51 cpan_mojo Statocles-0.073 by PREACTION https://metacpan.org/release/PREACTION/Statocles-0.073
23:00 acamari Grinnz_: you are Mojo::JSON::MaybeXS maintainer?
23:00 Grinnz_ yes
23:01 acamari I'm having some problem installing it... i'm gathering better info...
23:04 acamari sadly I also know nothing about build system of perl modules... :(
23:04 Grinnz_ what are you using to install it?
23:05 pink_mist and on what OS/distribution?
23:06 acamari i'm trying to isolate it, I'm porting it to OpenBSD/current - using the common procedure I used for lots of perl modules; maybe its' not related to Mojo::JSON::MaybeXS...
23:06 acamari my build system does: $ $ perl  Build destdir=/usr/ports/pobj/p5-Mojo-JSON-MaybeXS-0.010/fake-amd64 install
23:06 acamari Skip blib/lib/Mojo/JSON/MaybeXS.pm (unchanged)
23:06 acamari $ $
23:07 acamari I don't mean 'my', i mean OpenBSD/ports system~
23:07 Grinnz_ does the build system not know about Module::Build::Tiny? it requires --destdir
23:07 acamari s/destdir/--destdir/ ?
23:07 Grinnz_ right
23:08 Grinnz_ that --destdir=blah works on both MB and MBT
23:08 acamari well, its supposed to know - having an modbuild specific code, let me check it
23:08 Grinnz_ or is it a version of perl that doesn't understand configure requires as we were discussing before, so it's actually using MB?
23:09 acamari I don't know what MB means, let me google it
23:09 Grinnz_ also... if you're doing that with its dependencies, JSON::XS or Cpanel::JSON::XS are not going to work on openbsd if you compile them for amd64
23:09 jberger Module::Build
23:10 acamari oh
23:10 jberger there are essentially three build systems available for perl at the moment
23:10 jberger ExtUtils::MakeMaker (EUMM), Module::Build (MB) and Module::Build::Tiny (MBT)
23:10 Grinnz_ (what version of perl?)
23:11 acamari Grinnz_: json::Xs is already ported, cpanel::json::xs is not
23:11 acamari Grinnz_: 5.20.2
23:11 Grinnz_ oh, then thats fine
23:12 Grinnz_ i hope whatever you're using is following configure requires at least, or it will break on 5.22
23:13 acamari mm
23:14 Grinnz_ If you ended up falling back to Module::Build because it didn't follow configure requires, you'd see a giant warning in the log
23:16 acamari yeah, saw that before requireing Module::Build::Tiny
23:16 acamari any idea how to 'trace' debug Build
23:17 acamari apart from perl -d -> 't'
23:17 acamari s/destdir/--destdir/ does nothing, sorry for this poor report
23:17 Grinnz_ MBT doesn't have any debugging output really, but it takes --verbose
23:18 acamari anyway, $ cpan install Mojo::JSON::MaybeXS; works
23:18 acamari will check with OpenBSD/ports guys
23:19 acamari Grinnz_: thanks!
23:25 acamari jberger: thanks, I see, will research for into MBT
23:25 jberger acamari: its pretty bare-bones (as its name implies)
23:26 jberger note that App::ModuleBuildTiny is the "authoring tool" for it or else people use it in conjunction with Dist::Zilla (a generic authoring tool)
23:26 acamari or which one is the recommended?
23:26 jberger I'm starting to move my dists to MBT via App::MBT
23:26 Grinnz_ acamari: http://shadow.cat/blog/matt-s-trout/mstpan-11/
23:26 jberger at least the simple ones
23:27 jberger that article that Grinnz_ provided is useful, but keep in mind that it kinda conflates build and author tools a bit
23:28 jberger those three build tools are the only ones
23:28 jberger the others generate build tools for those three (if not used directly)
23:28 jberger MB and EUMM can act as their own authoring tool or else they can be generated from other authoring tools
23:28 Grinnz_ but note that other than MBT, they also conflate build and authoring themselves :P
23:29 acamari I see...
23:29 Grinnz_ MBT is strictly for building, dzil-based tools are strictly for authoring
23:29 jberger hmmm, I can already tell that I wish I had a better article to link to for this
23:29 * jberger ponders
23:30 jberger I wonder if MST would update his post
23:30 jberger otherwise I could write another one
23:30 pink_mist jberger: I'm sure he would if you poked him
23:30 Grinnz_ Module::Spilled::Tiny
23:30 pink_mist at least if you wrote the changes =)
23:31 jberger hmmm how did his initials get capitalized?
23:31 jberger too much MBT
23:32 acamari mod-install is deprecated?
23:32 jberger I love that there's a line in there just to keep me from yelling
23:32 jberger "This is still probably the least unpleasant way to do complex cross-platform compilation stuff where you need to customise the build process, but other than that there are better options."
23:32 jberger that's almost certainly aimed at me
23:32 jberger acamari: at least strongly discouraged yes
23:32 jberger it was the first attempt at an author tool
23:32 jberger I'd say all of the current author tools are better
23:33 Grinnz_ Module::Install is discouraged and has no active maintainers, Module::Build just has no active maintainers
23:33 Grinnz_ unfortunately as far as I know, neither of the docs for those modules say as such
23:34 jberger there are people who take tickets for MB when needed
23:34 jberger and EUMM is arguably worse
23:34 genio who owns Module::Install?
23:34 jberger though people realize that it has to be maintained or the house of cards collapses completely
23:34 acamari Grinnz_: hahah
23:34 Grinnz_ EUMM is actively worked on, but gets stuck when people break things
23:35 Grinnz_ genio: the last couple uploads were by ether, mostly dist cleanup stuff looks like
23:35 genio ah, it's maintained by the toolchain gang as well
23:35 jberger at least MB isn't tied to a network of platform-dependent ever-shifting external binary dependencies
23:35 jberger MB breaks when perl breaks it
23:36 genio It seems to me like the same core group of people maintain a HUGE portion of the modules on CPAN.  Spreading a bit too thin?
23:36 jberger genio: welcome to #toolchain
23:36 Grinnz_ genio: nobody else wants to do it
23:36 jberger speaking of which, all of this would probably be better suited to that channel anyway
23:37 Grinnz_ indeed
23:37 jberger acamari could we pursuade you to move this conversation over there
23:37 jberger ?
23:37 acamari yeah, sorry
23:39 genio I'm thankful that someone or some group are maintaining them, but if they're spread too thinly...
23:39 Adura Have them reproduce.
23:40 genio Excuse me, guys.  Can you create some clones of yourselves?
23:40 preaction they could just encourage contributors. that'd probably be a start
23:43 Grinnz_ it's tough. I'm sure they want to, but almost always contributors will have no regard for the impact of their changes (there was an extensive situation with EUMM somewhat recently)
23:43 Grinnz_ or, not enough regard
23:45 genio Yea, it's got to be hard to manage keeping things sane with people who want quick adoption of things
23:50 haarg yeah we tried to encourage contributors
23:50 haarg and ended up needing to do an emergency rollback
23:54 jberger its really hard to stay motivated when everything is basically mission critical and back compat is 100% required essentially forever
23:55 jberger thus the reason that the most excitement is in the newer tools
23:55 jberger and author tools
23:56 genio as well as modules in general
23:59 meshl joined #mojo

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary