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

IRC log for #mojo, 2016-06-01

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

All times shown according to UTC.

Time Nick Message
00:51 mcsnolte joined #mojo
00:52 tchaves joined #mojo
00:55 zivester joined #mojo
01:02 mattp joined #mojo
01:20 bpmedley https://gist.github.com/brianmed/844bf9fe70166d699ab8fec52be738ca <-- Here is my current under for saving response and request.. finish event for the win, it seems
01:36 jberger What is the loop for?
01:37 bpmedley Line 13-24?
01:38 bpmedley I guess that's the only loop.. that's in case the site gets popular.. :-P
01:38 jberger I don't see how it would help?
01:39 bpmedley I believe the mkdir's are atomic.
01:40 bpmedley So, the order of operations would be that an mkdir at time T5 for workers A and B would happen and one of the worker's mkdir would fail.
01:40 jberger Why not just redo if the mkdir failed?
01:41 bpmedley What if I get really popular and the 1 2 3 4 5 become a bottleneck..
01:41 bpmedley ;)
01:41 bpmedley redo would probably work well
01:41 jberger Then you probably shouldn't be logging in this way :p
01:41 bpmedley True, dat.
01:41 bpmedley I'm not sure what to use for logging at that scale.
01:41 jberger Optimize for that when that is a problem
01:42 bpmedley Sure, I hear ya.  I believe the foreach vs redo is a minor tribble at this juncture..
01:46 jberger You probably don't have to stash the name anyway
01:46 jberger You can just close over it right?
01:46 bpmedley That would be better; I was copying code around from a few methods.
01:46 jberger I can tell ;p
01:46 jberger Lots of repeated string interpolation etc
01:46 bpmedley I was excited that I got the res working - I didn't even do a double glance over the code.
01:46 jberger This might actually be a rare case for the after_build_tx hook
01:46 bpmedley Hrmm.  Dunno.
01:47 ilbot2 joined #mojo
01:47 Topic for #mojo is now 🍻 cheers | http://mojolicious.org | http://irclog.mojolicious.org | http://code-of-conduct.mojolicious.org
01:48 jberger That is what you say you are doing
01:49 bpmedley I don't see how an after_build_tx will work.  Doesn't that happen way before https://metacpan.org/source/SRI/Mojolicious-6.62/lib/Mojolicious/Controller.pm#L215 ?
01:49 jberger For establishing the on finish handler
01:51 bpmedley Hrmm; I would need an example.  Would the stash variable be needed in that scenario?
02:12 jberger I don't think a stash variable is needed in any case
02:42 noganex joined #mojo
03:47 disputin joined #mojo
03:53 dave if I use $c->delay where $c is a websocket controller, is there a way to override the default catch there at http://mojolicious.org/perldoc/Mojolicious/Plugin/DefaultHelpers#delay ?
03:56 bpmedley dave: Why doesn't passing a sub ref to catch work for you?
03:59 dave can I say $c->catch( sub { ... } )?
04:01 dave In the "longer version" it seems the catch method is attached to the delay object
04:01 bpmedley Are you wanting to provide a default catch for all your delay objects?
04:01 dave no just this one
04:02 bpmedley http://mojolicious.org/perldoc/Mojo/IOLoop/Delay <-- # Handle exceptions in all steps (see that e.g.)?
04:03 dave yes ... are you saying to use something like $c->delay( ... )->catch( sub { ... } )?
04:04 bpmedley That's how I read the documentation, yes.  Does that work for you?
04:04 dave I'll try it :)
04:06 dave actually, looking at the source ... that might not work
04:07 dave line 83 of Mojolicious/Plugin?DefaultHelpers.pm
04:07 dave he's already put a catch in there...if I do that after he does does mine win?
04:08 bpmedley dave: Mojo::IOLoop->delay( ... )->catch( sub { ... } ) <-- Perhaps this?
04:09 bpmedley dave: Answering you question, I would have to debug.  Perhaps you can try the Mojo::IOLoop->delay version?
04:09 dave yeah looks like I will have to do that
04:36 disputin joined #mojo
05:11 allison joined #mojo
05:12 bpmedley dave: Did you get catch working?
05:20 inokenty-w joined #mojo
05:32 dod joined #mojo
05:37 dod joined #mojo
05:38 che-quest joined #mojo
05:59 sri bpmedley, jberger: seriously, you can't do $c->tx->res->to_string
05:59 bpmedley sri: Ever?
06:00 sri yes
06:00 bpmedley My bad.. :(
06:00 bpmedley Gist is gone
06:01 sri if you do it before the response is written you risk destroying the response, if you do it after it is written then it might fail to generate the response again
06:01 sri this is why requests have a clone method that can fail
06:03 bpmedley Ahh, I didn't the bit about fail'ing to generate a response again.  Not sure I fully understand that part, either.
06:03 sri http://mojolicious.org/perldoc/Mojolicious/Guides/Rendering#Streaming
06:04 sri that response can only be generated once
06:04 bpmedley Cool, thx for the quick reply!
06:24 sri so, to make your approach somewhat reliable you'd need a response version of this https://github.com/kraih/mojo/blob/master/lib/Mojo/UserAgent/Transactor.pm#L85
06:26 bpmedley Understood, saving the req is more important for me.  I can wrap my response generators and save them via the app..
06:36 dave bpmedley: I think so. Hard to tell yet until I force an error in testing, but I am not using the ->delay helper now
07:06 cosimo joined #mojo
07:08 cosimo_ joined #mojo
07:12 coolo batman: I fail to find where the assets are stored in 1.13 so I could copy them to the cdn (or to the package in my case)
07:13 batman coolo: "git status" should tell you.
07:14 batman coolo: let me see if i can find a better answer to you...
07:14 coolo all I'm seeing is assets/cache but that have different filenames than what the static assets would look like
07:15 coolo so if I add route->to as described in your cookbook I end up with a link that goes like .../my-assets/asset/$HASH/app.css
07:15 batman not sure if i follow... maybe you have to run with MOJO_MODE=production to see the assets you don't see..?
07:16 coolo but I see nowhere below my app any directory with $HASH
07:17 coolo so I have /asset/3000aae959/bootstrap.css - but I have no such file anywhere I could package
07:18 coolo or I'm blind
07:18 batman no, that is true.
07:18 batman if you do, it's just a coincident.
07:18 coolo so how would I copy that to a cdn?
07:19 batman i don't copy anything to cloudflare (the cdn i'musing). they just cache the assets on the first request
07:19 batman *i'm using
07:20 batman maybe you want to read through this pr: https://github.com/jhthorsen/mojolicious-plugin-assetpack/pull/81
07:20 coolo ok, but that doesn't work for me. I need *static* assets to package them
07:20 batman what do you mean about "package them" ?
07:22 coolo we deploy from distribution packages
07:23 coolo so we have one (build) server to create the assets and bundle it with the application and other (production) servers that will install this package - those don't have sass and such
07:23 coolo and that worked fine with the previous version - we just packaged public/packed and we had it done
07:24 coolo you might remember that you had to fix one problem for us where assetpack tried to recreate the assets - but our production server user can't even write to the app directory outside of logs and database
07:25 batman but the same thing happens in this version. it's just that the files on disk doesn't change names all the time like they used to
07:27 batman so the checksum for the file on disk is composed by the input file name instead of the content of the file.
07:27 batman "input file name" = "input url or path on disk"
07:27 batman it was really annoying that assetpack created a gazillion assets on disk while developing.
07:28 batman but the checksum you see in the url in your browser is the checksum of the _content_ of the file.
07:28 coolo oh, I can see that. All I'm missing is really an easy dump for production so I don't require the app for serving static assets
07:28 trone joined #mojo
07:29 batman yeah, just make a unit test that sets MOJO_MODE=production or start morbo with MOJO_MODE=production while developing
07:29 batman there's never been a way for assetpack to generate both the production and development versions of the assets.
07:29 coolo I guess I could create something similiar to what you have with your cdn and ask the app during building about its assets and copy them away into the path
07:29 batman unless you go back to version 0.40 or something, which was full of bugs :(
07:30 coolo batman: that's what https://github.com/os-autoinst/openQA/blob/master/script/generate-packed-assets does
07:30 batman coolo: yes, you can. you can access the assets in startup{} with https://metacpan.org/pod/Mojolicious::Plugin::AssetPack#processed and do "whatever"
07:31 coolo I'll try - thanks
07:31 batman ok, then just run that script before packaging the app, and you will have all the assets you need.
07:32 batman i think maybe you want the same thing as the PR i linked to above.
07:32 coolo right
07:32 batman if you read through it and comment, you might convince me to reopen it.
07:32 Vandal joined #mojo
07:32 batman i do believe that the right way is to make the cdn cache the asset, since it's so lot simpler to set up.
07:33 batman but maybe my setup is just too simple...
07:33 coolo yeah, I comment just to get CCed on new events ;)
07:33 batman i think that PR is pretty much dead :/
07:33 coolo batman: your setup is basically impossibly by our policy
07:33 coolo but another topic - related to assetpack though
07:34 batman but... if you have a good usecase and description of the problem, please make a new issue and we'll make it work.
07:34 batman sounds like you're not alone. my world view is just too simple to get it :(
07:35 coolo I customize bootstrap through sass variables - and I failed to find a good way to make this work in the way you describe in the deprecation note of bootstrap3 plugin
07:36 batman yeah.... i need to fix https://metacpan.org/pod/distribution/Mojolicious-Plugin-AssetPack/lib/Mojolicious/Plugin/AssetPack/Guides/Cookbook.pod#Bootstrap
07:36 batman let me give you an example...
07:36 coolo batman: I guess you're having N developers vs 1 web page. What we're facing is having N developers vs M web pages all using the same app. And we want to avoid all M web pages to build assets themselves
07:37 coolo but I document in an issue
07:38 cpan_mojo Mojolicious-Plugin-QuickPg-1.5 by AZJADFTRE https://metacpan.org/release/AZJADFTRE/Mojolicious-Plugin-QuickPg-1.5
07:39 batman coolo: https://ssl.thorsen.pm/paste/22b5f36b6c76
07:40 batman i don't see the difference. just serve the assets from the same domain and the cdn should cache it just fine.
07:40 andrew_ joined #mojo
07:47 coolo batman: I would need to try your proposal for real. Possibly the problem I created myself but having one sass file that is basically doing what your .def file does - but then the sass interpreter wouldn't download the deps
07:48 batman are you sure? :)
07:49 coolo batman: wait a second - I'm still with the previous problem :)
07:49 stephan48 g44
07:49 coolo the more I describe the problem, the more I realize I don't care enough about the assets having the same path as long as it's not recreating them
07:49 coolo and that I have to try first
07:50 * coolo will save his nicely written preamble somewhere though ;)
07:53 batman coolo: sure...
07:53 batman sorry, but i'm very tired to day... i will hunt down some coffee now
08:04 batman coolo: let me know if i can be of more help
08:05 batman sorry = sorry if i'm not very constructive. i'll try to be better :)
08:06 coolo batman: I'm trying harder to accept that the assets are now served by the app - and let you know :)
08:06 batman also: "are you sure?" means: it should actually work, if you add a complete URL to the @import "..."; statements
08:06 coolo my beloved https://static.opensuse.org/ already died the rails dead - rails 4 asset pipeline also didn't like me :)
08:07 coolo batman: oh that I'm sure of - the @import ended up in the generated css
08:07 batman hehe... well, my statement still stands: make a killer argument, and i will 1) take a PR 2) implement it
08:07 batman it's not that i don't want it. i just don't see why it can't be solved in an easier way
08:18 batman coolo: does this make sense? https://github.com/jhthorsen/mojolicious-plugin-assetpack/blob/master/lib/Mojolicious/Plugin/AssetPack/Guides/Cookbook.pod#bootstrap
08:19 jasanj_ joined #mojo
08:20 jasanj_ I often do this: my $site = $c->param('site') // 'us'; $c->param(site => $site);
08:20 jasanj_ cause I need reuse the 'site' parameter in template
08:21 jasanj_ is there any good way to do this ?
08:22 sri oh wow, didn't realize the surface book even has a magnetic power connector
08:23 sri microsoft really went all out with that one
08:27 coolo batman: it looks easy to follow - I missed assetpack.def completely before actually
08:29 osfabibisi joined #mojo
08:31 batman ah! right you might want to look at the complete tutorial, but here is about assetpack.def: https://metacpan.org/pod/distribution/Mojolicious-Plugin-AssetPack/lib/Mojolicious/Plugin/AssetPack/Guides/Tutorial.pod#Defining-assets-in-a-definition-file
08:32 batman [mode==development] is quite handy. i use it when i include vuejs, since the development version gives me a lot more debug info in the browser console
08:32 * coolo only looked at it, but considered definining it in the application good enough and skipped the next paragraph :)
08:32 batman hehe
08:33 batman about the tutorial changes: i haven't actually run example myself, so i would very much appreciate questions and feedback on the example before i make a new release
08:33 batman "run the example"
08:34 coolo it's on my todo
08:34 batman thanks! i'll go back to work, but will be back when you mention me.
08:58 coolo SUCCESS! I have packages :)
08:58 coolo next step: testing ;)
09:27 batman coolo: awesomeness! so did you have to change anything or did you just adjust to the new way things are working?
09:28 Sound joined #mojo
09:29 coolo batman: before I brag about what I've done, i better finish my tests :)
09:30 coolo https://github.com/os-autoinst/openQA/pull/700/files isn't exactly a tiny change - but it's mainly about that script I showed you earlier
09:31 coolo batman: but what I feared of happens
09:31 coolo http://paste.opensuse.org/66026858 is from our staging server
09:32 coolo and /usr/share/openqa/assets/cache/needlediff-d5af99e8a7.min.js exists
09:32 coolo but the process call wants to double check it's the same thing - but I need to call process for %= asset 'needlediff' to work
09:34 coolo it's a bit recursive problem
09:39 batman coolo: i need to fix something @work. bbl
09:39 * coolo finishes the issues story and we pick it up from there .)
09:40 Sound Hello!  I'm stuck in a deployment issue with Mojo + nginx. It looks like the backend randomly drops the connection while serving response to nginx. But nothing is written in the Mojo log, and I have no clue about what to do.  How do I troubleshoot this?
09:44 che-quest drops on diffrent routes?
09:44 dod joined #mojo
09:45 che-quest i am normal use nginx+uwsgi+mojo
09:47 Sound che-quest: mostly on a couple routes that handle a form POST submission
09:48 punter joined #mojo
09:50 che-quest I'm trying to remember something about nginx POSTs, something that was long ago
10:17 che-quest Sound: which status code of the POST requests?
10:19 coolo batman: CSS::Minifier::XS is buggy. I thought I did something wrong, but I checked and it's the same on our public instance (using the old stack): only parts of the css file is minified
10:19 coolo it's only more obvious now as the first part is unminified
10:20 coolo and https://openqa.opensuse.org/packed/bootstrap-fa602e94cf94f5a7381026cbf851235c.min.css looks minified on first look, but if you scroll down like 30% tons of whitespace appears
10:21 coolo but I can't find anything more recent on cpan ;(
10:28 Kripton joined #mojo
10:40 batman coolo: are you sure you're including assetpack.db in the package?
10:41 batman coolo: can't help you with CSS::Minifier::XS. not my code :(
10:41 coolo batman: I forgot to prefix it with FYI
10:41 coolo but I have 200K with perl and 180K with cssminifier.com, which seems to use yui
10:42 coolo so there is potential :)
10:43 coolo batman: but yep, I have /usr/share/openqa/assets/assetpack.db packaged
10:44 batman pretty sure there's something messed up with $app->home or $app->asset->store->paths
10:45 meshl joined #mojo
10:46 batman Could not find asset 'javascripts/needleedit.js' is not from the generated asset. it's from failing to find an input file.
10:48 tchaves joined #mojo
10:51 batman maybe i should just use CSS::Sass to minify instead of CSS::Minifier::XS if available...
10:52 dvinciguerra joined #mojo
11:15 coolo batman: I don't get your last comment - https://metacpan.org/pod/Mojolicious::Plugin::AssetPack has the same order of pipes. I just added Javascript to it
11:22 batman coolo: that could be a documentation error. not quite sure
11:23 batman btw: my comment could also be wrong
11:23 marcus me fail english? Unpossible
11:24 batman marcus: was my sentence messed up or just making a joke?
11:33 che-quest Why Mojolicious::Routes does not allowed l10n action "잘" ?
11:34 sri che-quest: could be a bug
11:36 sri file a report if you want it fixed
11:36 che-quest sri: i want, ok
11:39 Zen is it possible to configure hypnotoad to use TLS-PSK ?
11:40 coolo batman: damn it, now you make it sound very appealing to consider the cache in git
11:40 marcus batman: just making a joke.
11:41 batman marcus: okidoki :)
11:42 batman coolo: that's a solution i support and can take bug reports on as well. your setup is unknown to me. and i really don't like submodules, so i won't try it out myself.
11:43 coolo yeah. it replaces git submodules with having two forks (development+production) of every asset in git
11:47 batman sorry. i don't follow. not sure if need to. i don't think your git workflow is a concern of assetpack...
11:50 dantti_laptop joined #mojo
11:52 coolo batman: sure, it's just me suffering in the balancing act between your ideals and the ideals of my coworkers ;)
11:53 batman i'm batman. resistance is futile.
11:53 batman :)
11:53 marcus batman of borg?
11:53 batman coolo: i'm not arguing that my solution is correct. i just believe it's simpler than what you have now.
11:54 jberger marcus++
11:54 * jberger yawns
11:54 marcus fail fast and simply?
11:54 batman coolo: i also realise that my knowledge of your setup is lacking, so i could be very wrong :/
11:54 batman marcus :)
12:16 jasanj_ joined #mojo
12:30 che-quest posted l10n action name report https://github.com/kraih/mojo/issues/966
12:31 dvinciguerra_ joined #mojo
12:34 coolo batman: just to avoid you getting bored - where are the actual fonts coming from if I use fontawesome.css from cache ?
12:35 coolo ah, missing pipe
12:35 zivester joined #mojo
12:40 punter joined #mojo
12:41 Zen is it possible to configure hypnotoad to use TLS-PSK ?
12:42 sri Zen: odds are nobody knows
12:42 Zen sri: I was afraid of this :)
12:44 zivester joined #mojo
12:54 ramortegui joined #mojo
13:21 gizmomathboy joined #mojo
13:38 sri looks like this might be controversial https://github.com/kraih/mojo/issues/966
13:41 jberger Zen: I'd imagine if IO::Socket::SSL supports it that there would at least be some way to make it happen
13:44 sri oh, did i mention what ups did now?
13:45 sri found a second note in the mailbox this morning that said they tried to deliver a second time yesterday evening -.-
13:48 sri and just as i write this the ups guy is at the door
13:51 Vandal joined #mojo
13:52 jberger \o/
13:55 genio got your mac back?
14:03 coolo joined #mojo
14:09 jabberwok http://blog.wlindley.com/2016/05/31/handy-tricks-with-sql-abstract-more-and-mojolicious/   getting back to the blog
14:12 Bean joined #mojo
14:26 jberger sri: I just filled in some of the facts since it will help me if the discussion goes into reasoning now at least I know where the code is
14:40 sri genio: they sent a new one
14:40 cosimo joined #mojo
14:41 sri so nice to look at a retina screen again
14:51 zivester joined #mojo
15:00 jberger so sri, I'm curious why action methods have to be /[a-z]/
15:01 jberger I understand checking for leading underscore and explicitly hidden methods
15:01 jberger (and note that I'm asking broadly, not just in the context of the question at hand)
15:08 asarch joined #mojo
15:22 PryMar56 joined #mojo
15:22 lluad joined #mojo
15:34 sri jberger: you tell me
15:37 jberger trying to avoid delegating to perl builtins like say AUTOLOAD?
15:37 bpmedley jabberwok++ # would you consider adding a complete example, or did i miss that?
15:38 sri jberger: seriously, i don't know where the regex came from
15:38 kaare_ joined #mojo
15:39 jberger ah
15:41 batman coolo: i've unsubscribed to https://github.com/os-autoinst/openQA/pull/700. Please mention me if you want input.
15:47 jasanj_ joined #mojo
15:48 jberger sri: looks like I was mostly right in my guess: https://github.com/kraih/mojo/commit/5a392932d5129167848e53f3f1339cf7cfd1153c
15:51 Grinnz_ || $method =~ /^[A-Z]+$/ # maybe?
15:54 Grinnz_ add a _ in there if it's also meant to catch TO_JSON and such
16:05 dod joined #mojo
16:08 jberger Grinnz_: I think you are right, probably just want to exclude methods that are all caps rather than exclude methods that contain any non-lowercase
16:08 jberger non-lowercase-ascii even
16:21 jasanj_ joined #mojo
16:23 dvinciguerra joined #mojo
16:59 mpapec return !!(
16:59 mpapec $h->{$method} or index($method, '_') ==0 or $method eq uc($method) or $method !~ /[a-zA-Z]/
16:59 mpapec );
17:11 mpapec code style nitpicking,
17:11 mpapec https://github.com/kraih/mojo/blob/5a392932d5129167848e53f3f1339cf7cfd1153c/lib/Mojolicious/Routes.pm#L79
17:12 mpapec =>my $route = $self->find($name) or return;
17:15 mpapec and return undef makes sense only in list context when returned list should have one list element
17:15 mpapec (undef)
17:15 sri i'm waiting for the argument
17:15 sri about bare return you're wrong though
17:16 mpapec in scalar context return or return undef are same things
17:16 sri bare return is an attack vector, so we don't do it
17:17 sri i know how it works
17:17 sri very well in fact
17:17 mpapec can you point me to some docs?
17:17 Grinnz_ https://metacpan.org/pod/Perl::Critic::Policy::Freenode::EmptyReturn
17:18 sri heh
17:18 mpapec Grinnz_: tnx
17:19 sri you should start with the assumption that we know what we are doing before quoting PBP ;p
17:19 mpapec actually this link confirms what I was talking about
17:19 Grinnz_ no, it doesn't
17:19 Grinnz_ if a function intends to return a scalar, all returns should explicitly return a scalar
17:20 Grinnz_ otherwise it is used like in the example there, and you get exploits
17:21 mpapec ok, but there is no exploit mentioned on your link
17:21 sri some context https://groups.google.com/d/msg/mojolicious/aJTYjRCPjOE/QJZB6ORfYJAJ
17:21 Grinnz_ there was a big one with bugzilla
17:21 sri not an exploit, an attack vector
17:23 mpapec ah, wantarray related?
17:24 sri my $foo = $c->param('foo'); $c-whatever(foo => $c->param('foo'))
17:24 preaction no. hash construction related: my %user = ( name => $c->param( 'name' ) ); $c->save_user( \%user ); <- if I have ?name=myname&name=is_admin&name=1, %user will contain is_admin => 1
17:24 sri context, not wantarray
17:24 mpapec yes, ok
17:25 Grinnz_ my policy bundle has a similar policy about wantarray for the same reason though https://metacpan.org/pod/Perl::Critic::Policy::Freenode::Wantarray
17:26 Grinnz_ it's the same conceptual problem, wantarray and bare return are both context sensitive
17:27 mpapec $c-whatever(foo => scalar $c->param('foo')) helps :)
17:27 Grinnz_ yes but that relies on the user knowing about the problem
17:27 preaction but nobody ever remembers to do that because most of the time it works
17:27 mpapec yes, ofcourse
17:27 sri the point here is that you should never have mixed context return values
17:28 preaction also, i'm not typing "scalar" 12 times while building a hash. context is about DWIM. forcing context isn't DWIM at all
17:28 sri stick to list or scalar
17:28 mpapec makes sense
17:30 sri for the record, damian now also teaches "return undef;"
17:30 sri so PBP is outdated on that
17:30 mpapec PBP book?
17:31 jberger PBP also recommended inside out objects
17:31 jberger but no one seems to remember that bit
17:31 preaction i remember it, vividly
17:31 Grinnz_ part of the problem is that the default perlcritic policies use PBP verbatim, even though PBP was intended more as guidelines than rules
17:31 preaction i spent years dealing with Class::InsideOut in WebGUI...
17:31 mpapec OO perl from 2003 is even better ;)
17:32 mpapec but that are not best practices
17:34 mpapec btw, had Damian any recent perl talks?
17:59 inokenty joined #mojo
18:04 haarg Class::InsideOut was at least a working implementation of inside-out objects.  what PBP recommended was just broken.
18:09 meshl joined #mojo
18:22 sri interesting, the replacement macbook thermal throttles even less than the last one Oo
18:52 mrallen1 joined #mojo
19:01 coolo batman: I merged it now. I think I found a good balance and only real life can prove me wrong ;)
19:07 jabberwok what is this Real Life of which you speak?
19:07 coolo jabberwok: no worries, only rumours
19:08 jabberwok =whew=
19:08 coolo damn
19:08 coolo Odd number of elements in anonymous hash at
19:08 coolo /usr/lib/perl5/vendor_perl/5.22.1/Mojolicious/Plugin/TagHelpers.pm line 171 (#1)
19:08 coolo now on to find which template where :)
19:09 preaction coolo: Devel::Confess might help
19:13 coolo preaction: it worked. the backtrace is pretty useless for templates, but it shows the arguments
19:14 coolo => (data => (remote => 'true')) => (id => 'restart-result')
19:14 coolo it looks almost correct
19:33 marcus https://medium.com/@qutheory/server-side-swift-vs-the-other-guys-1-input-ec48d7be37b7#.gfkouwqz5
19:49 punter joined #mojo
19:49 good_news_everyon joined #mojo
19:49 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vrAEk
19:49 good_news_everyon mojo/master 88cbb7c Sebastian Riedel: improve generator commands not to overwrite existing files (closes #965)
19:49 good_news_everyon left #mojo
19:51 good_news_everyon joined #mojo
19:51 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vrAEc
19:51 good_news_everyon mojo/master 8c0325d Sebastian Riedel: chdir a little later
19:51 good_news_everyon left #mojo
20:08 orev joined #mojo
20:15 good_news_everyon joined #mojo
20:15 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vrA2w
20:15 good_news_everyon mojo/master d3e6903 Sebastian Riedel: remove deprecated code
20:15 good_news_everyon left #mojo
20:20 sri allright, that means mojolicious is now officially 2 lines smaller than catalyst core :)
20:21 sri 8488 vs 8490
20:23 anparker is it some kind of contest?)
20:23 sri yes
20:24 sri it's the culmination of an old rivalry where catalyst devs used to complain that mojolicious reinvents too much
20:25 sri not thsat anyone cares anymore, but i get a kick out of it ;p
20:26 ribasushi wake me up when a project written against mojolicious in 2010 can run on latest-mojo-from-cpan with little to no changes
20:27 sri so, you'll be silent until that happens?
20:27 ribasushi no I will be visibly unexcited
20:27 sri meh
20:34 jberger sri++ nice patch for #965
20:36 jberger ribasushi: we are aware of your views but our policy is clearly stated, that it keeps you from using Mojolicious is unfortunate
20:37 jberger s/,/;/
20:37 jabberwok wake me up when my Visual Basic 3 programs using the Win16 API work on Windows 10
20:37 ribasushi I was pointing this out in the context of "omg Catalyst has more code"
20:37 sri wish i could think of a more snarky response, but today i'm too happy about having a decent laptop again ;p
20:37 ribasushi in any case - I said what I had to say, carry on
20:38 jberger sri: the new lappy is working out for you then?
20:39 sri yea, i'm suspicious of the spacebar though, it's almost a little bit too clicky --
20:39 sri -.-
20:50 sri i really need faster external storage for time machine backups though
20:52 Grinnz_ I think ribasushi was just trying to say Catalyst has more code because of backcompat, which IMO is a valid reason for a different userbase
20:54 Grinnz_ I wish laptops were more user serviceable, too many times I have a laptop that's dead for some simple reason that in a desktop could be easily replaced
20:55 sri i tend not to keep laptops long enough for that to happen
20:56 Grinnz_ that's one way :P
20:56 marty joined #mojo
21:09 sri guess i'll not do anything about https://github.com/kraih/mojo/issues/966
21:09 sri if a patch pops up we'll let a vote decide
21:22 bpmedley sri: did you get new stickers?
21:23 jabberwok http://blog.wlindley.com/2016/05/31/handy-tricks-with-sql-abstract-more-and-mojolicious/   including complete example, thanks bpmedley
21:24 bpmedley jabberwok: wow, it's even practical.. :)
21:26 jabberwok especially if you happen to need to know how many square miles of land and water are in the towns of Bernie Sanders' home state!
21:30 jabberwok good thing i reviewed it, just in case someone put &order_by=(drop tables all)
21:31 cfedde joined #mojo
21:37 disputin joined #mojo
22:16 PryMar56 joined #mojo
23:44 Phil21 joined #mojo

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