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

IRC log for #mojo, 2015-12-10

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

All times shown according to UTC.

Time Nick Message
00:04 henq joined #mojo
00:17 voldemortensen joined #mojo
00:20 voldemortensen1 joined #mojo
00:25 jlzhang joined #mojo
00:36 vanHoesel joined #mojo
01:33 jlzhang joined #mojo
01:40 kaare joined #mojo
02:47 voldemortensen joined #mojo
03:31 kafkaesque joined #mojo
03:32 kafkaesque with mojo::ua, what is the equivalent of curl URL -H 'Content-type:application/json' -d '[{id:test}]'
03:33 kafkaesque I'm doing my $blah = [{id => test}]; $blah = encode_json $blah; $ua->get(url => json => $blah); and it doesn't seem to work
03:33 kafkaesque encode_json coming from Mojo::JSON
03:37 kafkaesque woops, meant $ua->post
03:39 Grinnz the json generator does the encode part for you. just pass the data structure
03:41 kafkaesque There we go, that was my problem. Thanks Grinnz :)
03:59 noganex_ joined #mojo
04:23 inokenty-w joined #mojo
04:57 jlzhang joined #mojo
05:02 kafkaesque one more question: How do I set a cookie for my requests using Mojo::UA?
05:02 kafkaesque I've been doing it by setting the cookie header, but I'm assuming there's a righter way to do it.
05:06 preaction use the cookie jar
05:45 melo joined #mojo
06:18 sugar_ joined #mojo
06:19 tempire- joined #mojo
06:26 jberger Have shower, will sleep
06:34 tempire joined #mojo
06:48 dod joined #mojo
06:52 dod joined #mojo
07:06 melo1 joined #mojo
07:07 dod joined #mojo
08:03 osfabibisi joined #mojo
08:13 Vandal joined #mojo
08:16 AndrewIsh joined #mojo
08:19 melo joined #mojo
08:20 eseyman joined #mojo
08:29 nic Congratulations jberger.  My shower has a mystery leak, but I reckon using Fail-fast Agile I can work around it with minimal impact to development
08:30 trone joined #mojo
08:40 henq joined #mojo
08:45 melo joined #mojo
09:03 D4RK-PH0ENiX joined #mojo
09:04 d4rkie_ joined #mojo
09:08 kes joined #mojo
09:28 absolut_todd joined #mojo
09:30 arthas joined #mojo
09:57 punter joined #mojo
09:58 Adurah joined #mojo
10:01 vytas joined #mojo
10:02 Onigiri joined #mojo
10:15 Zoffix tun tun tun... based on this ticket Mojolicious 6.36 breaks my module, even tho it works with 6.35 https://github.com/zoffixznet/WebService-Rollbar-Notifier/issues/5
10:15 Zoffix JSON stuff hm
10:17 sugar_ joined #mojo
10:17 Zoffix "6.36  2015-12-08  - Improved Mojo::JSON performance slightly. (haarg)" :D
10:26 sugar joined #mojo
10:26 Zoffix This reproduces it: perl -MMojo::JSON=j -wlE 'say j {foo => $^V}'
10:26 Zoffix Though I doubt Mojolicious should care about this case.
10:26 Zoffix no point
10:29 scoopio joined #mojo
10:34 sue joined #mojo
10:58 haarg fucking version objects
11:00 cpan_mojo WebService-Rollbar-Notifier-1.002001 by ZOFFIX https://metacpan.org/release/ZOFFIX/WebService-Rollbar-Notifier-1.002001
11:10 absolut_todd joined #mojo
11:28 kaare joined #mojo
11:53 absolut_todd joined #mojo
12:42 irqq joined #mojo
12:43 inokenty-w joined #mojo
12:49 jberger nic: just be careful, all this started as fixing a leak almost a week ago
12:50 nic I'm already in big trouble
12:50 nic I started dismantling it piece by piece, on the assumption the problem would be something I had done
12:51 nic but by slow elimination it appears the problem is with the pipework _under_ the shower base (put in by the previous owner)
12:52 nic The clearly had a sense of humour, because they installed the base with zero spare clearance for the pipes, then tiled above it
12:52 jberger Hmmm yeah that sounds bad
12:52 nic So the only way to fix/replace the pipework is to angle-grind away the heavy-duty wall-tiles so that I can remove the base
12:52 jberger Do you have access from through the back?
12:53 jberger ie going through the wall in the next room over?
12:53 nic It'll create massive amounts of difficult-to-clean mess and I'll no doubt end up the with non-matching tiles in their place (the whole room is tiled in the same posh tiles)
12:54 nic actually, you may have a point.  The only other way to access is by cutting through the kitchen ceiling
12:54 nic but seeing as the leak has ruined the kitchen ceiling, there's little to lose
12:54 jberger Yeah you're going to have to replace that anyway
12:54 nic See, discussing on #mojo can help solve the most intractible of problems :D
12:55 jberger \o/
12:55 jberger Still sounds like a pain to work from that angle
12:55 jberger Good luck!
12:55 nic jberger++
13:23 orev joined #mojo
13:29 ashimema anyone here using Toadfarm.. I'm having issues attempting to use the run_as feature
13:36 ashimema ok.. you can completely ignore that..
13:36 ashimema it seems to have magicall started working now
13:36 ashimema not sure what I did.. must have been a typo somewhere.
13:41 nic Aaaggh
13:41 nic ->url_with->query(a => 'Aaaaggh') loses the original query
13:42 nic so I guess I need to use something like ->url_for->query(%existing_query, a => 'Aaaggh')
13:43 nic it would be mighty handy to have ->add_query(a => 'Aaaggh')
13:46 nic or do I just need to use 'merge'?  hadn't noticed that before
13:47 nic bah, penny has finally dropped; just need to get me some of them there square brackets
13:48 nic yay,  ->url_with->query([a => 'Aaaggh']) does the job beautifully
13:52 ashimema Toadfarm related.. anyone know if SNI is ever likely to land in Mojo::Server::Daemon?
13:53 ashimema I think batman pointed to it not being supported by ->listen.. but I'm not experienced enough to understand where the limitation comes from..
13:54 ashimema I'm likely being thick.. but it looks like listen jsut takes an array of locations.. which can include a ca and cert..
13:56 ashimema though looking at the examples.. I'm not seeing 'names' anywhere.. only * or ip's..
13:56 ashimema so parhaps that's where I'm missing the plot
13:57 ashimema or is it just generally better practice to stick apache/nginx in front and let it deal with the ssl side of things before adding proxy request headers
13:59 jjohn ashimema that's what I do.
14:00 jjohn I really like having SSL offloaded from the app server.
14:00 ashimema I may eventually need to do that anyway to support shibboleth authentication..
14:00 jjohn But I think AWS and other load balancers have influenced my thinking.
14:00 ashimema unless of course I want to write a saml2 parser and get to grips with shibboleth on yet another level
14:00 jjohn oh boy.  shib.
14:00 jjohn I used apache for shib stuff
14:01 ashimema me to so far..
14:01 jjohn it was always painful.
14:01 ashimema though I find the shibboleth service provider software super bloated and unreliable
14:01 nic ashimema: also, nginx says it is particularly optimised as an ssl terminator, so it makes a lot of sense
14:01 ashimema anyone ever used shib with nginx..
14:01 ashimema I kinda like the idea of nginx over apache..
14:02 ashimema but as I may need apache for mod_shib anyways I might just stick with it.
14:02 jjohn shib is such a nightmare that I think you're better off with apache.
14:02 jjohn I also use apache as the front end proxy.
14:03 jjohn But nginx definitely will give you more performance for static files.
14:03 jjohn Your app is likely to be the bottleneck with apache.
14:03 jjohn Benchmarks will, I think, support this.
14:04 jjohn Premature optimization is evil. :-D
14:04 CandyAngel Hm.. if I have a helper which parses a file (and either returns an array or runs a $cb on each row), is it better to pass the file to the helper or have the helper deal with the file itself?
14:04 jjohn How centralized does that information need to be, CandyAngel?
14:05 d4rkie joined #mojo
14:05 jjohn If other parts of the app might need to know about that file, perhaps your main app class should pass the filename to the plugin
14:06 jjohn If the file is in a "well-known" location, perhaps controllers, et. al. should ask the helper for that info.
14:06 D4RK-PH0ENiX joined #mojo
14:07 d4rkie_ joined #mojo
14:10 CandyAngel Currently, only that helper needs to know the name, nothing else uses it, but I might need to add an "age check" so the command quits if the file is too old (unless it is --forced)
14:10 CandyAngel So maybe I need another helper which does the check?
14:12 jjohn I wouldn't do that.
14:12 jjohn I would make the helper the gatekeeper of that information.
14:13 jjohn Otherwise, you have two plugins that might need to know too much about each others business.
14:13 jjohn But, the code is yours. :-D
14:13 CandyAngel I'm always up for writing better code :P
14:14 jjohn From the wonderful Code Complete comes the wonderful advice: Keep routines loosely coupled and highly coherent.
14:15 nic And don't store your Christmas pies near a radiator.
14:15 jjohn Or your gasoline.
14:16 CandyAngel Hm, so the helper would check the age of the file?
14:17 ashimema cheers chaps
14:18 d4rkie joined #mojo
14:19 jjohn Candy, that is the way I would build it.
14:19 D4RK-PH0ENiX joined #mojo
14:21 d4rkie joined #mojo
14:22 D4RK-PH0ENiX joined #mojo
14:23 d4rkie joined #mojo
14:25 cpan_mojo Clustericious-1.11 by PLICEASE https://metacpan.org/release/PLICEASE/Clustericious-1.11
14:25 d4rkie joined #mojo
14:27 D4RK-PH0ENiX joined #mojo
14:27 jjohn Because I am nuts, I added a local mojo command called "complexity" that analyzes all my .pm files for cyclomatic complexity and highlights the worst offenders.
14:27 jjohn For example:
14:27 jjohn lib/Xqursion/Controller/DependencyGroups.pm[edit]: lines:  17; complexity: 12
14:28 jjohn Any routine with a complexity above 10 is worth looking at rewriting, if possible.
14:28 d4rkie joined #mojo
14:29 jjohn That edit function isn't too bad, but I am sure that in 6 months, I'll wonder what it's doing.
14:29 CandyAngel :P
14:29 CandyAngel I've made a note to look into that, sounds useful
14:29 d4rkie joined #mojo
14:30 d4rkie joined #mojo
14:31 d4rkie joined #mojo
14:31 d4rkie joined #mojo
14:31 d4rkie joined #mojo
14:32 d4rkie joined #mojo
14:32 jjohn Candy, it can be.  The perl module that does the heavy lifting is: Perl::Metrics::Simple::Analysis
14:32 d4rkie joined #mojo
14:33 ashimema sounds interesting.
14:33 d4rkie joined #mojo
14:33 ashimema care to share the code being the complexity command you built then?
14:33 ashimema certailny sounds pretty useful
14:34 d4rkie joined #mojo
14:34 jjohn ashimema
14:34 ashimema jjohn
14:34 jjohn Sure, it's already in git as part of another project
14:34 d4rkie joined #mojo
14:34 jjohn (sorry, stupid fingers)
14:34 ashimema awesome :)
14:35 jjohn This is the project: https://github.com/taskboy3000/Xqursion
14:35 d4rkie joined #mojo
14:35 jjohn Here's the file: https://github.com/taskboy3000/Xqursion/blob/master/lib/Xqursion/Command/complexity.pm
14:35 d4rkie joined #mojo
14:36 jjohn Also, I want to plug the very excellent List::Utils.  It is the module I always wanted.
14:36 d4rkie joined #mojo
14:36 jjohn er , List::Util
14:36 jjohn http://perldoc.perl.org/List/Util.html
14:36 ashimema hehe..
14:36 ashimema yeah, I love List::Util since I found it :)
14:36 d4rkie joined #mojo
14:37 ashimema My coding has come on a long way in the past year of lurking in here and on the dbic channel.
14:37 d4rkie joined #mojo
14:37 ashimema it's tidbits like this that make all the difference :)
14:38 d4rkie joined #mojo
14:38 CandyAngel ashimema: I'm not sure if my coding has come a long way, but my coding style has definitely changed :P
14:38 d4rkie joined #mojo
14:39 d4rkie joined #mojo
14:39 jjohn ashimema, that was the case for me in 1997-2000 when I participated in #perl. :-D
14:39 ashimema haha.. I started at a very poor level.. so although it's come a long way.. it's still pretty poor comparitively ;)
14:40 ashimema I've started lurking in #perl too now..
14:40 pierrick joined #mojo
14:41 CandyAngel I don't think I'll ever be happy with my coding. Having both 'Jack of All Trades' and 'Perfectionist' traits really sucks :P
14:41 jjohn No one is born a coder.
14:41 CandyAngel For example, took me ages to be happy enough to put my OpenHMD module on CPAN (my first!) and now I'm stalling writing OpenHMD::OO because I don't want to make a poor API
14:42 Grinnz https://metacpan.org/source/JPEACOCK/version-0.9912/vperl/vpp.pm#L147
14:42 jjohn Also, I have just discovered magit, which is an unspeakably awesome git workflow thing for emacs.
14:43 jjohn Candy, good on you!
14:43 CandyAngel Some things you find out about just like.. change how you work completely
14:43 pierrick Hi, i know there is some Hearthstone players here, i've just launch a new website using Mojolicious : http://www.hearthstone-news.com/ still in beta and a lot of works to do but it s a start
14:43 nic In terms of intuition, it's a shame that ->query([..]) is for merge and ->query({..}) is for append, seeing as those datatypes would inutitively be the other way around -- [..] for multiple values and {..} for single values
14:43 CandyAngel Mojolicious, git-annex, Leo...
14:44 jjohn pierreck, cool. I like both of those things.
14:44 nic sri: Mojolicious::Plugin::TagHelpers::_form_for has "$url->query({_method => $method})"
14:45 nic I'm wondering if that should ([..])
14:45 jjohn Those twitter tags, those are to support twitter cards?
14:45 pierrick jjohn me too such a great game
14:45 pierrick probably soon yes
14:45 nic I can't think why you'd want method to append, but there's a strong chance there's something I'm not seeing
14:45 jjohn pierrick, I gather the Magic folks disagree, but that's OK.
14:46 jjohn I enjoy it.
14:46 jjohn Good way to kill 15 minutes.
14:47 jjohn nic, I think you do want that.
14:47 nic oh, I'm guessing it's cos if the method is 'PATCH' then you do want method => POST, method => PATCH
14:47 jjohn It is for RESTful stuff.
14:47 CandyAngel I have a secondlife-esque VR thing that uses all the main projects I use in some way
14:47 jjohn Yes, you got it.
14:47 CandyAngel I'm sure at some point, Mojo::IOLoop::Client will find its way in for the realtime netcode stuff :P
14:49 jjohn nic, I try to always use route names so that I do not have to worry about specifying the method.
14:49 jjohn That's a hold-over from RoR.
14:49 jjohn I sort of wish there were a simple_form thing for Mojo, but I am too lazy to write it. :-p
14:52 CandyAngel Sometimes I get tempted to write ACME::Babble::CandyAngel and just write up Perl modules and stuff I like to use
14:52 nic I'm trying to work out how to bash form_for into behaving like form_with, but it looks impossible
14:53 nic You can pass in {..} but I think that only helps with placeholders, not query params
14:54 CandyAngel nic: Do you mean like: form_for url_with->query(extra => 'param') or something else?
14:55 d4rkie joined #mojo
14:55 nic Perhaps form_for url_with->query([extra => 'param'])
14:58 nic amazing.  my login process now works a peach
14:58 CandyAngel Yay!
14:58 nic I certainly made a meal of that.  Thanks for humouring me :)
14:58 CandyAngel I'm just glad I was useful for a change :P
15:01 gryphon joined #mojo
15:03 d4rkie joined #mojo
15:15 d4rkie joined #mojo
15:16 asarch joined #mojo
15:29 jberger my coding style has certainly coalesced around most of the mojo style
15:29 jberger there are a few points upon which I differ
15:29 jberger but for the most part
15:29 jberger I don't care about line length (below some reasonable degree)
15:30 jberger and in documentation I like one line per sentence (from the LaTeX days)
15:30 jberger (which motivates the first one, since a sentence may be longer than 80 chars)
15:31 Grinnz i dont care about line length too much but i still like to wrap it for docs, since i'm often editing in nano
15:33 jberger https://xkcd.com/378/
15:34 Grinnz lol
15:36 voldemortensen joined #mojo
15:38 CandyAngel I'm a "<=78 character per line"-r
15:39 CandyAngel The main change in my coding since being in here is leaving out a lot of whitespace I used to put in
15:42 jberger with perl, if you put too much whitespace in, how will anyone recognize the line noise?
15:43 CandyAngel :P
15:43 CandyAngel I have a script from umm... 2 or 3 years ago.. that I still have to update. My coding has changed *so* much
15:44 sri ashimema: SNI in Mojo::Server::Daemon would be pretty easy, but due to lack of interest i doubt it will happen anytime soon
15:44 ashimema cheers sri, always good to get some insight.
15:45 ashimema shame few are interested.. would be nice to not need the extra layer in the stack
15:45 jjohn ashimema that was what I was saying 2-3 years ago.
15:46 jjohn Now I like the extra layer in stack.
15:46 ashimema but not a biggie... so i suppose I fall into the camp of 'not enough interest to pursue it' ;)
15:46 jjohn When you get to load balancing your app, you get extra layers anyway.
15:46 ashimema indeed you do ;)
15:47 jjohn apache and nginx are battle-hardened. Where is the gain to making a perl daemon that strong?
15:47 jjohn It is is a never-ending battle against hackers.
15:48 jjohn At least with a proxy upfront, you can worry more about the security problems that you introduced. :-D
15:48 ashimema fair enough point
15:48 jjohn Writing secure software is hard. Even BASH had a remote exploit and that's nuckin' futs.
15:49 ashimema out of interest.. do you guys tend to use Toadfarm at all.. or just reach straight for hypnotoad for deployment?
15:49 jjohn me? hypnotoad with apache.
15:49 ashimema question was wide open to anyone ;)
15:49 jjohn My answer stands.
15:50 ashimema cheers
15:50 ashimema do you have an init script for hypnotoad?
15:50 ashimema or just go in and start it yourself as whatever user
15:50 jjohn Good questions.
15:50 jjohn I have a deploy command
15:51 jjohn That restarts hypnotoad
15:51 ashimema deployment is the question i've been grappling with for the past few days..
15:51 ashimema too many ways to skin that particular cat ;)
15:51 sh4 joined #mojo
15:51 jjohn Ah. One way to do it: https://github.com/taskboy3000/Xqursion/blob/master/lib/Xqursion/Command/deploy/taskboy.pm
15:52 * ashimema reads
15:52 jjohn That is *almost* right.  There is a file of environment variables that needs to be sourced.
15:52 jjohn You would see the environment shell script in git. :-p
15:53 jjohn (I hope)
15:53 sri ashimema: https://github.com/kraih/mojo/issues/882
15:53 henq joined #mojo
15:53 jjohn Carton is really helpful for this too
15:53 ashimema indeed..
15:54 ashimema I'm using carton for development..
15:54 ashimema so my workflow for dpeloyment at the moment is to use carton to install perl deps and plenv for a newer perl
15:54 ashimema then my Toadfarm init script points at plenv for the perl
15:55 jjohn Sounds good.  I am using system perl on my server, because it is my server.
15:55 ashimema we have bare metal servers..
15:55 jberger tonight is Chicago Perl Mongers, tomorrow I think I need to zone out, then maybe over the weekend I can get back to the transaction refactoring
15:55 ashimema but as a company all are running stock debian
15:55 jberger unless someone else wants to take a stab
15:56 ashimema sri :)
15:56 ashimema I wondered if you could actually put the dns into the uri part?
15:58 ashimema so: listen => ['https://mojolicio.us*:443?cert=foo.crt&key=foo.key', 'https://mojolicious.org*:443?cert=foo-org.crt&key=foo-org.key']
15:58 ashimema oops.. meant to remove the *'s there
15:58 ashimema I presume there's something really obvious I'm missing there though?
16:10 sri nic: yea, i really wish we had used [] and {} the other way around too
16:10 sri (in ->query(...))
16:11 sri pierrick: \o/
16:12 * sri still plays hearthstone every now and then... but is stuck at rank 1 like forever :S
16:12 pierrick sri: rank 1 is good, still 1 step to legend
16:13 CandyAngel I'm going to fork Mojolicious and implement all these breaking changes.. I shall call it.. Jojolicious!
16:13 sri mojojojojojolicious
16:13 pierrick sri: i'm a modest rank 10 player for the moment
16:13 CandyAngel yeah! Every time you introduce a breaking change, you get an extra "jo"
16:14 jberger at $previous_company, when I came on as their mojo expert, they made my default password mojojojo
16:14 sri pierrick: i was about to quit because TGT was so bad, but LoE really saved the game
16:15 pierrick sri: so true
16:16 * CandyAngel is porting one of her scripts to a Mojolicious CLI
16:16 CandyAngel Mojo::Collection's 'map' is very handy for this :)
16:17 CandyAngel Wish it had a (cached) ->contains, then I could drop all these lookup hashes
16:18 CandyAngel Oh wait, what am I even doing? I got Array::Diff so I wouldn't need to do it this way! rgh
16:20 sri re Mojo::JSON number detection, is this a problem? perl -Mojo -Mversion -E 'say j({v => version->new("1.0")})'
16:20 sri "operation not supported with version object at /Users/sri/perl5/perlbrew/perls/perl-5.22.0/lib/site_perl/5.22.0/Mojo/JSON.pm line 252, <DATA> line 2125."
16:20 sri as it happens, version overloads the operators with die('...')
16:21 sri with the old implementation the result would simply be "{"v":"1.0"}"
16:22 Grinnz_ an object could also overload stringify with die
16:22 sri indeed
16:23 Grinnz_ hmm should the new check even apply to refs?
16:23 Grinnz_ (if you're talking about the recent change)
16:25 CandyAngel Hm, maybe I won't use Array::Diff. Obvious bug is obvious
16:25 sri Grinnz_: good question
16:34 PryMar56 joined #mojo
16:36 good_news_everyon joined #mojo
16:36 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/v0v2O
16:36 good_news_everyon mojo/master 7d5a82a Sebastian Riedel: fix a few reference encoding bugs in Mojo::JSON
16:36 good_news_everyon left #mojo
16:36 sri Zoffix/haarg: that should take care of the problem
16:38 CandyAngel Errr.. wow, Array::Diff seems really broken. Every time I run it, I get a different result :|
16:40 CandyAngel Oh wait, ignore that, I think I did something wrong. -.-
16:41 CandyAngel (this is why I don't like releasing my code to the public!)
16:41 sri public scrutiny tends to make you a better programmer though
16:42 CandyAngel There's no cure for using 'return' to exit a while loop several years into coding Perl :P
16:43 sri i use return to exit loops all the time
16:43 CandyAngel There's no cure for using 'return' to exit a while loop *when you just wanted next* several years into coding Perl :P
16:44 CandyAngel Is that better? :D
16:44 ashimema lol
16:44 sri yes, better... i mean... worse ;p
16:44 tyldis joined #mojo
16:45 CandyAngel That's why I was getting inconsistent results. Set::CrossProduct returns its tuples in a different order each time I run this command
16:46 CandyAngel So when I returned, it was jumping out of the command on some random tuple
16:46 CandyAngel Figured it out when I put 'die' at the end and it wasn't dying :D
16:46 * CandyAngel has awesome debugging skills
16:47 Grinnz_ heh
16:47 CandyAngel I'm porting this to Mojolicious commands for now, hopefully I can figure out a way of putting a webui on it :)
16:48 CandyAngel Seeing as a lot of the processing is in helpers now
17:14 dod joined #mojo
17:18 jberger Zoffix++ # Test::Mojo::Role::Debug
17:19 jberger its so awesome I almost want to propose it as an addition to Test::Mojo
17:39 lluad That looks rather useful.
17:47 sue joined #mojo
18:11 janus joined #mojo
18:14 sri mojolicious needs something like this http://www.django-cms.org
18:15 sri a polished 90% solution for all things cms
18:24 Grinnz_ time to fix up galileo!
18:36 jjohn sri, had a similar problem with RoR, believe it or not. The simple CMS (Radiant) could not work on recent ruby builds.
18:36 jjohn It seems like CMSs are often used to manage I18N issues.
18:36 jjohn So there is utility in a simple one.
18:37 jjohn Do you see the CMS as a separate app or as a Mojo plugin?
18:38 * pink_mist would assume app .. like Galileo
18:38 jjohn That would be easier to build
18:39 jjohn Why shouldn't people just use Galileo then?
18:39 pink_mist they should..
18:39 pink_mist but apparently it needs fixing up
18:39 jjohn Ah.
18:48 sri fixing up, polish and a pretty website
18:49 sri a good cms needs to be maintained well
18:50 sri if you don't have at least a monthly release with visible imporvements i don't think i'd let it count ;p
18:52 jberger monthly release :(
18:52 jberger jjohn: for something as large as a CMS you should make that an app, which you can mount into another application
18:52 jberger (at least that's how I see it)
18:53 jberger I do have ideas in mind for Galileo (or a replacement)
18:54 jberger and of course if anyone wants to offer pull requests
18:56 sri jberger: i think you're not passionate about cms design
18:56 sri or rather, passionate enough
18:56 jberger that's very possibly true
18:56 jberger I would love to collaborate with someone on something better
18:57 jberger the replacement I would have in mind would be api-backed
18:57 jberger I really want to try to make collborative editing possible
18:57 sri you make good tech choices, but you need someone who cares about tinkering with the little details until it's perfect
18:57 jberger but I keep striking out on that
18:58 jberger sri: yes, that's absolutely true
18:58 Grinnz ... a designer!
18:59 sri i would have said someone with a focus on the frontend
18:59 jberger I did the same thing in physics, I would get the solution to a rediculous problem down to an analytic result and often didn't even plug the numbers in at the end
18:59 jberger to me the result was the answer, the numbers were just a case example
18:59 Grinnz "web designer"
18:59 pierrick joined #mojo
19:00 sri "frontend engineer"
19:00 Grinnz i dont remember the job titles the frontend designers at my job have heh
19:00 sri when i hear web design, i think of folks mocking sites in photoshop
19:00 jberger oh, a web developer like this? https://remysharp.com/images/tbl-developer.jpg
19:00 Grinnz well that is part of it
19:00 lluad css-pusher, ux-ninja, ...
19:01 Grinnz jberger, lol
19:01 sri jberger: lol
19:01 jberger I need to memeify that one
19:01 jberger I'm sure I saw one before
19:01 pink_mist joined #mojo
19:01 jberger I CAN HAZ UNDERSTATE
19:07 trone joined #mojo
19:09 Quai css-pushers? like this guy? https://twitter.com/wiumlie/status/664802963915960324
19:26 gryphon joined #mojo
19:32 pink_mist oh god, that tbl pic, jeez
19:34 sri Quai: that doesn't belong here
19:37 irqq joined #mojo
19:41 miller joined #mojo
20:16 Quai sri: ok. I though we where posting pictures of web-tech inventors in embarrising moments... :P
20:20 disputin joined #mojo
20:22 sue joined #mojo
20:29 jberger so I'm trying to make a demo wherein I mount a PSGI app inside a mojo one
20:30 jberger is there an easy way to tell Catalyst that its base path is not /
20:30 jberger ?
20:30 pink_mist maybe it would be better to ask in #catalyst? :P
20:32 jberger I guess so
20:54 good_news_everyon joined #mojo
20:54 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/v0Jl7
20:54 good_news_everyon mojo/master 1172b9b Sebastian Riedel: better overload test
20:54 good_news_everyon left #mojo
20:54 disputin joined #mojo
20:58 disputin joined #mojo
21:08 jberger yay, adapted https://metacpan.org/pod/Plack::App::URLMap for the purpose
21:14 * sri adds more stuff to the todo list https://github.com/kraih/mojo/issues
21:30 harleypig Is there a way to setup routes in MyApp::Controller::MyController?
21:31 sri not without hacks
21:31 sri like http://blog.kraih.com/post/43198653667/mojolicious-hack-of-the-day-mojolyst
21:34 harleypig That's what I was beginning to think. I'm going to be having other devs working on the various controllers/models and I want to minimize the amount of possible damage they can do. That includes myself in 6 months. :]
21:35 sri have good tests
21:35 sri central routing shouldn't be a big deal
21:37 harleypig It's not really, I just like having as much related material in one place as possible. It makes sense to me the way the Controller and Model idea is broken up, but it doesn't make sense to me that routes should be separate from the controller.
21:38 sri decentralized routing might even be worse
21:38 sri since controllers have no prefix, your devs don't even see if they are stepping on eah others toes with similar routes
21:39 harleypig Yeah. Each has its plusses and minuses. Route collisions would be increased ... right.
21:39 harleypig There is not warning if a route is overwritten?
21:39 harleypig s/not/no/
21:42 irqq_ joined #mojo
21:43 sri the router doesn't know, you're just adding regular expressions to a stack
21:43 harleypig oic
21:44 harleypig So, even with centralized routing, there's a non-negligible chance of collisions, but it's easier to track down. Ok, that makes more sense to me now.
21:45 Grinnz_ consider the routes: '/foo/bar' and '/foo/:param'
21:45 Grinnz_ perfectly reasonable to use together if ordered properly
21:46 Grinnz_ but could be the result of an accidental collision, too
21:49 DFox joined #mojo
21:50 harleypig Right ... if /foo/:param comes first then /foo/bar won't be processed correctly (assuming /foo/bar wants to go to a different method).
21:51 sri i guess you might have expected a radix tree or so as internal data structure, but that just doesn't work with placeholders
21:52 harleypig Yeah, or at least it doesn't work efficiently enough. Way too much backtracking would be needed.
22:15 jwang joined #mojo
22:58 d4rkie joined #mojo
23:00 D4RK-PH0ENiX joined #mojo
23:00 d4rkie joined #mojo
23:03 cpan_mojo Mojolicious-Plugin-MountPSGI-0.08 by JBERGER https://metacpan.org/release/JBERGER/Mojolicious-Plugin-MountPSGI-0.08
23:04 d4rkie joined #mojo
23:09 jberger ^^ mount with rewrite on the url ala Plack::App::URLMap
23:14 disputin joined #mojo

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