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

IRC log for #mojo, 2015-04-07

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

All times shown according to UTC.

Time Nick Message
00:03 mattastrophe1 joined #mojo
00:34 good_news_everyon joined #mojo
00:34 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vez8w
00:34 good_news_everyon mojo/master 7efcb3e Sebastian Riedel: try not to resolve the application path
00:34 good_news_everyon left #mojo
00:37 mattastrophe1 left #mojo
00:54 good_news_everyon joined #mojo
00:54 good_news_everyon [mojo] kraih tagged v6.06 at a5682e1: http://git.io/vezEd
00:54 good_news_everyon left #mojo
00:55 good_news_everyon joined #mojo
00:55 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vezuC
00:55 good_news_everyon mojo/master f964d19 Sebastian Riedel: bump version
00:55 good_news_everyon left #mojo
00:56 cpan_mojo Mojolicious 6.06 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojolicious-6.06
01:08 inokenty-m1 joined #mojo
01:36 woz joined #mojo
01:49 klapperl_ joined #mojo
01:53 woz joined #mojo
01:53 frederico joined #mojo
01:54 inokenty-w joined #mojo
01:55 gryphon joined #mojo
02:11 noganex_ joined #mojo
02:30 woz joined #mojo
03:06 kaare joined #mojo
03:12 disputin joined #mojo
03:45 basic6_ joined #mojo
03:49 jberger today's the day cpan_mojo :'(
03:51 asarch joined #mojo
03:53 KCL joined #mojo
04:16 bc547 joined #mojo
05:04 rwp joined #mojo
05:28 arpadszasz joined #mojo
05:36 dotandimet joined #mojo
05:38 berov joined #mojo
06:04 hshong joined #mojo
06:07 dod joined #mojo
06:08 woz joined #mojo
06:08 Dandre joined #mojo
06:10 dod joined #mojo
06:11 Adurah joined #mojo
06:14 bramirez1 joined #mojo
06:15 melo joined #mojo
06:19 batman jberger: what about web scraping of https://twitter.com/cpan_new ?
06:53 ashimema joined #mojo
06:57 fhelmber_ joined #mojo
07:00 ashimema joined #mojo
07:05 woz joined #mojo
07:06 dotandimet joined #mojo
07:17 AndrewIsh joined #mojo
07:22 trone joined #mojo
07:29 melo joined #mojo
07:29 bramirez1 joined #mojo
07:35 KCL_ joined #mojo
07:45 eseyman joined #mojo
07:50 bramirez1 joined #mojo
08:05 odc joined #mojo
08:14 Vandal joined #mojo
08:15 bramirez1 joined #mojo
08:35 mattastrophe joined #mojo
09:05 frederico joined #mojo
09:14 amon joined #mojo
09:18 Vandal joined #mojo
09:41 meshl joined #mojo
09:56 mattastrophe joined #mojo
10:00 Zoffix Just ask GumbyPAN for sauce :)
10:01 Zoffix Or ask PAUSE admins for a hook to the upload form :D The second someone hits "Upload", the announcement happens haha :D
10:21 marcusr web hookers
10:30 meshl joined #mojo
11:00 Kripton joined #mojo
11:18 mishantil marcusr: web hookers? That's what developers are right?
11:18 mishantil I at least try to maintain the natural order of things and degrade mine as much as possible,
11:29 mattastrophe joined #mojo
11:42 marcusr mishantil: extra income?
11:43 risugg joined #mojo
11:46 harry joined #mojo
11:53 melo joined #mojo
11:57 mattastrophe joined #mojo
12:03 neilhwatson joined #mojo
12:05 inokenty-m joined #mojo
12:09 jberger batman: that will be the plan, miyagawa said that that will likely be his new feed
12:09 bzero Hello. How to use Plugin::Config not to merge, but only getting a config data?
12:10 jberger Zoffix: cpan_mojo is way faster than GumbyPAN and smarter
12:10 jberger bzero: you can just slurp
12:11 jberger there is very little magic in the config plugins
12:12 batman jberger: "said that that will" <--- what does that mean?
12:13 jberger @cpan_new used to depend on miyagawa's feed
12:13 jberger same as cpan_mojo
12:14 jberger miyagawa said that he will just push to @cpan_new instead now
12:14 jberger but I don't believe he has done it yet
12:15 bzero jberger: Yes. But I need to get a magic of .production.conf .development.conf and .test.conf
12:16 bzero I like P::Config, but it always merges data to the main config. ;)
12:16 jberger of course it does
12:16 bzero I want to use loading configuration files from the script.
12:16 jberger ?
12:18 bzero I use my $t = Test::Mojo->new('App'); my $config = $t->app->config; ;)
12:18 bzero And I want to load some data from one more config file (servers.conf)
12:18 bzero But like $config->{servers} = load_config('servers.conf');
12:20 bramirez1 joined #mojo
12:20 jberger subs like something that won't be too hard for you to accomplish
12:20 jberger *sounds
12:20 jberger (daily reminder: don't swipe before coffee)
12:21 bzero Yes. It was not a good idea to use P::Config for this purpose. ;)
12:21 jberger have you checked cpan, there are at least a few configuration modules available
12:24 jberger anyway what you are asking for is clearly beyond the scope
12:25 jberger to want mode file merges but not merge into the main config is going to be a very rare occurrence
12:39 mattastrophe joined #mojo
13:09 alexbyk joined #mojo
13:11 melo joined #mojo
13:21 zivester joined #mojo
13:29 mustardrat joined #mojo
13:30 panzana` joined #mojo
13:30 harry left #mojo
13:32 mustardrat Hi.  I wonder if someone can give me some advice.
13:32 hernan604 joined #mojo
13:32 oetiker joined #mojo
13:33 mustardrat I'm implementing a websocket service which serves up a log file of around 5000 lines.
13:33 HtbaaPi joined #mojo
13:33 mustardrat The file transfer works OK, but after the file has been received by the client, my webapp uses up 100% of one of the server CPUs for approx 5 minutes afterwards.
13:34 fhelmber_ joined #mojo
13:35 mustardrat The server is locked up and won't accept any more connections during this time.  I'm wondering what I have done wrong..
13:36 Lee joined #mojo
13:40 mustardrat To be more specific, the webapp spins at 100% CPU until the websocket inactivity timeout closes the connection.
13:43 genio can you provide a gist of code that shows how you're providing the file?
13:45 mustardrat I have tried a couple of different methods, but at the moment, as a simple test case, I'm reading the file into an array using File::Slurp, and then sending the first 5000 lines of the array using ->send().
13:46 mustardrat my @log = read_file ('/var/log/some.log');
13:46 mustardrat while ($n < 5000) {
13:46 mustardrat $self->send ($log[$n]);
13:46 mustardrat ++$n;
13:46 mustardrat }
13:46 mustardrat That bit seems to be working OK.  It's what happens afterwards that I'm confused about.  There's obviously some concept that I haven't grasped.
13:50 jberger mustardrat: please don't paste into the channel
13:50 genio Well, I don't think I know the answer to your actual question, but from the chatter about File::Slurp, I'd advise against it as it has some problems that won't be fixed.  Mojo::Asset::File will allow you to get chunks or slurp the entire file properly if you want that.
13:51 gryphon joined #mojo
13:52 jberger mustardrat: also, is there any reason not to just JSON encode \@log and send it?
13:53 mustardrat There's no real reason to avoid using JSON.  I'll try that.  Thanks also for pointing out Mojo::Asset::File.
13:54 mustardrat (last comment to genio)
13:54 Mikey is it freaking out with 'multibyte character in syswrite' or something like that?
13:54 Mikey logs might have damn near anything in em.
13:55 Mikey json encoding will utf-8 encode it for you and make everyone happy
13:55 jberger without seeing mustardrat's code, I'm guessing that rather than closing the websocket, its entering some kind of busy state
13:55 jberger which is why I suggested not looping
13:56 genio mustardrat: I'm thinking you should go to gist.github.com and provide a full sample of your code and then people can give better suggestions as to what might be wrong
13:57 Mikey also: strace is your friend.
13:57 mustardrat Mikey: you might be right about the utf-8 encoding.
13:57 Mikey i am pretty sure i am, but as others have said, i haven't seen your code.
13:58 genio oddly enough, that's part of the problem with File::Slurp
13:59 disputin joined #mojo
14:00 jberger note that Mojo::Util has a slurp function too
14:00 mustardrat Thank you everyone for your help and ideas.  I hadn't thought about the utf-8 encoding issue.  It seems that using JSON will achieve what I want in a more standard way, and fix the encoding at the same time. I'm going to give this a go.  And I'll look into Mojo::Asset::File.
14:01 odin22 joined #mojo
14:02 jberger sri: should Mojo::Collection have a TO_JSON?
14:02 jberger then the usage above could just be:
14:02 jberger $c->send({json => b($filename)->slurp->split("\n")})
14:03 jberger but the alternative isn't too bad I suppose
14:03 jberger $c->send({json => b($filename)->slurp->split("\n")->to_array})
14:04 Grinnz jberger, you're not actually planning on web scraping instead of using the twitter api right? :P
14:04 Grinnz https://github.com/Grinnz/zircbot/blob/master/lib/Bot/ZIRC/Plugin/Twitter.pm if you need some inspiration
14:05 jberger no I was going to port miyagawa's AE twitter stream api client
14:05 Grinnz ah right, the streaming api thing
14:05 jberger (which was what I did for friendfeed (at least initially))
14:07 jberger Grinnz: your bot seems to use the twitter api too
14:07 Grinnz yes that's why i linked it
14:07 jberger any thought to comparing it to https://metacpan.org/pod/AnyEvent::Twitter::Stream , factoring out the reusable bits and make a Mojo:: twitter library?
14:08 Grinnz well the streaming API is a bit different, but i'm sure it could be done
14:09 jberger oh, yours accesses the individual tweets
14:13 jberger Grinnz: I'm hoping to have some time to port AE::T::S soon
14:13 jberger at which point I will port cpan_mojo to it
14:14 jberger (and then poke batman to switch it over, since he currently hosts it)
14:14 marcusr I'd like to see a mojo::twitter too, for marvin =]
14:15 Grinnz honestly i rarely bother with modules to deal with APIs (either using or creating) for a few reasons; it's incredibly simple to use most APIs especially with Mojo::UA now and its builtin json/xml handling; i've had bad experiences with poorly maintained API modules; and usually i only need one or two functions of what may be a massive API
14:15 Grinnz but i can see the appeal in general
14:16 jberger mo modules, mo problems
14:17 marcusr Bigger modules!
14:17 marcusr modules of steel
14:17 jberger O
14:18 jberger I'm still proud of my friendfeed module, I think it did solve a very real problem of wrapping a not-very-mojoable mechanism into a mojo event emitter
14:18 jberger CURSE YOU FACEBOOK!!!!!
14:18 marcusr what is a friendfeed?
14:18 purl hmmm... a friendfeed is shutting down
14:18 jberger marcusr: it was a simple push notification service
14:18 jberger free
14:19 marcusr like pushbullet?
14:19 irq joined #mojo
14:19 jberger no, more like twitter I suppose
14:20 Grinnz you know, i dont use eventemitter nearly enough, i can think of one way at least i could use it in my bot
14:20 jberger the only reason I ever used it was miyagawa's cpan module stream
14:20 jberger http://friendfeed.com/cpan
14:20 purl rumour has it http://friendfeed.com/cpan is live, XMPP, comet long-poll and webhook ready
14:20 marcusr I built my whole bot around eventemitter
14:21 marcusr adapters emit messages to routes, and plugins emit messages to adapters
14:21 Grinnz jberger, DOESNT WORK!
14:21 jberger the long-poll api worked well, you request updates and wait, when it gets an update it sent a response and a cursor token
14:21 jberger then you request the same stream with the same token and you pick up where you left off
14:22 jberger Grinnz: it took a long time to respond for me, today is their shutdown day, so who knows
14:22 Grinnz now we have websockets! ;)
14:22 ZoffixWork joined #mojo
14:22 jberger Grinnz: sure, but you have to have a source
14:22 Grinnz indeed
14:23 ZoffixWork What happened to http://blog.kraih.com/mojolicious-is-changing-the-game ? :) I've just made a bowl of popcorn. Would love to read that, if there's any archived copy available.
14:23 Grinnz i'm more interested in what happened to the title on that page
14:23 Grinnz or the header rather
14:23 ZoffixWork heh
14:31 wingfold joined #mojo
14:31 ZoffixWork Well, this was wildly amusing :) http://showmetheco.de/articles/2011/1/why-i-chose-to-build-a-plack-framework-instead-of-using-mojolicious.html
14:32 * sri starts to regret the decision to test rel2abs https://github.com/kraih/mojo/issues/779
14:33 jberger Mojolicious has decoupled quite a bit since that article
14:34 jberger and Mojo is still built with an eye towards a hypothetical non-Mojolicious framework
14:34 jberger I don't know the history of the split with vti, but I do know that it is long in the past
14:34 ZoffixWork Yeah.. 2011
14:37 jberger and Plack still hasn't (and won't IIRC) caught up to the realtime features Mojolicious has
14:38 asarch joined #mojo
14:39 mst ZoffixWork: realistically, that was another case of "death by sri's personality", and then the technical reasons were after the fact rationalisations for how pissed off the guy was
14:40 ZoffixWork "another case"? There are more?
14:40 jberger do we have to do this again?
14:40 ZoffixWork I guess not. I'm just bored here :)
14:41 harry joined #mojo
14:41 mst the important part is the second part
14:42 punter joined #mojo
14:45 ferdinand joined #mojo
14:45 Mikey sri can slash my tires and hit on my wife if he wants, i'll still use mojo.  it's that good.
14:46 ZoffixWork Yeah, I'm loving it too.
14:47 harry left #mojo
14:48 sri i know drama is more fun, but i'd still like more opinions on https://github.com/kraih/mojo/issues/779
14:48 sri the patch that broke it was https://github.com/kraih/mojo/commit/7efcb3e184973040b1acabf8d975092dd92639a2
14:50 ferdinand hi all... i need some help with mojo
14:50 ZoffixWork Ask away, ferdinand.
14:50 ferdinand i'm building a web front-end for some cli scripts... they already have flags built in that can spit out html. this worked fine using ye olde cgi, but mojo wants to "stringify" the html output
14:51 ZoffixWork ferdinand, do you have code you can pastebin? What do you mean by "stringify"?
14:51 ZoffixWork Escape entities?
14:51 ZoffixWork s/entities/special HTML chars/;
14:52 nicomen maybe he is printing out already
14:52 * jberger hits on Mikey's wife
14:52 nicomen but the output needs to go into a variable to send to render?
14:52 ferdinand yes... so route calls a template...
14:52 Mikey lol
14:52 zackiv31 joined #mojo
14:53 ferdinand in the template it checks if a param was provided, if so, passes that param to another external script by way of backticks
14:53 ferdinand loads the result into a var
14:53 ferdinand when i later render that var in the layout, i get
14:53 ferdinand well...
14:53 ferdinand this : <table border="0" cellspacing="0" cellpadding="0">, etc
14:53 ZoffixWork ferdinand, a wild guess is you're using something like  <%= foo %> and you need to change that to <%== foo %> so it won't escape markup
14:53 ZoffixWork (note the extra =)
14:53 ferdinand ahhhhh
14:53 nicomen ;)
14:54 ferdinand yes, i figured there was something, but i didn't know what to search for in the documentation
14:54 ferdinand BOOM... and like that, my page is displaying as intended.
14:54 ferdinand thanks, that was quite painless
14:55 ZoffixWork ferdinand, no problem. See http://mojolicio.us/perldoc/Mojolicious/Guides/Rendering#Embedded-Perl
14:55 Grinnz :)
14:55 ZoffixWork That'll show what you can do in templates
14:56 ferdinand yes, i literally have the rendering guide open, though i must've skimmed over that part :/
14:56 ZoffixWork :)
14:56 ferdinand thanks again
14:57 nicomen the bad thing about mojo docs are that they are so short and concise, if you skim, you will lose out ;)
14:57 nicomen (which is the good thing though)
14:59 sri batman: i know you cared a little bit about this https://github.com/kraih/mojo/issues/779
15:00 ferdinand yes, docs are really great. i'm not a web dev by any stretch... i'm a network engineer who writes a lot of automation tools for my team... and mojo has made building the web frontend to these tools so easy
15:00 sri if there's no consensus on another solution, i'll release 6.07 soonish with s/rel2abs/abs_path/
15:04 ferdinand_dc joined #mojo
15:04 good_news_everyon joined #mojo
15:04 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/veVax
15:04 good_news_everyon mojo/master 8d64f62 Sebastian Riedel: use rel2abs in the test
15:04 good_news_everyon left #mojo
15:06 ferdinand_dc actually i have another q?  i have a list of links (css makes these into menu buttons)...
15:07 ferdinand_dc which i implement in a 'for' loop like so: <li><%= link_to $_ => $_ %></li>
15:08 mst ok, jberger was talking about doing this but clearly doesn't have time
15:08 ferdinand_dc only cause i couldn't figuire out a way if i can nest helpers? like can i do something along the lines of %= t li => link_to 'name' => 'url
15:08 mst can -somebody- please put something somewhere explaining what changes to make to upgrade to 6.x
15:08 good_news_everyon joined #mojo
15:08 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/veVwX
15:08 good_news_everyon mojo/master 01659dd Sebastian Riedel: change the path a little later
15:08 good_news_everyon left #mojo
15:08 mst I have some ideas that require that, and Changes only tells me what broke
15:08 mst not now people should be fixing it
15:09 mst "renamed X to Y" is obvious enough, but "removed X" entries don't say what to use instead :(
15:09 sri some things are just gone
15:09 mst right, but there must be some way to achieve the same goal, no?
15:10 sri like custom socket support, you can't do that any other way
15:10 sri it was just a bad feature, and nobody could come up with a sensible replacement
15:11 mst yeah, but e.g. the 'removed name listing support' thing, IIRC there's another method you can use to do that
15:11 nicomen I miss recipes for all the DOM related changes
15:12 jberger I think that one was removing the helper's param name support
15:12 nicomen and now I'm struggling a bit making something work in m5 and m6
15:12 jberger but the request's param object still can do that
15:12 * sri sets nicomen on fire
15:12 * jberger sets fire en fuego
15:12 nicomen ?
15:12 sri literally the one i made a list for in the list thread https://groups.google.com/d/msg/mojolicious/y_OH1hmxH0Q/PddqzFICiQYJ
15:13 absolut_todd joined #mojo
15:13 mst jberger: right, so there should be somewhere that says, "instead use the request's param object"
15:14 nicomen oh, nice, I am not on the mojolicious mailing list
15:14 ferdinand_dc left #mojo
15:15 sri mst: i always encourage people to post the stuff they stumble about to the release thread
15:15 mst sri: right, but you already know how to fix these - I'm just trying to find a line or two somewhere saying it
15:16 sri well, problem is $c->req->params->names; is not a complete replacement for $c->param;
15:16 mst how so?
15:16 sri if you want *all* names, there is no method
15:16 sri it used to cover route placeholders, uploads and stuff as well
15:17 mst ah, which is a concept that no longer exists, because it wasn't really a good idea?
15:17 sri yes, nobody could come up with a use case
15:17 mst ok, so, in that case it's "if you're using $c->param for the paremeters, use $c->req->params->names instead, if you're using it for something else, rethink" ?
15:18 sri yea
15:19 sri btw. regarding the rel2abs vs abs_path discussion from yesterday, it appears rel2abs has some portability issues
15:20 sri apparently rel2abs calls canonpath on windows too
15:24 Grinnz_ mst: also note $c->req->params->names returns an arrayref where $c->param returned a list
15:24 Grinnz_ same with the cookie names i think
15:27 mustardrat jberger, genio, Mikey:  Thank you for the tips.  I encoded the log file as json, and this has fixed the problem. I can now send > 100000 lines of log data without any problems, and the server is ready to accept new connections straight away afterwards.
15:28 jberger \o/
15:29 mst Grinnz_: right, what I'm saying is there should be some centralised place that documents this, an 'Upgrading from 5 to 6' type thing
15:29 mst or s/there should/that I was hoping there would/ maybe
15:31 Grinnz_ in a "release notes" sort of fashion, i agree
15:31 Grinnz_ could be helpful
15:31 Grinnz_ but where?
15:31 mst https://metacpan.org/pod/Catalyst::Upgrading seems to work for Catalyst's users
15:32 Grinnz_ i could try writing something up to start
15:34 nicomen I would wholeheartedly applaud that :-)
15:34 mst I think that would be a big win
15:35 mst I think part of what upsets people doing conversions is the "I've fixed X, damn, now there's another problem" whereas with a list of the things up front they'd be much less surprised
15:35 mst and while the Changes is sort of that, it still leaves them with the "wtf do I do now" feeling about translating things
15:41 batman sri: i read the issue, but i'm not sure if i get what i should care about...
15:42 batman i think i'm missing out on some details
15:42 Grinnz_ batman: its related to what you mentioned about hypnotoad not following symlinks
15:42 batman yeah. i read the issue :-)
15:42 Grinnz_ well if its reverted that behavior will stay as it was :P
15:43 batman i thought sri only changed from abs_path to rel2abs in the test files, and not in hypnotoad
15:46 Grinnz_ he changed it everywhere i believe
15:54 wingfold joined #mojo
15:55 mattastrophe joined #mojo
16:00 sri Grinnz/mst: feel free to add something like that to the wiki
16:00 Grinnz_ i dont think that would have enough visibility
16:01 sri core docs would not be a good place, since it can't be updated fast enough
16:02 Grinnz_ why not? it only needs to update on breaking changes
16:02 sri people will stumble over things you overlooked, and then the upgrading docs need to be quickly updated
16:02 Grinnz_ at a minimum, major releases
16:02 Grinnz_ i could see a wiki working but only if its linked from somewhere
16:03 sri if it is done well i wouldn't mind linking to it from the official release announcements
16:11 * Grinnz_ is slightly worse at markdown than at pod
16:12 jberger github wiki can be pod
16:12 Grinnz_ huh, how about that
16:12 purl somebody said that was a much more rare case
16:13 Grinnz_ i think thats a better idea actually, for easy links
16:13 Grinnz_ although i think github pod still links to s.c.o
16:13 Grinnz_ so meh
16:14 mst Grinnz_: if you can get the actual information together I wouldn't completely hate fixing up the markdown
16:18 alexbyk joined #mojo
16:18 Onigiri joined #mojo
16:29 risugg joined #mojo
16:32 sri this attitude makes me angry... "yea, some test fails a lot... it's not critical..." https://github.com/kraih/mojo/pull/778#issuecomment-90617083
16:33 Grinnz_ sri: i think he meant (in his mind) that the test was wrong, not that it was failing because it found an issue
16:33 Grinnz_ i'm not sure if that's the case
16:34 sri anyway, back to abs_path
16:34 sri don't want to deal with this anymore
16:37 Oleg joined #mojo
16:41 sri if anyone here really cares, now is your final chance
16:41 sri like batman
16:41 sri pick up the issue and deal with it, or i will
16:42 batman sri: i tried to ask you about it earlier... didn't you just change the tests?
16:42 sri my commit is right there in the issue
16:42 sri https://github.com/kraih/mojo/issues/779#issuecomment-90592126
16:44 punter joined #mojo
16:45 batman ETOOMANYTABS :(
16:45 batman sorry sri. looking now
16:49 Grinnz_ mst / sri: https://github.com/kraih/mojo/wiki/Upgrading what i have so far, any other major section other than adding all the various deprecations and renames?
16:50 Grinnz_ (also i'm not sure all the details on the custom socket support that was removed)
16:53 sri Grinnz++
16:55 sri "Removed accept mutex: Just enjoy the better performance."
16:55 Grinnz_ heh
16:56 sri "Removed Mojo::UserAgent custom socket support: Don't know what this was? Neither do we!"
16:56 batman sri: what's wrong with https://github.com/kraih/mojo/commit/8d64f62bd9721090d6335b41a48d7ed71b0dc41a ?
16:56 sri batman: it's untested
16:57 sri it may work, or it may cause new problems
16:57 sri and i don't want experiments anymore
16:57 batman ok. then i'm +1 on reverting the whole thing :(
16:57 sri ok
16:58 batman i would be very surprised if i came up with any solution that doesn't include experimenting.
16:58 sri that makes the decision to revert unanimous
16:59 good_news_everyon joined #mojo
16:59 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vew2U
16:59 good_news_everyon mojo/master 693a3cf Sebastian Riedel: just use abs_path again
16:59 good_news_everyon left #mojo
17:02 jberger sri: can I assume that the minion plugin will always take two arguments, arranged as a one key hashref?
17:03 jberger at $work we are about to start using the pg backend for some servers
17:03 sri yes
17:04 jberger so I need to $app->home->rel_file a relative path to File, but obviously not touch Pg paths
17:04 jberger cool
17:04 jberger thanks
17:05 sri if i need to pass along more data i'll just make the value a data structure too
17:05 cono Ah, here was a dicsussion...
17:05 cono anyway thanks for the experiment, and sorry
17:06 sri pretty much everything is discussed here
17:06 Grinnz_ even zoidberg
17:07 cono I thought you will wait until I finish my pull-request, but you applied it w/o testing :(
17:07 sri the dangers of opening pull requests early ;p
17:07 Grinnz_ cono: it cannot easily be tested on every platform before release
17:07 melo joined #mojo
17:07 cono I was fooled with labe "work in progress" :p
17:07 cono label*
17:08 cono Grinnz_: why? we have Virtualbox
17:08 sri and yes, we cannot test on many platforms before release
17:08 sri i don't even have any windows licenses
17:08 cono bug which we discovered is simple as hell, I could find this on VirtualBox
17:08 jberger why not zoidberg?
17:08 purl (V) (;,,;) (V)
17:08 Grinnz_ linux is easy enough, and sri runs on os x, but other than that...
17:09 cono I have win7 on virtualbox
17:09 sri in fact, we don't even have a core developer on windows
17:09 sri all windows support is irregular volunteers
17:09 cono and I'm from Ukraine, I don't need a licence for Windows :p
17:09 Grinnz_ lol
17:10 neilhwatson Last devopsdays event I was at, one guy asked about other's XP with Windows. Poor guy, everyone laughed.
17:11 jberger when I use windows, it has been either my ancient win2k iso
17:11 dod joined #mojo
17:11 * Grinnz_ runs win 7, but there's really no alternative for modern gaming yet
17:11 sri you also need to test on many different versions of windows
17:11 jberger or else one of the temporary images m$ gives away
17:11 sri and combinations of different windows/perl versions
17:11 cono sri: /agree, but we could atleast discover this stupid bug earlier
17:12 jberger https://www.modern.ie
17:12 cono there can be potential more bugs, but this one, makes me frustrated
17:12 sri jberger: those virtual machines expire pretty quickly
17:13 jberger which is why I only use them when I really need to
17:13 cono sri: use VirtualBox snapshot mechanism :)
17:13 cono amd you can use this windows forever )
17:13 cono and*
17:14 sri is that compatible with the licensing terms?
17:14 jberger tbh I think their docs say to do that, IIRC
17:14 cono sri: not sure, as I already said - I'm from Ukraine :) I don't understand why people paying for software
17:15 sri i'm from germany, it's illegal here
17:15 trone joined #mojo
17:15 cono I thinks its illegal in our country too. But no1 cares about one particular person
17:15 cono only about companies
17:16 sri legally speaking, i am a company
17:16 cono ah
17:17 sri also, some people pay for software to support the devlopers
17:17 sri you know, to make sure they can continue making new versions
17:18 neilhwatson and fixing bugs ;)
17:18 sri feed their children and stuff
17:22 mst admittedly, it's probably safest that you and I never breed
17:25 ZoffixWork joined #mojo
17:26 ZoffixWork The VMs on https://www.modern.ie last for 90 days... Downloading them 4 times a year doesn't seem too bad. IMO.
17:26 cono in my ideal world, developers should get money by selling service (cpu time) or support (adding additional features). But selling the code is kinda weired for meh :)
17:26 Grinnz_ well, the world is never ideal
17:27 ZoffixWork No word on snapshots in the license agreement and it says "You may use the software for 90 days after it is downloaded to the licensed computer." So I'd say making them would be a violation.
17:27 cono Grinnz_: indeed :)
17:28 ZoffixWork Or the world is *always* ideal :D
17:28 cono it depends in whos hands it is :)
17:28 purl i think :) is all good, and for the records I'm pretty laided back guy
17:29 Grinnz_ purl: forget :)
17:29 purl Grinnz_: I forgot :)
17:29 Grinnz_ purl: you oddball
17:29 purl Grinnz_: what?
17:29 cono :D
17:29 ZoffixWork purl, I though you were a girl! :O
17:29 purl ZoffixWork: what?
17:30 Grinnz_ sri: does this section look accurate? https://github.com/kraih/mojo/wiki/Upgrading#mojodom-renaming
17:31 jb360 joined #mojo
17:32 absolut_todd joined #mojo
17:33 mattastrophe joined #mojo
17:35 cfedde joined #mojo
17:36 good_news_everyon joined #mojo
17:36 good_news_everyon [mojo] kraih tagged v6.07 at 1a06159: http://git.io/vewQ2
17:36 good_news_everyon left #mojo
17:37 cono Nick is very representative for meh :D It was like from professor mouth :D
17:37 good_news_everyon joined #mojo
17:37 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vew7U
17:37 good_news_everyon mojo/master b219f63 Sebastian Riedel: bump version
17:37 good_news_everyon left #mojo
17:40 cono was nice to meet you guys. cu
17:40 cono left #mojo
17:40 berov joined #mojo
17:42 wingfold joined #mojo
17:49 basic6 joined #mojo
17:52 risugg joined #mojo
17:56 Grinnz_ sri: one other thing i'm not sure about, "Removed auto_render method from Mojolicious::Routes."
17:56 Grinnz_ am i remembering correctly, that was removed because it's unnecessary (automatic rendering happens automatically)?
17:59 Grinnz_ other than that i think that page has all the 6.0 changes
18:06 Grinnz_ ill add sections for earlier in the 5.0 cycle later (like Mojo::JSON's OO API)
18:10 k-man joined #mojo
18:18 disputin joined #mojo
18:31 mustardrat joined #mojo
18:33 disputin joined #mojo
18:33 mattastrophe1 joined #mojo
18:47 inokenty-m joined #mojo
19:00 jb360 joined #mojo
19:16 woz joined #mojo
19:21 risugg joined #mojo
19:24 punter joined #mojo
19:47 meshl joined #mojo
19:53 meshl joined #mojo
19:56 odin22 joined #mojo
19:57 irq joined #mojo
20:00 risugg joined #mojo
20:00 martin joined #mojo
20:41 woz joined #mojo
20:51 dod joined #mojo
20:53 d4rkie joined #mojo
20:56 * tempire has successfully installed mojolicious as a dependency in a large corporation
20:56 tempire CONQUER!
20:57 pink_mist \o/
20:58 pink_mist which corp if I may ask?
20:58 D4RK-PH0ENiX joined #mojo
21:00 noganex joined #mojo
21:00 hernan604 lol
21:01 hernan604 i did it too.. =)
21:05 Zoffix I'm hoping to do so in a couple of months :)
21:05 riche joined #mojo
21:06 Zoffix Grinnz_++ # upgrade wiki
21:07 asarch joined #mojo
21:10 neilhwatson joined #mojo
21:11 Grinnz_ https://github.com/kraih/mojo/wiki/Upgrading -- this is all the breaking changes i can find back to 5.0
21:14 sri Grinnz_: looks good, might be worth adding a few links too
21:14 mst awesome.
21:14 Grinnz_ sri: yeah it would be ncie if i could add autolinks to mcpan :P
21:15 Grinnz_ or even mojolicio.us
21:15 sri like the one for the security thread https://groups.google.com/d/msg/mojolicious/aJTYjRCPjOE/QJZB6ORfYJAJ
21:15 Grinnz_ right, more context there would be helpful
21:15 Grinnz_ no pun intended
21:15 Grinnz_ <_<
21:16 sri aww... i liked the pun
21:17 sri hahahaha... you actually did it https://github.com/kraih/mojo/wiki/Upgrading#removed-mojouseragent-custom-socket-support
21:17 Grinnz_ hehe
21:20 mst Grinnz_++
21:20 mst sri++
21:20 Zoffix heh.. I think ZofCMS is vulnerable to that attack :)
21:20 sri Grinnz_++
21:20 Grinnz_ Zoffix: yeah i went through our codebase with a fine toothed comb that day ;)
21:20 Zoffix Or to variation of it. I'm using Vars() from CGI.pm
21:20 Zoffix Grinnz_, did you really?
21:21 Grinnz_ well, a comb named "grep"
21:21 Zoffix oh lol
21:21 Zoffix What did you find? :)
21:21 punter joined #mojo
21:21 Grinnz_ it wasn't too bad, our codebase is mostly newer
21:21 Grinnz_ the other project though probably is still vulnerable
21:22 Grinnz_ Vars isn't a problem i don't think, that returns a hashref doesnt it?
21:22 Zoffix Actually yeah, it's not a problem. It returns a list that I'm storing in a hash. It's just right now I tested it and it concatenated values of multiple same-name params, which kinda threw me off
21:22 Zoffix Since I didn't know it did that.
21:22 Grinnz_ yeah, thats why CGI->Vars is a bad idea :P
21:23 mst join "\0", @values;
21:23 Zoffix mst, what's that?
21:23 Grinnz_ how it concatenates them, i think
21:24 Zoffix Oh, yeah.
21:24 mst wat, Vars returns a hash TIED to CGI
21:24 Grinnz_ hah
21:24 mst and, yep, sub FETCH is
21:24 mst return join("\0", $_[0]->param($_[1]));
21:25 Grinnz_ thats pretty amazing
21:25 mst I'm just surprised I remembered that
21:25 Grinnz_ can people just stop tying things to other things unless its actually helpful
21:25 mst I probably need to drink more
21:28 reneeb joined #mojo
21:33 Grinnz_ hmm... github wiki doesnt have any way to "watch" a page does it?
21:34 Grinnz_ i dont see anything
21:34 mst Ctrl-R
21:34 Grinnz_ lol
21:37 meshl joined #mojo
21:47 harry joined #mojo
21:55 mattastrophe joined #mojo
22:07 mattastrophe joined #mojo
22:25 Grinnz_ " * We're renaming some API every few months just to not let you get bored." someone thinks hes funny :)
22:32 woz joined #mojo
22:35 Onigiri Grinnz_: page2rss \
22:35 Grinnz_ huh?
22:35 Grinnz_ oh, for the watching thing
22:36 Grinnz_ i meant a mediawiki page watch, anythign more complicated i wont bother with :P
23:02 sri Grinnz_: sadly he's not trying to be funny
23:02 sri he writes anti mojolicious blog posts in russian if i remember correctly
23:03 sri and then acts nice around here
23:03 neilhwatson joined #mojo
23:07 sri (yes, i do remember that stuff and will keep an eye on you!)
23:14 Zoffix Who? where?
23:14 sri https://groups.google.com/d/msg/mojolicious/kNWdJsG2lr4/C7RqCGMGNu4J
23:15 jberger https://mytvandme.files.wordpress.com/2011/06/leela-eye-groening-futurama-ecards-someecards.jpg
23:16 Zoffix heh
23:18 Zoffix http://powerman.name/ ? :D
23:19 mattastrophe joined #mojo
23:21 jb360 joined #mojo
23:22 sri this might have been one of them http://habrahabr.ru/post/227493/
23:23 Zoffix "Mojolicious Docs: The Lost Chapters" :D
23:23 * Zoffix is excited
23:25 Zoffix I don't know why he's raging about backwards compatibility so much in that article. You don't have to rush to cpanm every time a new update comes out :S
23:25 Grinnz_ as has been mentioned many times, if you have such a system you should use carton or similar to maintain a predictable perl lib
23:26 * Zoffix nods
23:27 Grinnz_ i mean, Mojo can do all the deprecation policies it wants but if IO::Socket::SSL or even something in core breaks something, that's something else you have to know when updating
23:28 Grinnz_ there are no general guarantees whenever you hit the cpan button ;)
23:28 Zoffix Bahaha. He writes: "And lastly — alphabetised list of methods in the docs, SERIOUSLY?" :)
23:29 Grinnz_ lol
23:29 Grinnz_ he should see the code!
23:29 Zoffix rofl
23:29 Grinnz_ at this point if i saw any non-alphabetized list in the Mojo source i would consider reporting it as a bug
23:29 sri and you would be right
23:30 Zoffix :D
23:30 sri it's literally in the rules :)
23:30 Grinnz_ heh
23:32 * Grinnz_ thinks alphabetizing "new" is still silly, though
23:33 sri for a long time new was actually an exception
23:34 sri but i couldn't really think of a good reason why
23:34 Grinnz_ because it's where everything starts
23:34 * Grinnz_ shrugs
23:34 sri not always
23:35 Grinnz_ it just feels awkward whenever i see it in the middle of some random methods
23:35 sri there's constructors like http://mojolicio.us/perldoc/Mojo/Exception#throw
23:35 Grinnz_ yeah, true
23:35 Zoffix Been a while since I wrote a sub new { ...
23:35 sri or http://mojolicio.us/perldoc/Mojo/Cookie/Request#parse
23:36 sri or http://mojolicio.us/perldoc/Mojo/ByteStream#b
23:36 Grinnz_ heh
23:36 Grinnz_ that last one is just cause of ojo :P
23:37 sri it might actually predate ojo
23:37 sri and might have inspired ojo in the first place
23:37 sri memory is fuzzy though
23:38 Grinnz_ i kind of wanted to steal ath.pm so i could do perl -Math -E 'say calc "2+2"' but i think that might be too silly
23:38 sri anyway, that's why i thought new is not really that special
23:38 sri unlike AUTOLOAD or DESTROY
23:39 Grinnz_ it isn't special to perl, but it's still special conceptually, i think
23:39 Zoffix I think the most clever one is the: perl -Moo
23:39 Zoffix :D
23:40 sri java apis always give me a headache http://netty.io/wiki/using-as-a-generic-library.html
23:40 sri soooo many objects
23:40 Grinnz_ i wonder if with some forbidden magic i could make perl -Math -E '2+2' work
23:44 woz joined #mojo
23:51 Zoffix Grinnz_, http://fpaste.scsys.co.uk/471310
23:51 Zoffix :D
23:52 Zoffix echo "package ath; use Filter::Simple sub { s/\A/say eval / }; 1;" > ath.pm
23:52 Zoffix ^_^
23:55 pink_mist mv Devel/REPL.pm ath.pm
23:55 pink_mist (not sure that'd actually do it though :P)

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