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

IRC log for #mojo, 2018-02-05

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

All times shown according to UTC.

Time Nick Message
00:08 sockmister joined #mojo
00:09 ferreira joined #mojo
02:00 aborazmeh joined #mojo
03:07 JimOK joined #mojo
03:08 JimOK Is this a pretty active channel?
03:10 JimOK I'm new to building websites, and all I want to do is make a clan website for a gaming community.  My only problem is I'm confused about sessions, and data base driven website.  There doesn't seem to be much online for perl programing.  Is there somewhere I should start?
03:25 ferreira joined #mojo
04:13 hesco1 I have a route like this: $r->get('/v1/app/aup/latest_version/*aup_component/do/:action')->to('acceptable_use_policy#action');
04:14 hesco1 but the wildcard placeholder is proving more greedy that I would hope.  Any suggestions?
04:30 Grinnz hesco1: if you want it to allow . but not / then using #
04:31 Grinnz you can't use * if you want to allow further / routing
04:32 hesco1 I rewrote the route as do/:action/*aup_component instead.  but now its grabbing: do/copy/party_loyalty/ten_key_values.txt
04:32 Grinnz JimOK: https://metacpan.org/pod/Mojolicious::Guides#BASICS has some links to start
04:33 hesco1 sometimes it seems like my code version is being cached between tests.  Because subsequent tests do not always seem to respond to changes I have made.
04:36 Grinnz JimOK: Mojo::Pg has some example apps using the database https://metacpan.org/pod/Mojo::Pg#EXAMPLES
04:36 hesco1 JimOK: you want to check out perlmonks.com, as well.
04:57 hesco1 issue related to my test being caught by a previously defined route.  Hope I remember that lesson.  That was borderline painful to sort out.
05:04 dboehmer joined #mojo
06:31 inokenty-w joined #mojo
06:58 Vandal joined #mojo
07:18 berov joined #mojo
07:33 noganex joined #mojo
07:34 McA joined #mojo
07:58 dod joined #mojo
08:00 AndrewIsh joined #mojo
08:02 dod joined #mojo
08:09 dod joined #mojo
08:40 trone joined #mojo
08:54 Edward joined #mojo
09:09 karjala_ joined #mojo
09:13 reetp joined #mojo
10:01 dshadowukraine joined #mojo
11:14 berov joined #mojo
11:15 tchaves joined #mojo
12:35 tcohen joined #mojo
12:58 gizmomathboy joined #mojo
13:00 sri last chance to comment on https://github.com/kraih/mojo/pull/1167
13:16 exp-innit a role seems intelligent
13:16 exp-innit also I'm done with reviews and the like, i'll try and get that PR finished shortly
14:06 jacoby joined #mojo
14:31 mib_01syku joined #mojo
14:42 Pyritic joined #mojo
14:46 eseyman joined #mojo
15:03 maschine joined #mojo
15:06 jamesaxl joined #mojo
15:48 gryphon joined #mojo
15:49 sh14 joined #mojo
15:49 anony joined #mojo
15:59 ChmEarl joined #mojo
16:12 sri i have no idea what to do about this https://github.com/kraih/mojo/pull/1183
16:12 sri it should be impossible
16:13 sri unless someone here has an idea i'll close the issue
16:18 maschine have you read the stackoverflow thread recently?  They mention it could just be the way the docs describe how to set up Hypnotoad with systemd
16:18 maschine https://stackoverflow.com/questions/48242004/why-cant-i-stop-my-mojo-app-using-systemd
16:19 maschine I don't use systemd so I have no idea if that works or not
16:20 sri but that is wrong
16:21 sri yes, systemd will only kill the manager, but when the manager goes away the workers will die on their own
16:23 exp-innit I assume that can't just block indefinitely for whatever reaso
16:25 maschine what's the drawback to killing the workers directly?  not letting them finish whatever task they're working on?
16:26 orev joined #mojo
16:27 Grinnz i've been using KillMode=process on several hypnotoad daemons for months and never had that issue
16:28 sri i just looked at servers at work, and i do in fact not specify a killmode for prefork
16:28 Grinnz i'm not using the prefork daemon of course but that shouldn't make a difference
16:28 sri it is fairly different actually
16:29 sri hypnotoad is a type=prefork
16:29 sri umm, forking
16:29 Grinnz type=forking you mean yeah
16:30 good_news_everyon joined #mojo
16:30 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vNh10
16:30 good_news_everyon mojo/master 393daba Sebastian Riedel: everybody seems to be using no KillMode (closes #1183)
16:30 good_news_everyon left #mojo
16:30 Grinnz the only difference between those is with type=simple systemd will start it and immediately move on, with type=forking systemd waits for it to exit then reassigns the service to its child via the pidfile
16:32 Grinnz so i dunno why it would make a difference to the killmode needed but whatever works
18:01 dod joined #mojo
18:09 tcohen joined #mojo
18:31 nohuhu joined #mojo
18:53 Seth joined #mojo
19:18 trone joined #mojo
19:22 mantovani left #mojo
20:16 polettix joined #mojo
20:30 tchaves joined #mojo
22:21 sri exp-innit: you forgot to update your pull request again *nudge*
22:28 spicyjack all your (re)base are belong to us?
22:39 dustinm joined #mojo
22:43 * sri is really happy we embraced Role::Tiny for extensions
22:44 sri it just works so well
22:50 jabberwok it's Rocky instead of Bullwinkle ... (couldn't resist)
22:51 * Grinnz considered writing the ISO8601 role but looked at the PR code and reconsidered :P
22:52 Grinnz https://en.wikipedia.org/wiki/ISO_8601#Week_dates and https://en.wikipedia.org/wiki/ISO_8601#Ordinal_dates being a thing makes things way more complicated...
22:52 sri Grinnz: yea, that part of the spec is really shitty
22:53 Grinnz i would happily write "iso8601 but only calendar dates" though :P
22:53 sri one of the reasons i like RFC 3339, it's a sensible if maybe a little too strict subset
22:54 Grinnz it could also be super lazy and just depend on https://metacpan.org/pod/distribution/DateTime-Format-ISO8601/lib/DateTime/Format/ISO8601.pod
22:54 sri whitespace instead of the T is friendlier for humans
22:57 Grinnz i like how the module has one 3star review saying to use DateTime::Format::RFC3339 which as expected doesn't actually parse anything but RFC3339
22:57 Grinnz cpanratings at it again :P
22:57 sri i'd like cpanratings a lot more if there were more comedy reviews
22:58 sri like those funny amazon reviews
23:00 jberger Grinnz: a MAJOR reason to recommend that iso8601 support be a role is to let it drag in a dependency
23:00 jberger guess I shoulda mentioned that in my comment too
23:00 Grinnz indeed, role support makes that feasible
23:00 jberger honestly, one might even consider a role that brings in a big date parser module
23:00 jberger a dwim one
23:01 Grinnz eh, i don't like that
23:01 Grinnz those*
23:01 jberger it would be a different role of course
23:01 Grinnz it's useful when you have human-entered dates, but every other use case you should be expecting a defined format
23:01 Grinnz if not iso8601 or rfc3339, then something you can pass to strptime
23:01 jberger I know I mentioned strptime
23:02 Grinnz a Mojo::Date::Role::Strptime would be neat though
23:02 jberger maybe there could a role that has a mapping of accepted patterns
23:02 Grinnz but again, not something you want to reimplement yourself
23:03 Grinnz since it only stores epoch, it would be easy to implement a strptime role using Time::Piece

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