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

IRC log for #mojo, 2016-07-29

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

All times shown according to UTC.

Time Nick Message
00:00 Grinnz_ that's ... quite a web page
00:00 sri a few more methods in mojolicious and it will have chrome crashing qualities
00:00 Grinnz_ ahh sattelite internet... unavoidable 750ms latency or so, last time i used it
00:00 dave chrome crashes with too many methods?
00:01 dave this web index is very...comprehensive :D
00:01 sri any big site will do ;p
00:01 Grinnz_ seems fine in chrome here, mostly because there's no CSS whatsoever :P
00:01 nicomen marcus: add WebRTC support?
00:02 dave is the DOM size the issue or is there any reference you have for that idea?
00:02 sri that's true, i guess big files on github crash it because of syntax highlighting
00:02 dave I never see that, maybe that's cause I use chromium?
00:02 Grinnz_ on metacpan the syntax highlighting is in JS and that can be a huge slowdown on pages like DBI.pm
00:03 pink_mist dave: ... I'm curious what was up with "this irish station" :P
00:03 dave there's an irish jazz fusion station that just played a bunch of really nice jazz fusion, then they played some led zep
00:03 sri they play riverdance 24/7?
00:04 dave *rimshot*
00:04 Grinnz_ dave: well... england is close to ireland!
00:04 dave Grinnz_: the DBI page is not slow at all on my chromium
00:04 dave https://metacpan.org/pod/DBI
00:04 Grinnz_ dave: the source page I meant
00:05 dave ...oh yeah ... 10 sec now and climbing
00:05 dave and .. the unresponsive dialog
00:05 Grinnz_ yep... always fun digging through DBI bugs :P
00:06 dave I'm going to probably get laughed at in -this- channel, but I'm reminded that the browser was originally intended as a documentation tool ...
00:13 dave Mojo::DOM needs a ->diff() method :)
01:08 zivester joined #mojo
01:33 Kripton joined #mojo
01:57 PryMar56 joined #mojo
02:04 Janos joined #mojo
02:37 noganex joined #mojo
02:40 sri don't forget to get your free windows 10 upgrade in the next 24 hours ;p
03:05 kaare_ joined #mojo
04:05 ningu joined #mojo
04:19 dave omg omg omg windows 10 ... I must rush and get it ... maybe DOOM will run faster ;)
04:56 sri i think it actually might
04:57 sri the windows anniversary update in a few days should be pretty cool too, with bash for windows going public
05:32 ningu windows bashes itself, no need for help
05:38 * sri bashes ningu
05:41 Grinnz :)
05:47 ningu haha Trump complained there were too few American flags at the Democratic convention
05:47 ningu just saw a pic from there with flags everywhere
05:54 inokenty-w joined #mojo
06:16 genio joined #mojo
06:23 dod joined #mojo
06:34 Gedge joined #mojo
06:34 che-quest joined #mojo
06:36 Jonis I'm running windows 10 at home, and have had no issues with it
06:37 Jonis although I'm running the insider preview
07:01 marcus jonis inside
07:02 marcus nicomen: not seeing how that would help convos?
07:03 marcus unless you want to call someone on irc, who's also running convos, I guess.
07:06 trone joined #mojo
07:14 Vandal joined #mojo
07:15 Phil21 sri: so today's discussion led to me thinking this out a bit more long-term and talking to some internal folks
07:16 Phil21 I'm going to get a dedicated cabinet setup for this stuff
07:16 Phil21 tldr; expect mid next week, need to get a cabinet assigned and switches configured
07:19 dod joined #mojo
07:22 dod joined #mojo
07:28 AndrewIsh joined #mojo
07:57 sugar joined #mojo
08:13 osfabibisi joined #mojo
08:41 tchaves joined #mojo
08:49 nic Hopefully you're much better at putting together a dedicated cabinet than Theresa May
09:08 ningu joined #mojo
09:15 nicomen marcus: I thought using a webrtc data channel which is udp based might help when roaming interruptions ruined the tcp connection?
09:15 nicomen https://webrtc.github.io/samples/src/content/datachannel/basic/ ?
09:31 orev joined #mojo
09:40 Zx3 joined #mojo
09:48 cuechan joined #mojo
10:06 tchaves joined #mojo
10:51 Jonis
11:07 eseyman joined #mojo
11:13 eseyman_ joined #mojo
11:36 sugar joined #mojo
12:09 eseyman joined #mojo
12:22 tchaves joined #mojo
12:25 mpapec sri: should I see this warning https://metacpan.org/source/SRI/Mojolicious-7.0/lib/Mojo/UserAgent.pm#L53
12:26 mpapec when running https://gist.github.com/mpapec/c857fb3f8c472300bfae906377acca0b ?
12:30 kid51 joined #mojo
12:31 nicomen mpapec: only if you have set MOJO_USERAGENT_DEBUG ?
12:35 cpan_mojo Mojolicious-Plugin-OpenAPI-0.08 by JHTHORSEN https://metacpan.org/release/JHTHORSEN/Mojolicious-Plugin-OpenAPI-0.08
12:41 neilhwatson joined #mojo
12:42 mpapec nicomen: yes, I have: https://gist.github.com/mpapec/c857fb3f8c472300bfae906377acca0b
12:45 nic What debugging proxies to people here use?
12:46 nic (I might buy a licence for charles, but don't want to then discover I'd rather use a different tool)
12:53 nicomen mpapec: why would you not want to get that warning?
12:56 mpapec nicomen: I want it, but I don't get it
12:56 pink_mist mpapec: you may need to wrap the ENV setting in a BEGIN block
12:57 mpapec pink_mist: sec
12:57 nic pink_mist: pretty sure that won't work
12:57 pink_mist nic: pretty sure it might
12:57 mpapec pink_mist++
12:57 Kripton joined #mojo
12:58 nic the env setup has to be in place for the surrounding shell before compile-time
12:58 mpapec I'm not sure why BEGIN is needed
12:58 nic it's not
12:58 pink_mist it is
12:58 gizmomathboy joined #mojo
12:58 nic you need to define the *_DEBUG on the commandline _before_ any code runs
12:58 pink_mist because the use statements are implicitly also wrapped in a BEGIN
12:59 pink_mist so the code that reads the env variable and sets the DEBUG constant gets run before the ENV setting gets executed unless it's also in a BEGIN
12:59 mpapec ah,
12:59 mpapec it is constant => https://metacpan.org/source/SRI/Mojolicious-7.0/lib/Mojo/UserAgent.pm#L14
12:59 mpapec not object property
12:59 nic Essentially it has to be in place before _anything_ uses Mojo::UserAgent
13:00 pink_mist nic: yes, that's what the BEGIN does
13:00 nic not necessarily
13:00 pink_mist uhm, yes
13:00 nic what you're saying hinges on which file and which position you're talking about
13:00 pink_mist because it was above the use statements
13:01 pink_mist we're talking about one specific file
13:01 pink_mist stop being silly
13:01 nic stop being rude
13:02 pink_mist perhaps you should have looked at the file before saying anything
13:08 ningu joined #mojo
13:10 abmoji joined #mojo
13:15 ababu joined #mojo
13:17 Janos joined #mojo
13:19 ababu Can’t get static contents work. domain:port/public/scripts/javascript.js is attempted to be routed.
13:19 ababu http://pastebin.com/Jebebzzb
13:19 ababu this is my code, can any one please take a look
13:20 pink_mist ababu: your $static isn't being connected to the Mojolicious server at all; it's just an object you're creating and setting, and then not using anywhere ever again
13:21 ababu I am new to this, can you show me a sample. If you don’t mind
13:21 Jonis ababu: instead of my $static = Mojolicious::Static->new; you should use my $static = app->static();
13:22 ababu Ok. Thnks . Let me try that
13:22 pink_mist instead of my $static = Mojolicous::Static->new; do my $static = $self->static;
13:22 pink_mist it's not a lite app, Jonis, so 'app' won't work
13:22 Jonis ah yes, sorry.
13:22 Jonis could even push to @{$self->static->paths}
13:22 pink_mist ababu: and you can remove the "use Mojolicious::Static;" too
13:23 pink_mist true, Jonis
13:27 ababu still its not working :(
13:27 ababu http://138.68.11.176:3000/public/index.html
13:27 ababu No clue so far.
13:28 Jonis http://138.68.11.176:3000/index.html
13:28 Jonis public is not part of the path :)
13:28 ababu ohhh
13:28 ababu great
13:28 ababu hoo
13:28 ababu Thanks Guys. Life saver  d:)  [:)
13:29 pink_mist np, happy to help =)
13:31 Jonis np :)
13:41 ababu joined #mojo
13:56 HtbaaPi joined #mojo
14:06 mcsnolte joined #mojo
14:21 PopeFelix How would I find the minimum Mojolicious version I need for Mojo::UserAgent::Mockable? I'd like to stop getting this test failure report from a box w/ Mojo 4.88 on it.
14:21 ptolemarch joined #mojo
14:22 pink_mist well, figure out what features it's using, and check which version those features were implemented :P
14:23 pink_mist the actual error you're getting might help there
14:26 PopeFelix lol
14:27 PopeFelix Well, I'm seeing Mojo::IOLoop::stop_gracefully not being defined.
14:27 jberger I usually pick the last major release before I created a module as its minimum Mojo starting point
14:27 pink_mist https://metacpan.org/changes/distribution/Mojolicious#L510
14:27 pink_mist 5.80
14:27 PopeFelix jberger, I created this under Mojo 6.x.
14:27 PopeFelix pink_mist, thanks
14:27 PopeFelix So 6.0 is a fairly safe minimum, then.
14:29 sri 6.0 is 1.5 years old already too
14:29 sri Phil21: \o/
14:31 sri nic/pink_mist: http://www.amandablain.com/wp-content/uploads/2012/07/smiles.jpg
14:33 PopeFelix Where do I even find Mojo 6? CPAN only seems to have 7.0, 5.17, and a couple of 4.x releases
14:33 sri backpan
14:34 sri you can even install old versions with cpanm and a special syntax
14:34 sri think it was like "cpanm Mojolicious@6.0"
14:34 jberger sri: yep that's it
14:36 PopeFelix thanks
14:36 ababu joined #mojo
14:38 meshl joined #mojo
14:43 haarg metacpan indexes backpan
14:44 ningu joined #mojo
14:56 AirDisa joined #mojo
15:07 orev joined #mojo
15:18 y1mmm joined #mojo
15:35 dod joined #mojo
15:39 cpan_mojo Mojo-UserAgent-Mockable-1.42 by POPEFELIX https://metacpan.org/release/POPEFELIX/Mojo-UserAgent-Mockable-1.42
15:56 cuechan_ joined #mojo
16:04 cuechan joined #mojo
16:19 sri btw. here's an idea for better systemd support https://gist.github.com/anonymous/d9a34d7d41c03886e5a5bb58e4b66dbd
16:20 sri re what Grinnz mentioned yesterday http://irclog.perlgeek.de/mojo/2016-07-28#i_12929879
16:23 dod joined #mojo
16:25 sri imagine zero downtime software upgrades with systemd might be interesting for some
16:27 ningu yeah. so how would you do the upgrade, use sytstemctl reload?
16:28 ningu is that update going to be in an upcoming release?
16:29 sri no, it's untested so far
16:31 ningu ok, well, I look forward to seeing a tested version :)
16:31 ningu I haven't tried hypnotoad with systemd yet
16:31 ningu currently using supervisord which is kind of lame because you can't do the hot upgrades
16:31 * pink_mist is real happy his distro of choice isn't using systemd
16:34 Kripton joined #mojo
16:38 kamran joined #mojo
16:44 yuni How do I run minion worker as background process? "./app.pl minion worker" seems running always as foreground process.
16:46 Kripton joined #mojo
16:46 jberger yuni: either start from init system somehow or else append an &
16:46 jberger I use init scripts for my production minions but I just start backgrounded in my dev env most times
16:49 yuni thanks for the info
17:09 yuni it works well with an &. i was confused - it would die when i logged out. thanks jberger
17:14 melo joined #mojo
17:21 mpapec I'm doing something wrong,
17:21 mpapec https://gist.github.com/mpapec/9688b87ecf089bba1d050b0d9926b47a#file-gistfile1-txt-L21
17:22 mpapec jberger: can you see why second ioloop step doesn't execute?
17:22 genio mpapec: http://mojolicious.org/perldoc/Mojolicious/Plugin/DefaultHelpers#delay
17:23 genio The default delay helper and examples there can probably help you find it quickly.  the helper makes it so you don't have to remember to ->render_later()
17:24 pink_mist mpapec: use $c->delay() instead of Mojo::IOLoop->delay()
17:24 mpapec hm, why singleton?
17:25 mpapec I've thought that ioloops can be nested
17:25 genio take a look at the document I pointed to.  It glosses over the benefits pretty clearly, I think.
17:25 mpapec tnx genio, I'll check
17:25 pink_mist mpapec: your other delay will be reaped by the gc too soon I believe
17:25 pink_mist mpapec: either way it's easier to just use $c->delay
17:26 mpapec pink_mist: when doing nonblocking redis, I don't have such problems
17:26 mpapec only with user agent
17:26 pink_mist I don't believe your code to be the same
17:28 mpapec sure, I'm using websockets, but for simplicity sake I've made shorter test
17:30 Grinnz_ sri: still same issue :/
17:30 sri Grinnz_: Oo
17:39 Grinnz_ it still seems to be created the pidfile after the old manager stops (in the hypnotoad log)
17:39 Grinnz_ let me make sure i'm not doing anything stupid...
17:39 sri well, the logic should work, but odds are i made a typo
17:40 sri hmm
17:41 sri i suppose a small change could be made to avoid timing issues
17:42 sri https://gist.github.com/anonymous/0e0fc3b5a24cc1c286b506ef3941e597
17:42 dvinciguerra joined #mojo
17:45 mpapec no changes with $c->delay,
17:45 sri things get weird now if the upgrade times out though
17:46 mpapec second step is a no go, https://gist.github.com/mpapec/9688b87ecf089bba1d050b0d9926b47a#file-gistfile1-txt-L22
17:46 sri mpapec: why don't you just start with the examples in the docs if you have problems?
17:47 mpapec sri: didn't found mojo lite with useragent example :/
17:47 sri i mean, it's literally right in the tutorial http://mojolicious.org/perldoc/Mojolicious/Guides/Tutorial#User-agent
17:47 mpapec sec
17:48 Grinnz_ sri: i was doing something dumb, it works from the first gist, though a timing issue is probably possible
17:49 sri Grinnz_: ah, the first shouldn't have timing issues :)
17:49 Grinnz_ if the other process finishes exiting before the pidfile is unlinked and recreated
17:50 sri the pid file would still be there though
17:50 * sri shrugs
17:50 sri guess i would have to know how systemd works
17:51 Grinnz_ in order for systemd to track the new process, the pidfile has to be updated before the old process exits
17:51 sri ok, then the logic is not good enough
17:51 mpapec sri++
17:52 mpapec ok, so Mojo::UserAgent->new is out and only $c->ua should be used
17:52 PryMar56 joined #mojo
17:56 mpapec pink_mist: this is working version https://gist.github.com/mpapec/9688b87ecf089bba1d050b0d9926b47a
17:56 mpapec tnx to all
17:57 ivi joined #mojo
17:58 genio mpapec: Again, you're forgetting a render_later.  use the delay helper rather than Mojo::IOLoop->delay()
17:59 yuni i guess $c->delay calls render_later for you?
17:59 mpapec genio: you're right but there was a bigger problem with user agent
18:00 sri Grinnz_: this should work https://gist.github.com/anonymous/87e6d7bee2be1a4175741307cb97bdb4
18:00 genio yuni: Yes.  http://mojolicious.org/perldoc/Mojolicious/Plugin/DefaultHelpers#delay   shows the ->delay() helper version vs the longer way with Mojo::IOLoop->delay()
18:01 sri it moves all the pid file logic to the original master
18:01 genio sri: There seems to be an issue in the delay helper documentation's example.  http://mojolicious.org/perldoc/Mojolicious/Plugin/DefaultHelpers#delay  The longer version example kind of bounces around
18:02 sri genio: wat?
18:03 yuni genio: thanks. i should remove redundant render_later in my code.
18:03 genio sri: It appears as if there're two examples in one kind of.  Not really delineated clearly
18:04 genio Apologies for not wording things well.  I'm somewhat scatterbrained at the moment
18:05 sri genio: you make no sense
18:05 sri this is what i see blob:http://imgur.com/429c82c1-2f04-4454-aeb1-1c8e06f9e175
18:05 sri whoops
18:05 sri http://i.imgur.com/mWc2lqy.png
18:09 genio yea, just showing both ways in the same example block.  Maybe I'm slower than normal today or something, but some extra comments explaining long vs helpered short way would help there?
18:10 pink_mist I think I agree with genio; it's not clear enough that the longer version is actually longer since the sub {...} bits are shortened :P
18:10 pink_mist compared to the non blocking request example that is
18:11 jberger mpapec: why are you calling back into your own app?
18:11 sri Grinnz_: i've opened https://github.com/kraih/mojo/issues/983
18:12 sri in case there is more interest in full systemd support
18:12 genio sri: Apologies again for using the word issue as I htink that sparked the thought that something was wrong there when in my mind it was just not absolutely clear at a glance.
18:12 sri genio: sorry, you're too confusing for me
18:13 genio sri: No worries. I'm shutting up and getting back to work.
18:14 mpapec jberger: it's only example for user agent
18:16 mpapec I've found out that my own Mojo::UserAgent->new didn't play well with ioloop
18:16 mpapec unlike $c->ua
18:16 sri mpapec: that's not true
18:16 jberger calling into your own application is always going to be complex
18:16 jberger there are lots of problems to consider
18:16 sri mpapec: yours died because $ua went out of scope
18:16 jberger or there's that ^^
18:16 sri it's just perl garbage collectiona t work
18:17 tchaves joined #mojo
18:17 mpapec oh, so it did silent GC?
18:18 mpapec I mean,no log/warning
18:18 mpapec so, if I declare my $ua out of the sub it should not be GC-ed
18:18 pink_mist why would there be any of that? you specifically didn't store it anywhere
18:20 mpapec ok, I'll try to make my own ua helper
18:24 sri hmm, so basically all big linux distributions are using systemd now?
18:24 mpapec so it was GC biting me for n-th time :)
18:24 preaction from what i understand, they basically have to
18:25 mpapec why is that?
18:25 mpapec centos started using it with 7
18:26 preaction though systemd defines a bunch of APIs, they also are currently the only implementation of them, which means they've wormed their way into a lot of parts of the OS, like the window manager
18:27 preaction like, GNOME requires systemd
18:27 sri oh my
18:27 preaction (or, to be pedantic, they require something that conforms to the systemd-logind API, which currently only one thing does)
18:28 mpapec is there problem with API or just systemd?
18:29 sri ok, that makes #983 a bit of a priority i guess
18:30 preaction the problem, from my opinion, is that systemd is doing an unprecedented grab for control over the whole of the OS. rather than the standard Unix philosophy of "do one thing well" and "small things working together", systemd is a monolith that extends its tendrils into basically everything the OS is doing
18:31 sri and our first 8.0 feature perhaps ;p
18:31 sri "full systemd support"
18:31 preaction the very fact that "systemd support" needs to be a thing that a web framework(!!) has to implement is _mind-boggling_
18:32 sri i don't blame systemd in this case though
18:32 sri the pid file vanishing for a short time is a bit of a problem to handle
18:32 preaction lennart doesn't have linus's philosophy on development: don't break user-space...
18:33 mpapec so there is a room for another player implementing same API?
18:33 preaction and yes, technically #983 is hypnotoad could do that a bit better, but it's systemd's arrogance as to "correctness" which is revealing this very tiny problem
18:34 preaction mpapec: in the case of GNOME, yes. well volunteered! i look forward to hearing about your progress
18:35 mpapec preaction: I've meant third player; personally not into this kind of stuff :)
18:35 preaction mpapec: right. that's the problem. that's why systemd is required.
18:36 preaction if you want to live without systemd, parts of the linux ecosystem are being closed off from you
18:36 mpapec you mean like GNOME?
18:36 preaction yes
18:36 mpapec I'm running redhat/centos and don't care about that
18:38 mpapec (ssh/nginx/redis/postgres is what matters to me)
18:41 mpapec time to move to containers and away from distribution?
18:41 mpapec distributions*
18:47 preaction time to move back to BSDs :p
18:48 dvinciguerra joined #mojo
18:49 Grinnz_ preaction: while I disagree with lennart on lots of things, systemd is only monolithic in idea; it's made up of a lot of very individual parts
18:49 Grinnz_ also, personally I don't care what gnome does :P
18:49 Grinnz_ they lost my interest as soon as gnome 3 came out
18:50 Grinnz_ i help maintain the cinnamon spin for fedora specifically to have a better option than gnome for myself :P
18:54 Grinnz_ sri: will try the patch in a bit, things coming up at work :P
18:55 sri preaction: well, macOS is a BSD ;p
18:56 sri coolo: now i'm curious what the most popular laptop at suse might be
18:56 neilhwat1on joined #mojo
18:57 sri from what i hear at redhat they still love high end thinkpads
19:00 tchaves joined #mojo
19:19 sri mishanti1: do you know yet when your designer will have some time for minion?
19:19 Grinnz_ sri: looks good with patch
19:20 sri \o/
19:20 Grinnz_ if i'm interpreting correctly, you made it so when SIGQUIT is sent to the old master, the old master then unlinks and adds the new pidfile itself?
19:21 sri correct
19:21 Grinnz_ cool
19:25 Grinnz_ here is the systemd unit file, pretty straightforward https://paste.fedoraproject.org/397442/69820309/
19:26 sri now that systemd is everywhere, i wonder if we should have a systemd unit file in the cookbook
19:26 Grinnz_ thats a lite app that i was using to test that uses a pg database
19:29 itaipu joined #mojo
19:56 sri mishanti1: asking so i can put some time aside to try and make a minimal minion ui app before
19:56 ksmadsen Hi. Apropos hot deployment: Are there any way to detect if the hot deployment actually worked or not?
19:57 sri it's in the log
19:59 Grinnz_ if it didn't work, the pid won't change, also
20:00 ksmadsen In our case the logs are syslogged to another machine, and not stored locally, so we can't watch the logs from our init-script.
20:01 ksmadsen Watching the pid file is what we do now. However it leaves the question: How long should you wait before declaring the hot deployment a failure?
20:01 Grinnz_ i guess you'd just have to base it on the upgrade timeout, which defaults to 60 seconds
20:01 Grinnz_ https://metacpan.org/pod/Mojo::Server::Hypnotoad#upgrade_timeout
20:04 neilhwatson joined #mojo
20:09 ksmadsen Would it be possible to have hypnotoad create a pidfile for the hotdeploy child? When that file disappears, the deployment went wrong of the main pid file hasn't changed.
20:12 ksmadsen The real reason I'm interested in this is because we also use hypnotoad in our dev-environment (our codebase is too big for morbo). We then have a grunt-like script using inotify to watch for changes and restart hypnotoad.
20:13 sri too big for morbo? Oo
20:13 Grinnz_ you mean for morbo to watch all the files? maybe an option would be to subclass Mojo::Server::Morbo with an inotify version of modify_files
20:14 Grinnz_ modified_files*
20:15 Grinnz_ otherwise... not sure what codebase too big for morbo means :P
20:17 ksmadsen Exactly. It ends up spending all the time watching for file changes.
20:17 sri ah
20:17 tchaves joined #mojo
20:18 ksmadsen Before we switch to hypnotoad, we worked around the issue by having morbo watch a directory in /tmp, and the having our script touch a file there when it wanted a restart.
20:19 ksmadsen s/switch/switched/
20:19 sri yea, i'd subclass Mojo::Server::Morbo for inotify support, like Grinnz said
20:19 Grinnz_ morbo and hypnotoad work differently enough that it tends to be easier to get them to do what you want, than to shoehorn morbo into prod or hypnotoad into dev
20:19 sri ksmadsen: that workaround still sounds way easier than your hypnotoad hack ;p
20:20 dvinciguerra joined #mojo
20:21 sri i mean, hypnotoad pid file handling could totally be rewritten to use two pid files for upgrades, but it's unlikely you'll find someone who would want to work on that and get it release ready
20:22 sri and when i say unlikely, i mean yea... it's not gonna happen ;p
20:23 ksmadsen Hmm... it might be worth to revisit morbo... I can't remember if we had other reasons. Other than the "then dev and prod are as similar as possible".
20:23 sugar joined #mojo
20:23 sri morbo is already very very similar to hypnotoad
20:24 sri just a different process management strategy
20:25 jabberwok joined #mojo
20:33 Kripton joined #mojo
20:44 ribasushi joined #mojo
20:50 melo joined #mojo
20:51 go|dfish joined #mojo
21:07 cpan_mojo Mojolicious-Plugin-XML-LX-1.0 by RSHADOW https://metacpan.org/release/RSHADOW/Mojolicious-Plugin-XML-LX-1.0
21:17 good_news_everyon joined #mojo
21:17 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vKhbJ
21:17 good_news_everyon mojo/master 2547129 Sebastian Riedel: improve support for systemd
21:17 good_news_everyon left #mojo
21:17 sri is this a reasonable default systemd unit file https://github.com/kraih/mojo/commit/25471294ef714e2116dd2bbe98de4893285c1121#diff-b659d9bfeffa6b78a1f2950c7e72259bR151
21:17 sri ?
21:19 Grinnz_ i think After=network.target and WantedBy=multi-user.target, the default boot target doesn't really figure into running a webserver
21:19 Grinnz_ otherwise looks good
21:20 good_news_everyon joined #mojo
21:20 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vKhb4
21:20 good_news_everyon mojo/master 80bf7f8 Sebastian Riedel: better unit configuration file
21:20 good_news_everyon left #mojo
21:20 sri Grinnz++
21:22 Grinnz_ here is nginx's on centos 7 for comparison http://paste.fedoraproject.org/397480/98273461
21:23 Grinnz_ i suppose ExecReload could also be a kill with USR2
21:24 sri i like calling the script more
21:24 sri one more layer of protection
21:25 Grinnz_ sure, and it already won't call it if it sees that it's not running, anyway
21:25 Grinnz_ (which is why it was important for that to work)
22:22 daemonic joined #mojo
22:43 sugar joined #mojo
22:49 mishanti1 sri: Sorry for the slow response. Been afk. The designer has some time off int the coming two weeks, but we can pace things out so we'll be able to contribute when you've started focusing on the functionality.
22:50 mishanti1 "some time off" eq "notworking on our application, but is at work"
22:50 Vitrifur joined #mojo
22:50 mishanti1 I should have said "is available". :)
23:41 gizmomathboy joined #mojo
23:58 sugar joined #mojo
23:59 gizmomathboy joined #mojo

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