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

IRC log for #mojo, 2017-07-18

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

All times shown according to UTC.

Time Nick Message
00:16 genio Craftsmanship: Wow. Awesome. Hopefully they prove me really wrong and keep up the maintenance. Like I said, the product is pretty great to use
00:19 genio arcanez: We use it in production for one app. It's been mostly stable. The usage of the product (API) is plain amazing. My concern was that there wasn't really any activity for a while as far as merging code into master. I don't want to be stuck maintaining an app that uses a DB that's no longer maintained
00:20 genio but, if they're going to stay on top of it, I give it an overall positive review
00:47 arcanez genio: are there perl bindings?
02:43 firnsy joined #mojo
02:59 anparker joined #mojo
03:06 firnsy joined #mojo
03:10 noganex joined #mojo
05:12 zach jberger: What I mean is that I have protected routes that have a logged_in sub and if not logged in you get a 302 to the login instead of a 200, so I'd like the mojo tests to ALSO be able to test a test user logging in and reaching things requiring that, so I'll do what you said and thanks
05:20 inokenty-w joined #mojo
06:05 Lee joined #mojo
06:22 dod joined #mojo
06:28 dso joined #mojo
06:28 dod joined #mojo
06:30 dso Hi, in Minion do I have a chance to reset a job to inactive after I dequeued it? I have some strange constraints that can be only checked at possible run time of the job, in the worker. So the worker gets the job and does some checks. If it determines that the job cannot be executed (now), how can I "give it back"?
06:31 dso I see the retry method. But doesn't this increase the retry counter?
06:37 schelcj joined #mojo
06:39 AndrewIsh joined #mojo
06:52 prg joined #mojo
06:53 Vandal joined #mojo
07:01 dod joined #mojo
07:01 zach is it even worth having an email function all on its own
07:36 trone joined #mojo
07:57 dso okay, retry works fine :-)
07:57 dso left #mojo
08:14 rshadow joined #mojo
08:55 sri i suppose retry in minion should also allow to increase the attempts value
09:12 jonasbn joined #mojo
09:15 jonasbn I was reading the FAQ on inactivity timeout, to me it is not completely clear, whether MOJO_INACTIVITY_TIMEOUT both influences both server and client
09:15 jonasbn can somebody clarify?
09:17 pink_mist it does not
09:17 jonasbn pink_mist: thanks :-)
09:18 CandyAngel How does it not? It's in both UA and Server::Daemon?
09:18 pink_mist oh, actually, never mind, I'm mistaken :P
09:18 jonasbn damn
09:19 * jonasbn has timeout issues
09:19 CandyAngel https://github.com/kraih/mojo/blob/f8a0eaf360718dc4230fd495ec69d73de180c274/lib/Mojo/UserAgent.pm#L21
09:19 CandyAngel https://github.com/kraih/mojo/blob/8c63cdfec3a8821a5ad7a2bb2a43f4b759fb5edf/lib/Mojo/Server/Daemon.pm#L16
09:20 sri how is the documentation unclear about that?
09:20 sri this links to both attributes, which mention it again http://mojolicious.org/perldoc/Mojolicious/Guides/FAQ#What-does-Inactivity-timeout-mean
09:23 ksmadsen joined #mojo
09:24 jonasbn sri: the description can be interpreted both ways
09:24 jonasbn perhaps in could be emphasized: "MOJO_INACTIVITY_TIMEOUT environment variable for both"
09:40 irqq joined #mojo
09:49 dod joined #mojo
10:07 gregf_ joined #mojo
10:24 zerah joined #mojo
11:18 yukikimoto joined #mojo
11:40 tchaves joined #mojo
12:24 irqq joined #mojo
12:47 genio arcanez: Yes. There's a RethinkDB dist out there in CPAN. It's not as fluid as the JS client, but it's usable
12:49 sri using rethinkdb is insane at this point
12:55 Janos joined #mojo
13:07 gizmomathboy joined #mojo
13:12 Pyritic joined #mojo
13:55 gryphon joined #mojo
13:55 Pyritic joined #mojo
14:05 jberger I'm happy to see that there was a community release, but I think I'd wait a bit to see that really establish before starting a new project using it
14:06 jberger if and when that happens I might be interested though
14:06 jberger I was always curious about it
14:09 sri a bit = 1-2 years
14:11 sri i mean, for a fun project you can of course use whatever you like
14:12 sri but investing in it professionally would be madness
14:15 jberger yes, I should have said "new project for $work"
14:15 jberger and yeah, 1-2 years is probably where I'd want to see it
14:16 jberger though it does help that it was an established project so all one really needs to see is that its governance and development work
14:16 sri also no news about basho it seems, guess they might be trying to sell riak
14:17 jberger I really don't know anything about riak
14:17 jberger what's it raison d'etre?
14:17 jberger its*
14:17 sri super high availability
14:21 sri distributed key value store, based on amazon's dynamo paper (ring of nodes and stuff)
14:22 sri and it used http as its main protocol
14:23 sri i guess if you're coming from mongodb it's worth mentioning that it auto scales
14:23 sri you add nodes to the ring, and it rebalances
14:24 sri and you decide on how many nodes your data should be replicated
14:26 sri it's very good to build something like S3 on top of
14:27 sri which is actually what they did, think they wanted to earn money selling an S3 addon for enterprise use
14:31 sri they made some very popular tools for the erlang community in the process and those are now in limbo too
14:32 sri ah https://www.theregister.co.uk/2017/07/13/will_the_last_person_at_basho_get_the_lights_oh_too_late/
14:32 sri looks really bad
14:33 sri hard times ahead for nosql
14:33 sri who would have thought just a few years ago
14:33 jberger yeah
14:33 jberger looked like nosql was on a rampage
14:41 jberger also kinda hard to believe that after all the bad press MongoDB looks like they might be one of the few survivors
14:48 jberger despite all the negative press covfefe
14:49 jberger oh! did everyone see andy serkis reading Trump tweets as Gollum?
14:49 jberger so good
14:50 PryMar56 joined #mojo
14:50 plicease joined #mojo
14:58 sri wasn't that two weeks ago?
14:58 sri like a dozen trump scandals ago...
15:00 spicyjack petition to change the universal time standard from astronomical to one based on trump scandals
15:02 haarg we'll need 256-bit numbers for that
15:02 pink_mist sounds unstable
15:02 spicyjack quantum math
15:17 jberger :D
15:18 CandyAngel We could just all move to Australia and get them to pass a law saying all numbers are 256-bit
15:19 jberger math is no longer the law in Australia
15:19 jberger therefore no numbers are 256-bit
15:30 Grinnz I was watching Arrow and there was a "scandal" where the mayor was under investigation for a coverup and I was just thinking, Trump creates a bigger scandal than this every day
15:32 jberger satire shows and even scandal shows are having a hard time competing
15:37 spicyjack talk shows are doing very well
15:37 spicyjack as is news shows
15:37 spicyjack as are, rather
15:46 sh14 joined #mojo
15:47 Pyritic joined #mojo
15:47 zivester joined #mojo
15:47 Grinnz so i played with postgres fulltext searching a bit. definitely workable, if the features you need are what they implemented. but it's very far behind elasticsearch in flexibility and features
15:48 Grinnz what they do have implemented is mostly way easier to do in elasticsearch, too
15:49 sri elasticsearch is not a database though
15:49 Grinnz sure, not saying it's bad, it's just not an elasticsearch replacement in any sense in my mind
15:49 sri did anyone say it was?
15:50 Grinnz there was that article saying to use postgres instead of x, y, z :P
15:51 sri that's the case for many postgres features, they are often just a 90% solution
15:51 sri and i'm pretty sure that's what the article said
15:52 sri most of the time what postgres provides is enough
15:52 pink_mist http://renesd.blogspot.com/2017/02/is-postgresql-good-enough.html
15:52 Grinnz well that's what i mean... it's good enough (barely) for the small project i was using it on, but for anything i want a real fulltext search in, i don't think that it is
15:53 sri so, you're saying the article is correct :p
15:53 jnbek joined #mojo
15:54 sri and it looks like i missed the vacuum improvements in postgres 9.6 that are supposed to help with job queues
15:54 sri good to know
15:57 jberger I tried using postgres full text search for a multi field typeahead and it isn't quite good enough
15:57 jberger but if I was doing proper text search on text fields (mine were often more numerical, like ips etc) then yeah, I expect it is plenty good for that
15:58 jberger like if I was searching movie titles, actors, descriptions etc it would be just fine IMO
15:59 Grinnz fulltext search systems are always designed for searching these documents of prose, which never seems to be how I end up needing to use it
15:59 Grinnz luckily it's pretty easy to customize how elasticsearch works for that
15:59 Grinnz and sphinxsearch before that was passable
16:00 Grinnz i'm glad i made the jump to elasticsearch when i did though... sphinxsearch seems mostly abandoned
16:11 jberger Grinnz: do you abuse the _all field? that's the way I know how to do multi-field matching
16:11 Grinnz no, i'm using query_string matches and that takes a list of fields
16:12 jberger the next time I implement this I'm going to have a lot of reading to do
16:12 jberger I was like 50 releases ago (more?)
16:12 Grinnz yes, definitely worth reading through each and every filter
16:12 Grinnz different ones will be applicable to different use cases
16:12 Grinnz for mine, i don't really have any option other than query string, because i need infix matching
16:12 Grinnz e.g. wildcard at beginning of terms
16:21 jberger this is what I used to implement my search before: http://blog.qbox.io/multi-field-partial-word-autocomplete-in-elasticsearch-using-ngrams
16:30 Grinnz I haven't used the suggesters feature but it seems to be designed for that use case https://www.elastic.co/guide/en/elasticsearch/reference/current/search-suggesters-completion.html
16:30 Grinnz ah, thats what it's talking about
16:31 madalynn joined #mojo
16:31 madalynn durr
16:32 Grinnz ...k
16:34 sri that reminds me... i've been wondering if we should put "Chat" in the mojobar :)
16:34 jabberwok How ingrained is 'lib' as the name of the application class?  I have written a more customizable replacement for »mojo generate app« but find that Mojo::Home special-cases 'lib' and fails if you put the application and controllers under e.g. 'app' instead
16:34 jabberwok http://www.pastebeest.com/113
16:34 sri leading to the usual mibbit link https://chat.mibbit.com/?channel=%23mojo&server=irc.perl.org
16:34 Grinnz jabberwok: yes, i wrote my own "home" helper that basically reimplements it with a different parent dir
16:35 Grinnz i don't know if there's a good way to make it customizable
16:35 Grinnz but lib is pretty standard in perl
16:36 mib_8d2n6x joined #mojo
16:36 jabberwok right Grinnz - the script always adds Find::Bin + ../lib to @INC, so perhaps if Mojo::Home took "the leaf directory name of the last thing in @INC if it looks like... umm" yeah it's complex
16:37 sri jabberwok: after all the home detection bugs we've had to fight, any changes would have to be absolutely rock solid to be accepted
16:40 jabberwok absolutely, sri.  'lib' aside, would it be useful to have generate/app.pm use potentially-customizable directory and class names ?
16:42 sri doesn't seem very useful to me
16:42 sri never even thought about it
16:43 jabberwok might be handy when developing alongside other projects and trying to keep consistent file structures
16:44 good_news_everyon joined #mojo
16:44 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vQpfY
16:44 good_news_everyon mojo/master bb95f1b Sebastian Riedel: try having a chat link in the menu bar
16:44 good_news_everyon left #mojo
16:46 mib_o38h48 joined #mojo
16:48 good_news_everyon joined #mojo
16:48 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/vQpJo
16:48 good_news_everyon mojo/master 4342838 Sebastian Riedel: looks like the project no longer exists
16:48 good_news_everyon left #mojo
16:49 Pyritic joined #mojo
17:01 dod joined #mojo
17:02 sri hmm, the vscode terminal is so good that i'm now getting too lazy to switch to iterm2 to run my tests
17:14 Pyritic joined #mojo
17:40 jberger Grinnz: neat thanks!
17:41 rshadow joined #mojo
17:43 jberger jabberwok: I see generate as a nice thing for starters but I don't really think of it a generic tool
17:44 jberger commands can be released to CPAN too of course
17:44 jberger if you want to make one that is more generic
17:44 jberger honestly I never use generate when I start a new app, but I might not be the target audience :-P
17:48 gryphon joined #mojo
17:53 sri i use the generators to test stuff quickly
17:53 sri not to start projects
18:00 Grinnz i already have a generator to start projects: dzil new :P
18:00 Grinnz https://metacpan.org/pod/Dist::Zilla::MintingProfile::Author::DBOOK
18:04 sri genio: what's the state of Mojo::Reactor::UV? :)
18:05 genio sri: The base UV::Handle object is doing what it's supposed to do. UV::Loop is also behaving as expected.  Things are testing well with ::Timer, ::Idle, and ::Prepare objects.  So, I think for Reactor::UV all we need is a UV::Poll Handle object. That object is very, very similar to the Timer object and shouldn't be too difficult to implement.
18:06 genio I'm looking at doing that tonight and getting all of the libuv test suite https://github.com/libuv/libuv/tree/v1.x/test for the test-poll-*.c tests working. Once that's good, minor tweaks to Reactor::UV and documentation need to be done.
18:08 genio At this point, every test I've Perl-ified from their test suite is behaving exactly as their tests do. I'm pretty happy with it. Documentation and ironing out a TRIAL release soon
18:08 sri \o/
18:09 genio https://github.com/genio/p5-UV/blob/master/lib/UV/Handle.pod    UV::Handle base class is documented
18:09 pudgetta joined #mojo
18:10 jberger genio++
18:20 dod joined #mojo
18:20 sri looks like Alien::libuv installed fine here
18:20 genio \o/
18:20 sri just pkgconfig was missing and i had to install it
18:21 genio Testing's going pretty well for the small XS dist I made to help test Alien::libuv  http://matrix.cpantesters.org/?dist=UV-Util+0.008
18:21 plicease sri you shouldn't need pkg-config what dist are you on?
18:22 sri tested on macos
18:22 plicease Alien-Build didn't pull in PkgConfig.pm ?
18:22 Pyritic joined #mojo
18:23 sri this was how it failed for me https://gist.github.com/anonymous/ee27bc1fa0f79d9123dfa3967730c146
18:25 sri brew install pkgconfig fixed it
18:25 sri on linux i have the whole toolchain anyway
18:25 plicease yeah it uses real pkg-config if available.
18:25 plicease which version of AB if you don't mind my asking?
18:26 sri 0.66
18:30 perlpilot joined #mojo
18:32 plicease sri: thanks I think I can follow up on this
18:33 sri plicease++
18:43 Pyritic joined #mojo
19:07 dod joined #mojo
19:19 tchaves joined #mojo
19:30 bianca joined #mojo
19:32 chandwki left #mojo
19:35 bianca joined #mojo
20:09 brunoramos joined #mojo
20:53 trone joined #mojo
22:14 qmq joined #mojo
22:45 qmq Hi, I am new in IRC and Mojolicious. In the company that I work I only have Windows servers. I manage to run perl-cgi scripts on IIS7. Now Im trying to move to Mojolicious and but I think with Windows IIS7 will be a dead end. Any server suggestions or Documnetation to run Mojolicious on Windows.
22:47 qmq Thanks in advance.
22:47 Grinnz if you have a server that runs CGI scripts, you can run mojo apps transparently with that
22:47 Grinnz it's not going to be as efficient as running a persistent server, but it works
22:47 Grinnz https://metacpan.org/pod/Mojolicious::Guides::Cookbook#Apache/CGI
22:48 Grinnz and you don't get async or websocket features of course
22:49 qmq ok...I am using it for databases
22:51 qmq maybe that will be enough...I will try that.  Thanks :)
23:40 brunoramos joined #mojo

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