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

IRC log for #mojo, 2017-08-16

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

All times shown according to UTC.

Time Nick Message
00:02 haarg please link me to where the aclu talked about the "alt-left" and the "SJWs"
00:04 marty haarg.  The ACLU does not care what labels people use.  They are concerned with protecting peoples rights.
00:06 haarg i'm not debating your right to assemble.  i'm debating your right to not be an nazi apologist.
00:07 marty Ahhh, there is that emotion again.  Then you should really take that up with the ACLU because they have been representing neo-nazis for decades.   And I agree with the ACLU
00:08 haarg please keep explaining how much you love neo-nozis
00:08 marty believe it or not the ACLU represents the KKK also.
00:08 haarg s/no/na/
00:08 marty they also represent imigrant workers, single parents etc...
00:09 marty A true liberal organization.  I really admire them
00:09 mattp_ marty: the first amendment went out the window at some point in the past. I'm not sure when
00:09 mattp_ socially anyway
00:09 marty mattp_  It may be down, but it aint out.   :)
00:11 marty And organizations like the ACLU a pillar of rational thought.  They always (almost) look at the big picture, beyond the individual or group and strive for equality and fairness for everyone.
00:12 CandyAngel I like people being able to express their politic opinions. Makes it much easier to figure out whose opinion I should give little or no weight too :P
00:12 CandyAngel s/to/too/
00:12 haarg how did a reddit bot end up on perl irc?
00:12 CandyAngel Wow, can't even regex
00:13 marty CandyAngel  I am in total agreement.
00:14 marty "the personal attack is the last refuge of the scoundrel who has nothing of substance to say"--Edmund Burke
00:16 Grinnz it's not a good idea to discuss political opinions in a community such as this. evidenced clearly by the last couple hours
00:16 marty Grinnz: Agreed.  :)
00:17 CandyAngel Eh, it's not like everyone here is going to go by the same nickname in.. wherever the appropriate place to discuss it is, so still useful to me!
00:19 CandyAngel It is er.. dangerous though. I've had friends stop being friends with each other because of similar topics >.<
00:19 marty Ya, I guess any discourse which is civil is worth the effort.  Personally, I'll rack it up as a positive exchange of opinions.
00:20 haarg a breakdown in civility :(
00:20 haarg i'm sure that's the problem
00:20 * marty shakes haarg's hand
00:21 haarg no thanks
00:21 CandyAngel It was especially annoying because I have different views to both those friends and I manage to stay friends with both of them :|
00:22 marty CandyAngel: Agreed.  It takes a lot of self esteem to disagree and still be friends.   But it's worth it
00:25 CandyAngel And to some degree, they are good friends. Neither has said that I need to "pick one" to stay friends with.. if that makes sense?
01:31 vicash left #mojo
02:03 noganex joined #mojo
05:35 PopeFelix joined #mojo
05:48 eseyman joined #mojo
05:53 PopeF joined #mojo
06:19 dod joined #mojo
06:25 dod joined #mojo
06:32 trone joined #mojo
06:48 AndrewIsh joined #mojo
07:02 fxbrain joined #mojo
07:07 panitaliemom joined #mojo
07:15 inokenty-w joined #mojo
07:28 ashimema is it silly to work on websocket code using morbo as a development server?
07:29 * ashimema doesn't understand enough of the fine differences between daemon, morbo, hypnotoad etc yet
07:33 ashimema I ask because as soon as I have more than one client connected I get random behaviour
07:36 sri marty: that's a stupid argument, you're a nazi apologist
07:36 petru joined #mojo
07:37 sri being a nazi is not a valid political opinion
07:38 sri it's purely about hate and inciting violence, exactly the same as jihadists
07:40 sri even your 1st amendment right is not absolute, you can't yell "fire!" with a large group of people in a closed room and bear no responsibility when people get hurt
07:41 sri same applies to being a nazi
07:41 sri anyone who thinks fascism is a valid political opinion is a fucking idiot
07:42 sri it's only ever a precursor to violence
07:47 sri CandyAngel: same goes for you, if you're friends with nazis you're part of the problem normalizing hate
07:58 sri Grinnz: this is not a political discussion at all, this is taking a stand against hate groups
08:12 sh14 joined #mojo
08:18 good_news_everyon joined #mojo
08:18 good_news_everyon [mojo] kraih tagged v7.41 at 85e1e70: https://git.io/v7bws
08:18 good_news_everyon left #mojo
08:19 marcus Wish we could drop the political discussion in here.
08:19 marcus (Note: not in any way endorsing or suppporting nazis).
08:19 good_news_everyon joined #mojo
08:19 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v7bw4
08:19 good_news_everyon mojo/master f7a0b66 Sebastian Riedel: bump version
08:19 good_news_everyon left #mojo
08:19 CandyAngel sri: I can be friends with people who have different opinions, not so much with people who want to kill other people :|
08:20 CandyAngel (or more generally, harm other people)
08:36 prg joined #mojo
08:52 rshadow joined #mojo
09:08 rshadow joined #mojo
09:11 aborazmeh joined #mojo
09:56 sh14 joined #mojo
10:06 jasanj joined #mojo
10:08 jasanj batman: hello, i found by lastest Mojolicious 7.41, I can't install Mojo::IOLoop::ReadWriteFork
10:12 jasanj it stucked on t/minion.t, i tried do some debug, in line231 of ReadWriteFork.pm, if ( $chld eq 'DEFAULT' ), by the old version Mojo, it's ture, by the lastest version Mojo, it goes to the else condition, and the EV::child seems not worked on my side, it not catch the child finished, and never call _sigchild()
10:14 CandyAngel jasanj: What version were you on before? Child handling got changed recently to remove a delay for fast-completing jobs
10:14 jasanj batman: i mean lastest Minion7.05 actually, not Mojolicious i tink
10:15 jasanj CandyAngel: use Minion7.01 before, it's OK for me
10:15 jasanj Minion7.05 has this issue
10:16 CandyAngel Can you try 7.02? That's when the handling changed which might have broken ReadWriteFork
10:19 jasanj https://github.com/kraih/minion/commit/b39eaa9db4f165c3dfe81db6ea61d4306d1a4e92
10:20 jasanj this change make $chld not be 'DEFAULT', and caused my issue
10:21 jasanj i wondered why EV::child in 'else condition' not worked
10:24 CandyAngel That commit should just be optimisation, this one should be the "breaking" one: https://github.com/kraih/minion/commit/84d43e3ba3e2442c7e4bc3a769d24b000da4b8d0
10:24 CandyAngel Maybe RWF can be fixed by setting a local signal handler back to DEFAULT?
10:25 frogonia joined #mojo
10:36 batman jasanj: sorry, but i don't have time to look at it today. would very much appreciate an issue/pr based on your findings.
10:56 gregf_ joined #mojo
11:02 tchaves joined #mojo
11:06 Zx3 joined #mojo
11:06 ashimema sri.. can you Get your opinion on https://metacpan.org/source/SRI/Mojo-Pg-4.01/lib/Mojo/Pg/PubSub.pm#L16
11:07 ashimema this allows for the exact same callback to exist multiple times for a channel..
11:08 ashimema should/could this not be better served by a hash keyed on cb?
11:09 ashimema thus dissallowing the duplication of callbacks for a single channel?
11:14 sri ashimema: no, it works the same as Mojo::EventEmitter by design
11:15 ashimema aha.. ok.. so by design
11:15 ashimema so is there a way to ensure the exact same callback isn't added multiple times from a level up then?
11:24 ashimema aha.. so I can always call an unlisten with the callback before calling a listen with the callback to prevent dupes
11:24 ashimema brain back on track.. thanks sri
11:44 petru joined #mojo
12:22 marcus sri: You like Bobiverse?
12:23 sri marcus: my favorite scifi series so far
12:23 sri sadly it looks like book 3 was the end of it
12:24 marcus sri: Guess I should give it another try. Started with the first book as audio book but didn't really get into it.
12:24 marcus It's free on amazon unlimited tho.
12:24 sri got all three as audiobooks
12:25 marcus sometimes I have a hard time focusing in the beginning on audio books
12:26 sri bobiverse is scifi for programmers imo
12:26 marcus Almost finished The Atlantis Gene now tho, really enjoyed that one.
12:26 sri worth another try
12:26 marcus Cool.
12:27 sri at least try it until he gets into space and has to solve problems there
12:27 sri that's what the rest of the books are like
12:29 marcus yeah, I actually got to a part with some space battle, but I'd not paid enough attention and didn't quite get why the other guy wanted to kill him, so I got a bit bored.
12:34 ashimema ack.. nope still stuck
12:34 CandyAngel Is it available as a book? You know, without DRM stuff? Summary of book 1 has gotten my interest :P
12:35 CandyAngel Never mind, why do I prematurely ask questions? >.< (i.e. found it)
12:35 ashimema for each websocket connect or disconnect I'm getting a new callback added for all existing connections.. so for one notification the exact same callbac fires multiple times and thus sends the message multiple times down each connected socket
12:35 ashimema I'm missing something obvious here :(
12:40 ashimema anyone able to take a moment to look at this and point out any mistakes :( https://ssl.thorsen.pm/paste/e29ba71be325
12:41 ashimema I've been bashing at it for a week or so now..
12:48 ashimema and with most of the debug lines removed in case it's any clearer:https://ssl.thorsen.pm/paste/25f0d2652ae7
12:49 mib_d1xqsj joined #mojo
12:53 salva joined #mojo
13:15 PryMar56 joined #mojo
13:26 gryphon joined #mojo
13:48 Pyritic joined #mojo
14:06 dod joined #mojo
14:11 mcsnolte joined #mojo
14:24 jberger ashimema do you actually upgrade the connection?
14:24 ashimema er...?
14:24 ashimema not explicitly.. I thought calling ->on() did that for me
14:24 ashimema ->on(finish in my case?
14:25 jberger well, two things
14:25 dotan_convos joined #mojo
14:25 ashimema thanks for taking the time to look jberger
14:25 jberger first, you don't actually call on, you call once (which I think is actually ok) but ...
14:25 jberger you call it on the transaction, not the controller
14:25 gizmomathboy joined #mojo
14:25 * ashimema listens instently.. but is about to go collect kids from school so may not be able to take much in for another hour I'm afraid
14:25 jberger I don't think the transaction's on method does the upgrade
14:26 ashimema ah.. ok
14:26 jberger anyway, I have no idea if that's the problem
14:26 ashimema I do get the socket upgrade at the browser
14:26 jberger but it is a thing I look for when people have websocket connections
14:26 ashimema so thought that bit was working
14:27 jberger which callback gets fired multiple times?
14:27 ashimema it's the pg pubsubs.. either notifications_cb or updates_cb
14:27 jberger why do you unsubscribe before subscribing?
14:28 ashimema because I thought the issue was that we were somehow registering the callback with the channel multiple times
14:28 ashimema it didn't help so I was going to remove it again.. left it in so one to could what I'd already tried
14:29 jberger even if you were subscibing a callback with this same exact behavior, it would have a different memory address and so you wouldn't remove it original one on a subsequent connect
14:29 ashimema make sense
14:29 * ashimema removes that.. claims stupidity
14:29 jberger nah, its ok
14:29 jberger I mean, remove it, but it makes sense to come to that conclusion
14:30 jberger these are the Hard Problems (TM)
14:30 ashimema :)
14:30 ashimema glad they're the hard problems and it's not just me being entirely stupid ;)
14:30 ashimema right.. I best go get the kids before they think I've abandoned them..
14:30 ashimema any chance I could grab you for a chat/advice some time tomorrow pretty please?
14:31 jberger I'm around often :-P
14:31 ashimema :)
14:31 ashimema we're coming out of conferance season now too
14:31 jberger the thing I'd check next is that pg isn't actually sending out the notification multiple times
14:32 jberger no sense running down multiple firings of a callback if the notification is being sent multiple times
14:33 * ashimema runs to car, will try to check that when i get back.. pretty sure I call the ->notify only once and can see it only once int he logs.. so would be surprised and am not sure how to check at the pg end for it getting multiplied there
14:34 Janos joined #mojo
15:03 bwf joined #mojo
15:37 stomper joined #mojo
15:39 stomper Is there a way to simulate a "click" when Testing. I can't see anything in the docs of Test::Mojo.
15:39 Grinnz no, that's a javascript action, not html
15:40 Grinnz if the click would trigger a form submit or similar, you can do that directly
15:41 Grinnz otherwise, you need a javascript capable test like https://metacpan.org/pod/Test::Mojo::Role::Phantom
15:41 stomper I guess I'm after a selenium/webdriver action.
15:44 stomper Ahh Role::Selenium may be an option
16:04 petru joined #mojo
16:39 sh14 joined #mojo
16:48 jberger one of these days I want to learn about the headless chrome stuff
16:49 jberger sad that the phantom project stalled/stopped
16:49 jberger but I can't blame the guy
18:04 arcanez there's chromelss
18:04 arcanez *chromeless
18:07 itaipu joined #mojo
18:12 CandyAngel jasanj: Did you manage to get a fix working?
18:20 petru why does I18N warn $host if I set supported_hosts?
18:42 trone joined #mojo
18:54 marcus So excited about this https://www.theguardian.com/books/2017/aug/15/good-omens-david-tennant-michael-sheen-neil-gaiman-terry-pratchett-tv-adaptation
19:24 itaipu joined #mojo
19:34 jberger arcanez: chromeless requires node
19:35 arcanez I was on that hype train for a bit.. but I got more into it and don't like it .. and prefer Go
19:35 arcanez (so far)
19:36 jberger I guess I have nothing specifically against it, I don't see the desire to do javascript on the backend, but it gets better with each new JS/ES version
19:36 jberger but, there is no reason to require that for a mojo module
19:36 arcanez oh for sure
19:36 jberger IIUC the chrome headless protocol is just websocket/json
19:56 zivester joined #mojo
20:53 jberger hahahaha, nice job preaction: http://preaction.me/title/
20:55 preaction powered by Mojolicious: https://github.com/preaction/MillennialTitle
21:28 marcus yeah, neat!
21:38 dmanto joined #mojo
21:46 good_news_everyon joined #mojo
21:46 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v7AZY
21:46 good_news_everyon mojo/master 87030e6 Sebastian Riedel: improve Mojo::Base role support with shorthands, so "MyClass->with_roles('MyClass::Role::Foo')" becomes "MyClass->with_roles('+Foo')"
21:46 good_news_everyon left #mojo
21:49 jberger sri++
21:54 good_news_everyon joined #mojo
21:54 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v7Anm
21:54 good_news_everyon mojo/master 823eb0c Sebastian Riedel: use the shorthand in the testing guide
21:54 good_news_everyon left #mojo
22:12 sri i've chosen the +Foo syntax since it's the only one that makes sense to me
22:25 Janos joined #mojo
23:24 nicomen how about "no ::" means relative?
23:24 nicomen you would be very strange to have a role named just MyRole no?
23:25 preaction no
23:25 nicomen I think you mean yes :-P
23:25 preaction Throwable
23:26 nicomen like outside your project, or root namespace Throwable? lib/Foo/…, lib/Throwable.pm ?
23:26 nicomen I would call that strange, although plausible
23:27 preaction as in there is a cpan module right now called Throwable
23:27 nicomen oh you mean _the_ Throwable, gah
23:30 nicomen someone made NG that contains Object, great
23:56 Grinnz the NG distribution is really screwed up

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