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

IRC log for #mojo, 2015-08-03

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

All times shown according to UTC.

Time Nick Message
00:09 Zoffix Huh? Mojo::Reactor::EV: I/O watcher failed: Bad arg length for Socket::unpack_sockaddr_in, length is 28, should be 16 at /home/zoffix/perl5/perlbrew/perls/perl-5.20.2/lib/5.20.2/x86_64-linux/Socket.pm line 830.
00:09 chhikv joined #mojo
00:12 Zoffix Weird. This is the code I'm getting that error from https://github.com/XTaTIK/XTaTIK.org
00:55 meshl joined #mojo
01:04 Grinnz Zoffix, some kinda ipv6/ipv4 mismatch? best guess
01:30 jontaylor_ joined #mojo
01:43 mattastrophe joined #mojo
01:45 jontaylor_ joined #mojo
02:46 noganex_ joined #mojo
02:51 kaare joined #mojo
03:02 davido___ joined #mojo
03:49 davido___ joined #mojo
04:12 buu BAH
04:44 melo1 joined #mojo
04:47 chhikv joined #mojo
04:52 buu joined #mojo
05:59 dod joined #mojo
06:11 dod joined #mojo
06:15 dod joined #mojo
06:33 trone joined #mojo
07:08 AndrewIsh joined #mojo
07:10 Vandal joined #mojo
07:16 eseyman joined #mojo
07:16 jontaylor_ joined #mojo
07:18 marcusr BUU
07:48 buu_ joined #mojo
07:56 odc joined #mojo
08:03 jontaylor joined #mojo
08:11 vytas joined #mojo
08:44 berov joined #mojo
08:52 damaya joined #mojo
09:10 vytas joined #mojo
09:23 vytas joined #mojo
09:30 dotan joined #mojo
09:44 meshl joined #mojo
10:32 damaya joined #mojo
10:33 mattastrophe joined #mojo
10:50 dp_ joined #mojo
11:11 ribasushi re earlier conversation on why I encourage people to use the RT interface
11:11 ribasushi Grinnz: because it has a practical zero-barrier to entry
11:13 ribasushi to combat the wider issue of people being unwilling to invest considerable time, which is understandable given the general attitude towards regressions "Oh - this is now declared wrong, using things in an outdated way is bad, and you should feel bad!"
11:13 ribasushi more here https://twitter.com/ribasushi/status/628149968017223680
11:24 nicomen RT has like the highest barrier of any reporting system I have ever seen, other than that it has a zero-barrier
11:29 eseyman does it? I wish all FLOSS projects allowed you to report bugs by mail like RT does
11:29 hummeleBop joined #mojo
11:30 nicomen eseyman: that requires you to a) know that there is a mail interface, 2) how the pattern of the mail is, so in the end is not as easy
11:30 nicomen it was new to me, and I've been programming perl since 4.x
11:30 nicomen (saw the lightning talk mentioned)
11:31 nicomen but once you are looking at the web interface of RT
11:31 nicomen oh man
11:31 nicomen should just say with large warnings sign <a href="mailto:rt-Foo-Bar@rt.cpan.org">Report bug here</a>
11:32 eseyman nicomen: once you read the main page, you're all set
11:34 nicomen you mean the main page of "thank you for signing out" you get?
11:35 eseyman I'm looking at https://rt.cpan.org/
11:35 nicomen you never end up at that page
11:36 nicomen rt is most of the time a link from cpan
11:36 nicomen for most people
11:36 eseyman ah
11:36 nicomen why does that page have more than one sentence anyway?
11:36 eseyman I would have thought it would be a mention in a distribution's README or perldoc
11:37 nicomen imagine http://facebook.com having a start page like that
11:38 eseyman well, reporting a bug involves more commitment than using FB
11:38 eseyman different audiences, different main pages
11:39 nicomen well, other systems are easier
11:39 nicomen for instance, github
11:39 nicomen or sourceforge
11:39 Zoffix joined #mojo
11:39 nicomen to compare another system with 90s architecture and design
11:40 Lee some users don't like GH due to the TOS and the fact they have to register (can't interact with just e-mail). i don't have this problem, but it's something to consider
11:40 nicomen I guess my point is, before complaining that users are bad at reporting, fix everything else so it actually has a low barrier, then we can talk again
11:40 eseyman ugh. Now I remember using sf's bug tracker and how painful it was
11:42 nicomen Lee: the benefit of being able to interact with RT with just email without logging in is as good as non-existing given its invisibility on the pages users are sent to when reporting bugs
11:43 Lee yes, i prefer GH over RT. other's don't (which i guess is ribasushi's original point)
11:44 eseyman I suspect's everybody's entry point is "https://rt.cpan.org/Public/Bug/Report.html?Queue=<MODULE>"
11:44 mattastrophe joined #mojo
11:45 eseyman which contains "The quickest way to report a bug in <MODULE> is by sending email to bug-<MODULE> [at] rt.cpan.org."
11:45 ribasushi nicomen: noted, and tldr-ed: https://twitter.com/ribasushi/status/628169595283750912
11:45 ribasushi the point isn't "I prefer to use X as an author"
11:45 nicomen ribasushi: I think it's great that you create awareness
11:45 nicomen but I think it should be part of the initial page on RT
11:46 ribasushi the point is "what's the cheapest way for a drive-by "it doesn't work anymore" reports"
11:46 ribasushi nicomen: that's a very good point (front page of RT)
11:46 ribasushi and something to emphasize that no account is necessary
11:46 ribasushi in order to participate
11:47 Zoffix What could be causing this? Mojo::Reactor::EV: I/O watcher failed: Bad arg length for Socket::unpack_sockaddr_in, length is 28, should be 16 at /home/zoffix/perl5/perlbrew/perls/perl-5.20.2/lib/site_perl/5.20.2/x86_64-linux/Socket.pm line 831.
11:47 nicomen hm, actually when you eventually find the "report a new bug" button written in <small> at the top left
11:47 meshl joined #mojo
11:47 nicomen (https://rt.cpan.org/Public/Dist/Display.html?Name=Moose)
11:47 Zoffix It worked fine before, I just made a minor change and tried to restart the site. This is the code: https://github.com/XTaTIK/XTaTIK.org
11:47 nicomen ther penultimate paragraph contains this: The quickest way to report a bug in Moose is by sending email to bug-Moose [at] rt.cpan.org.
11:47 Zoffix Grinn z yesterday said it could be IP6 vs IP4, which I don't know what that even means...
11:48 eseyman nicomen: what are you doing at https://rt.cpan.org/Public/Dist/Display.html?Name=Moose
11:48 nicomen *top left = top right
11:48 eseyman I can't imagine why you would go there to file a bug report
11:49 nicomen eseyman: clicked "View / report bugs" here: http://search.cpan.org/~ether/Moose-2.1600/
11:49 Zoffix ribasushi, the problem with that tweet is that some dists use GH not RT so those RT emailed tickets will just rot
11:50 eseyman why do that instead of what Moose's documentation tells you to do : "Please report any bugs to bug-moose@rt.cpan.org, or through the web interface at http://rt.cpan.org. You can also submit a TODO test as a pull request at https://github.com/moose/Moose"
11:50 ribasushi Zoffix: yes, this will happen in some select cases. Tthat's not the fault of the user, this is the fault of the maintainer. But even that is beyond the point - look at the big picture
11:51 ribasushi I mean hell - some people will even receive an opinionated autoresponder from Schmorp
11:51 nicomen eseyman: because luckily I still behave like most humans, and reach for what I find first
11:51 ribasushi does this mean the entire mechanism is null-and-void because of a few bad actors?
11:52 nicomen lol
11:52 Zoffix ribasushi, it does for people who don't use RT :P
11:52 Lee RT will e-mail you new tickets though
11:53 ribasushi I just told these people "well, fuck you" on twitter in a more civil manner, and hence consider that aspect of the problem closed for me ;)
11:53 Zoffix lol
11:53 Lee so at least you can act on them (migrate them to GH :)
11:53 ribasushi Lee: bingo
11:53 ribasushi Lee: the only thing not possible via the email interface is the closing/metadata edits
11:54 eseyman ah, search.cpan.org
11:54 nicomen eseyman: googled Moose
11:54 nicomen I know now that on page 9 in perldoc Moose what you said is stated
11:55 eseyman googling Moose gives me the wikipedia page on Moose (the animal)
11:56 eseyman that said, I think everybody is in agreement that search.cpan.org could be improved
11:56 Zoffix Any idea on my issue? If I want long enough, it eventually stops with an asset pack timeout error: http://fpaste.scsys.co.uk/496480
11:56 Zoffix Even though I can fetch that file just fine with wget
11:57 Zoffix And the app starts fine on my local box
11:57 Zoffix I don'yt
11:58 Zoffix eseyman, I don't. That improvement already happened and it's called metacpan.org
11:58 Zoffix The only better improvement would be to remove SCO from search results
11:59 eseyman I don't see how metacpan.org separates viewing tickets from reporting a new one
12:01 Zoffix Doesn't need to. It just points to the Issue tracker, whether it is RT or GH (dunno what else it supports) and it's the Issue tracker's responsibility to provide the UI for creation of new tickets
12:04 eseyman typically, viewing and writing bug reports happen on different pages
12:05 eseyman nicomen's problem is that he's taken to the viewing bug page and that he then has to search for the creating page (and this is non-obvious to him on search.cpan.org)
12:06 * eseyman is looking at a github issues page and has no idea how to create an entry
12:06 nicomen eseyman: I have lots of problems all the time ;)
12:06 eseyman ah, "New issue"
12:07 nicomen but I've heard people complaining about poeple not submitting bugs throughout my time, and most bts are hard for normal persons to report in
12:07 Zoffix Ah, c'mon. We're talking about programmers here, not your grandma.
12:07 eseyman dear $ENTITY, how can anybody use this website to get work done ?
12:07 mudler joined #mojo
12:07 Zoffix If you can't figure out how to use an arbitrary bug tracker to report a bug, you've got bigger problems to worry.
12:07 Zoffix That's my stance on it.
12:08 nicomen Zoffix: in my work life I wanted journalists, moderators and other to report bugs. I have made simpler interfaces like for instance mail gateways for trac, or simplified forms
12:08 nicomen but RT was a monster I never was able to do anything with
12:08 Zoffix I hate RT with vehemence.
12:41 chhikv joined #mojo
12:59 chhikv joined #mojo
13:01 mattastrophe joined #mojo
13:12 chhikv2 joined #mojo
13:17 zivester joined #mojo
13:46 gryphon joined #mojo
14:10 eitz joined #mojo
14:16 Grinnz <ribasushi> and something to emphasize that no account is necessary
14:16 Grinnz while true, it's also "no account is possible" unless you want to sign up for PAUSE
14:16 Grinnz which is the problem i had with reporting bugs via RT; i didn't *want* to communicate by email
14:16 Grinnz i wanted a web interface
14:17 Grinnz but no solution will please everyone... the important thing is RT makes the alternate bugtrackers as visible as possible, which it does
14:18 ribasushi Grinnz: this is a valid desire on your part, which however puts you entirely outside of my target audience
14:18 Grinnz does it? i'm talking about when i was just using perl modules and found bugs in them
14:19 ribasushi basically... I tend to consider IT professionals who "don't want to use email" as invisible
14:20 ribasushi better for me and better for them
14:20 ribasushi (this has zilch to do with perl itself, it's much broader)
14:21 PopeFelix joined #mojo
14:21 Grinnz alright, i'll not entertain this further
14:21 jberger <ribasushi>: "yes, this (bug report rot in RT) will happen in some select cases. Tthat's not the fault of the user, this is the fault of the maintainer" <--- this is unfortunately what makes Schmorp's rediculous stance valid
14:21 jberger he wants to actually opt his modules out, but cannot
14:22 jberger he doesn't want to be responsible for that scenario
14:22 Grinnz his stance is valid, his reaction is not
14:22 jberger I can't really blame him
14:22 jberger Grinnz: of course
14:22 ribasushi jberger: this is all true, and is acceptable collateral damage
14:23 Grinnz and the situation with RT won't change for various reasons, so his only courses of action are to get over it, keep what he's doing, or stop submitting cpan modules
14:23 stephen joined #mojo
14:24 jberger ribasushi: says yoyu
14:24 jberger I don't particularly want that burden either
14:24 jberger I have dozens of modules (as many of us here do)
14:24 * jberger ponders an idea
14:25 ribasushi jberger: well - sue me ;)
14:25 jberger is there a dashboard in RT per-author which can show all the RT bugs for that author's modules?
14:25 jberger it would probably need opt-in for modules which you aren't the current releaser or some such
14:25 ribasushi yes there is (maybe not default, but composable)
14:26 jberger I wonder how hard a "migrate RT bugs to GH" tool would be
14:26 Grinnz RT allows openID now, so that's slightly better than it used to be, though it still requires signing up for something external which i wouldnt' have done back then
14:26 Grinnz because apparently creating account systems is hard these days
14:27 lluad joined #mojo
14:27 ribasushi jberger: http://www.dagolden.com/index.php/2397/moving-cpan-rt-tickets-to-github-now-improved/
14:27 Grinnz jberger, and maybe something that goes through dists and finds them with external BT listed in metadata and open RT bugs
14:27 Grinnz cool
14:28 Grinnz jberger, https://rt.cpan.org/index.html? "bugs in my distributions" has a list but i have no idea what it's from, it has a ton of bugs that i have nothing to do with
14:29 Grinnz oh, sri made me a maintainer of a bunch of things
14:29 Grinnz heh
14:34 Grinnz i would hate RT a lot less if the navigation of the site was just a little more sensible
14:35 Grinnz it's less intuitive than bugzilla IMO
14:36 eseyman Moving to 4.2.x would be a step up, in that regard
14:37 Grinnz 4.2.x of what? bugzilla?
14:37 eseyman no, RT 4.2.x
14:37 Grinnz oh didnt know there was a version not being used
14:37 eseyman consider Bugzilla 5.0 for the UI overhaul although that's quite new
14:38 Grinnz although, i imagine most people that are happy with RT would shit fury if they made it more usable
14:39 eseyman I seriously doubt that
14:39 Grinnz https://xkcd.com/1172/
14:39 eseyman they probably won't use the new features but I doubt they'll complain that they exist
14:40 Grinnz "usable" to me would be completely changing the layout of navigation and ticket actions
14:41 eseyman which would be completely invisible to people using the email interface
14:41 eseyman but you're right, we'll never get a UI that pleases everybody
14:41 Grinnz right, i don't care about that, since i don't
14:43 Grinnz its not necessarily pleasing everyone that's the problem, but people accepting change, as inevitably some are used to the interface by now
14:45 eseyman It's the same old story : people will tolerate UI change if there's a substantiel improvement to their workflow
14:45 ribasushi this.
14:45 eseyman because they need to get to know the new UI and that take time and effort
14:45 eseyman if the improvement isn't substantiel, it's not worth it
14:46 mst jberger: how do I submit a bug to github via email without accepting their onerous terms of service?>
14:46 mst jberger: answer: I can't, and I can't, so you MUST handle bugs on RT as well, and that's your problem for picking github
14:46 ribasushi mst: you *can't* start a bugreport via email at all
14:46 mst if anybody ever comes up with a system that doesn't have those problems, I'll be happy to try and find a way to get bug-Foo mail for those mirrored there
14:47 ribasushi you can participate later on
14:47 mst ribasushi: I just said that.
14:47 mst ribasushi: which part of "I can't" didn't you understand?
14:47 ribasushi I mean even if you have account, TOS accepted and all
14:47 * mst sighs
14:47 mst that was the exact point I was making
14:47 mst will you stop assuming disagreement when you've just misunderstood
14:47 nicomen err
14:48 mst but in the mean time, sending to bug-Foo and hoping maintainers aren't dicks about it is the least worst option
14:49 mst ribasushi: wonder if we could offer xdg's code as a service
14:49 Grinnz personally, i don't care if someone submits a bug to RT, i'll just reopen it on github (or maybe use that thing you linked to move it)... so thanks for increasing awareness ribasushi
14:49 mst right
14:50 mst I'm wondering if we could make it so authors basically poke through a web UI somewhere and it starts auto-moving for them or something
14:50 ribasushi mst: it's still flawed as (last time I checked) you can't attach a random email to a github issue
14:50 mst ah, of course, so people still need ... bah
14:50 ribasushi so the original RT submitter will be left voiceless
14:51 mst two way sync maybe?
14:51 chhikv joined #mojo
14:51 ribasushi *perhaps* this has changed though, there was a lot of churn this year
14:51 mst doubtful
14:51 mst they really want an account+TOS
14:51 Grinnz show me a bugtracking system that isn't terrible to use like RT and bugzilla, and ill happily switch
14:52 Grinnz also one that's integrated to CPAN and/or my source control
14:52 eseyman re moving bugs, you probably want to read Olivier Berger's paper on Bugtrackers triplification
14:53 eseyman Grinnz: what, no pony?
14:53 Grinnz i don't like livestock
14:54 eseyman seriously though, redmine is the best exemple I've seen of an easy to use bug-tracker
14:55 eseyman but again, no email gateway (last time I checked) and it's a pain to admin
14:55 mst every redmine admin I've ever said says basically "nice UI, wish I didn't have to admin the damn thing"
14:56 Grinnz i mean i understand why some wouldn't want to sign up to github to report issues, but much like ribasushi with people who don't want to interact with bugtracking via email, i don't care that much
14:56 eseyman amen... a status workflow for every project? wtf?
14:56 mst I'd rather have a bug filed on RT than not at all
14:56 Grinnz also true
14:58 CHYC Hi, just a quick word to all the developers of Mojo to say thank you. The codebase is fantastic, the documentation is fantastic, it's all fantastic. You've all made me a very happy web developer :)
15:18 PryMar56 joined #mojo
15:20 Grinnz_ CHYC: it is pretty awesome :) but if you come up with any suggestions we're open to them, we're running out of cool new features!
15:21 Grinnz_ for the RT situation... in summary i'm glad it's there, but i'll avoid using it whenever possible
15:22 CHYC Grinnz_: No suggestions at the moment, but I'll bear that in mind. Thanks.
15:28 chhikv joined #mojo
15:30 jberger CHYC: glad to hear it!
15:31 jberger mst, ribasushi, Grinnz_ : sorry I had a meeting
15:31 jberger I do wish that github had unauthenticated bug submission but it doesn't
15:32 jberger that said when you say "I can't, and I can't, so you MUST handle bugs on RT as well, and that's your problem for picking github"
15:32 jberger I don't see how that's my problem, I am making a choice to limit the avenues for bug submission
15:33 jberger I chose github as opposed to competing services because of its ubiquity
15:33 jberger many of the people who would offer a bug report are already on it
15:33 jberger yes, that's an approximation, but it is one that I chose
15:33 jberger I don't see how "MUST" applies to me there
15:34 jberger authors can and have sent me emails directly to ask questions, report bugs and request features
15:34 jberger how is that different than a bug tracker that I didn't choose and don't check
15:34 jberger ?
15:34 Grinnz_ github won't get unauthenticated bug submission, and i don't blame them. it introduces a whole new level of spam and abuse
15:35 hernan605 joined #mojo
15:35 jberger Grinnz_: I don't question their choice either
15:42 eseyman "basic" user accounts that can report bugs but don't have to agree to the TOS?
15:42 eseyman I'm not holding my breath on that one, though
15:42 disputin joined #mojo
15:42 Grinnz_ eseyman: that's not a whole lot different from allowing arbitrary emails
15:42 jberger eseyman: of course you can make a throw-away account to report a bug
15:43 jberger I'm not sure how scary TOS are if the account is only ever used to open one bug
15:43 eseyman Account Term 7 : "One person or legal entity may not maintain more than one free account."
15:44 Grinnz_ it's not about scary, it's about laying out their reasoning for banning you when you abuse the system
15:44 eseyman note that different people will have different reasons to not agree to the TOS
15:45 eseyman and some will not want to have to maintain yet another (username, password) combo
16:05 jberger eseyman: then I will be sad not to have their contribution, though of course they can email me personally if they choose
16:06 eseyman I'm trying to think of a situation in which emailing the module author shouldn't be considered sufficient and I'm not seeing it
16:07 Grinnz_ sometimes the author doesn't list their email in the module docs and doesn't have their cpan email forwarded properly, but i would consider that "their fault"
16:07 eseyman there's a higher chance of getting dupes but that's mainly a worry for the author, not the reporter
16:07 Grinnz_ you should always list a proper email in the module docs
16:07 jberger "I've decided to send you a bug report and I choose not to use Github. Here are the details. -- sent to email addres X" when X is my email address then everything is ok, when X is "service that I do not monitor" (like RT) they are not going to be happy with the result
16:08 Grinnz_ RT effectively forwards it to your email, though, as long as you have that set in pause..
16:08 Grinnz_ (configuration of pause is another issue, we don't need to get into)
16:13 jberger yes my email is correctly configured in pause
16:21 batman_ joined #mojo
17:09 dod joined #mojo
17:29 mib_1bziie joined #mojo
17:29 mib_1bziie sri: take a look https://github.com/kraih/mojo/issues/823
17:31 Grinnz_ i'm not sure i understand what's being requested. or even what intellJ IDEA is
17:33 preaction it's an IDE
17:34 Grinnz_ where would mojo mention this? idgi
17:34 Grinnz_ on the wiki somewhere perhaps
17:35 Grinnz_ also, how does everyone in here keep losing op
17:35 genio maybe in the installation section?  "in the oddball case that you're using intelliJ, here ya go!"
17:35 Grinnz_ marcus: you're our only hope
17:36 meshl joined #mojo
17:52 eitz joined #mojo
17:55 pink_mist obi wan marcus?
17:59 amon joined #mojo
18:07 ZoffixWork joined #mojo
18:14 chhikv2 joined #mojo
18:15 jontaylor joined #mojo
18:29 jberger marcus: you are the last op left
18:30 jberger perhaps op Bender and a few of us?
18:30 jberger names
18:30 mst Bender: trust jberger
18:31 Bender OK, mst
18:31 jberger I wonder if sri would consider registering the channel with chanserv
18:31 jberger it does work nicely
18:38 hummeleBop joined #mojo
18:42 ribasushi jberger: um... ever considered registering it yourself?
18:42 mst ribasushi: I imagine he has.
18:44 jberger I would rather sri do it, which seems proper given the structure of the mojo team
18:45 jberger I will happily do so if he would delegate that responsibility to me
18:49 ZoffixWork The content_* and text_* methods in Test::Mojo are strangely similar
18:49 ZoffixWork I have two ways to test content, but am still looking for something to test raw HTML markup
18:49 preaction you should be able to pass ownership of a channel off after you've registered it
18:52 jberger preaction: that assumes that sri would want it registered
18:52 preaction of course
18:52 jberger ZoffixWork: nodes vs DOM
18:52 ZoffixWork tx->res is the way for my thing. As for content_* sounds like it's just text_* methods with ':root' as selector :(
18:52 gryphon joined #mojo
18:52 ZoffixWork jberger, I'm not following.
18:53 jberger for example, a comment <!-- --> is a node but not a DOM element
18:53 jberger read the documentation carefully looking for the distinction in operation over nodes or elements
18:55 ZoffixWork Oh, ok. I've made an erroneous assumption that '"text" in Mojo::Message.' meant text-only stuff, stripped of HTML markup
18:57 jberger ZoffixWork: you can do that with $rx->res->dom->all_text I believe
18:59 jberger hmmmm, that doesn't quite do what you want perhaps
18:59 ZoffixWork jberger, I think content_is does what I want.
18:59 ZoffixWork I just thought it only dealt with text stripped of markup, but I do need the markup
19:00 jberger oh, yes sorry i forgot that your original question was about Test::Mojo
19:10 PopeF joined #mojo
19:19 PopeFelix joined #mojo
19:56 bwf joined #mojo
20:18 genio joined #mojo
20:46 ashimema_ joined #mojo
20:47 kaare joined #mojo
21:45 mattastrophe joined #mojo
21:49 meshl joined #mojo
21:52 thowe joined #mojo
22:19 stephen joined #mojo
22:23 thowe Hi, Mojo people.
22:28 hernan605 Hi thowe
23:01 njlg joined #mojo
23:11 mattastrophe joined #mojo
23:45 mattastrophe joined #mojo
23:52 jontaylor joined #mojo

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