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

IRC log for #mojo, 2016-11-06

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

All times shown according to UTC.

Time Nick Message
00:17 jberger joined #mojo
00:27 jberger joined #mojo
01:29 sri omg, the new macbooks get worse and worse
01:30 sri now there are usb-c devices that you can't use with them if you want to use wifi at the same time :O
01:30 sri https://www.youtube.com/watch?v=NYVjIjBMx6o&feature=youtu.be&t=209
01:30 suede no way
01:30 suede i can't believe that
01:30 suede that's obscene
01:31 suede sri: holy cow, they are just GIVING away market share at this point
01:31 sri and it's not even unusual devices... just hdmi dongles and external harddrives :O
02:34 asarch joined #mojo
03:29 jberger Yowza
03:53 noganex joined #mojo
04:09 aborazmeh joined #mojo
05:25 janus joined #mojo
05:31 Paddi joined #mojo
05:55 polettix joined #mojo
06:59 polettix joined #mojo
07:03 Vandal15263 joined #mojo
07:55 CHYC joined #mojo
08:14 rshadow joined #mojo
08:47 Paddi joined #mojo
10:18 dod joined #mojo
10:23 dod joined #mojo
11:00 dod joined #mojo
11:03 dod joined #mojo
11:03 dod joined #mojo
11:15 sh14 joined #mojo
11:15 polettix joined #mojo
11:18 sri i guess this might redeem the new macbook pro :D https://github.com/avatsaev/touchbar_nyancat
11:18 * sri pokes tempire
11:27 Paddi joined #mojo
11:36 dotan_convos joined #mojo
11:50 Atog joined #mojo
11:56 Atog hi, is there anyone around using perl and par::packer on windows that could try to create me a executable file from this? https://github.com/oprietop/mai_rename
12:18 Paddi in mojo lite, how can I delete a cookie which i set from another app (diff subdomain)?
12:21 suede Paddi: doesn't cross-origin apply between subdomains as well? I'm not even sure you can
12:21 suede Can you access the cookie from the other sub?
12:22 Paddi suede: yup i can create and read, but I'm not sure how to delete it
12:22 Paddi actually I use subdomain cookie for SSO... I wanna delete it after successful login.
12:22 suede Paddi: try delete $c->cookies->{whatever}? That's how you get rid of session keys
12:23 Paddi @suede: okie let me try that ... I tried $c->cookies("name_of_cookie", { expires => -1 } ) but doesnt work
12:26 Paddi hmm ... can't delete $c->cookies
12:28 suede hrm, i'm not sure then
12:28 suede sorry man
12:29 Paddi @suede: no worry bro, thanks anyway :)
13:07 pink_mist suede: session keys are parts of one cookie, not cookies in themselves
13:07 pink_mist suede: the way you delete a cookie is by setting its expiration date in the past
13:07 suede pink_mist: I know how session keys work- I assumed cookies were the same way I suppose.
13:08 sri you delete cookies by expiring them explicitly
13:08 suede i mean what are cookies but storing key/value pairs on a browser
13:08 sri oh, pink_mist was faster
13:15 rshadow joined #mojo
13:32 inokenty joined #mojo
13:33 sri Grinnz: sombra looks super fun
14:45 flow joined #mojo
14:46 flow my helper $c->reply->client_error($caught_exception) works all fine. :-)
14:48 flow Considering now to put the relevant code section into my issue-closed reply.
14:49 flow But would anyone read it there?
14:49 lluad joined #mojo
14:59 sri it's always nice to add more information to issues
14:59 sri just in case someone stumbles over the same thing and searches before opening a new issue
15:05 * sri is still not sure if he hates Mojo::SQLite::PubSub
15:07 sri i can see it being used by people testing stuff that was originally meant for Mojo::Pg
15:07 sri but at the same time it's incredibly inefficient and ugly code
15:08 sri think i prefer the Mojo::* database modules to focus on what makes the specific databases great
15:08 sri if i was to fork Mojo::SQLite to clean it up, what would i name it?
15:09 genio Mojo::SQLite3
15:09 genio eww :/
15:10 genio don't fork it. you know the author. work with him to clean it up
15:11 sri i remember talking with Grinnz about it before, and he seems kinda invested in being compatible with Mojo::Pg
15:11 rshadow joined #mojo
15:12 genio and I'm sure you can see value in that.  make improvements
15:12 sri breaking so much just to "clean up the code" is also not nice for the users
15:12 sri there is no way to do pubsub efficiently in sqlite
15:12 sri the feature is just bad for the database
15:13 genio So, what would you change? other than ridding the dist of the ::PubSub feature set?
15:14 sri i mean, there's already differences between Mojo::Pg and Mojo::SQLite, like Mojo::Pg::search_path
15:14 sri the better the modules get, the more database specific optimizations they should get imo
15:15 sri yea, i would get rid of pubsub, in favor of better query performance, and focus a little more on making temp databases easier to use
15:18 sri i'm kinda invested in sqlite now, and will prolly use it much more at work, so i might even consider an official spin-off project
15:20 sri note that i'm not saying Grinnz has done a bad job, quite the opposite in fact
15:20 sri but our goals don't really align
15:27 sri of course i'd also like to hear opinions from others on a Mojo::SQLite-ish spin-off project
15:28 sri btw. really like that SQLite has transactional DDL
15:34 genio hrm. I was looking to see how difficult it'd be to make ::PubSub its own dist that subclasses ::Database to make it possible.
15:35 genio that'a where my mind went. fixup ::SQLite, making ::PubSub an optional additional install.  It'd take me a lot more time to digest the source to see how complicated that'd be
15:36 sri that would still be a breaking change
15:36 genio yes, but at least not a terrible one to overcome.
15:38 sri guess i'll just make a personal fork for myself
15:38 flow *typiditype* ... *submit* ... for those interested: https://github.com/kraih/mojo/issues/1014
15:39 sri and we can talk about official spin-offs when there is actual interest, since i don't want to step on anyones toes
15:46 asarch joined #mojo
15:54 dod joined #mojo
16:02 asan13 joined #mojo
16:16 ajr_ joined #mojo
17:03 PryMar56 joined #mojo
17:49 polettix joined #mojo
17:51 sri *crickets*
18:22 sri anyway, it really bugs me that every time i tell someone about Mojo::SQLite i also have to warn them not to use the pubsub functionality
18:22 sri that's just bad design
18:45 stryx` joined #mojo
18:51 suede sri: sqlite... and pubsub?
19:06 trwww joined #mojo
19:08 sri exactly
19:20 Grinnz sri: i'm definitely open to adding discouraging language (recommending actual pubsub) or removing it with a deprecation cycle, it's not worth keeping around
19:20 Grinnz does it actually adversely affect performance when not being used? i feel like that at least could be fixed
19:22 sri don't know, but as far as i'm concerned it should just not exist
19:42 Adura joined #mojo
19:51 jamesaxl joined #mojo
19:51 jamesaxl hi
19:51 purl bonjour, jamesaxl.
19:58 suede hi jamesaxl
20:03 stryx` joined #mojo
20:18 dp_ joined #mojo
20:19 jberger it would be nice if there were a more pluggable pubsub solution
20:19 jberger that's my $0.02
20:20 jberger I mean I kinda rolled on for Minion::Notifier
20:20 jberger (see Minion::Notifier::Transport and subclasses)
20:20 jberger Mojo::SQLite::PubSub isn't so much pubsub as Null-transport for testing
20:21 jberger that said, Minion::Notifier use's preaction's Mercury as its "null transport for testing"
20:21 suede jberger: it doesn't seem like a pub/sub abstraction would really take much work
20:22 jberger and it is a much more capable null transport (it is a pubsub bus built as a mojo app relay using websockets as the transport)
20:22 suede it'd be what, like 3-4 classes total?
20:22 jberger suede: right, I'm saying I've already done basically that
20:22 jberger but it wasn't intended for general use, just the use of notifying on minion job state movement
20:22 jberger but there's no reason it couldn't be extended slightly to do that
20:23 jberger https://github.com/jberger/Minion-Notifier/tree/master/lib/Minion/Notifier/Transport
20:47 jberger on another topic, with all the recent news about Vue.js (especially given the Vue 2.0 release and the stories about its speed and even hints at a Vue native) I thought I would repost my Vue chat example here: https://gist.github.com/jberger/8d09d4fa19aa54cff2d999c614380da3
20:48 jberger I probably ought to get around to writing a blog post about it
20:48 jberger maybe for perltricks
20:50 preaction or a tutorial / walk-through presentation...
20:50 jberger both are very possible
20:54 sri gotta say vue.js does look quite elegant
20:54 sri never really got into react.js, but that example just makes sense
20:56 sri Grinnz: so, you're deprecating the pubsub stuff in Mojo::SQLite?
20:56 sri then i'll hold off with the fork
20:58 suede sri: React is very very good
20:59 suede It's the best approach I've seen for programmatically generating/modifying HTML uis
21:00 sri sorry, but jsx is some of the most stupid bullshit i've ever encountered in my programming carreer
21:03 suede it's not exactly like vanilla javascript is any better
21:04 preaction but comparing jsx to dom is deliberately ignoring everything else that came between
21:04 sri templates in vue.js make sense to me
21:15 polettix joined #mojo
21:29 mishanti1 jsx is a layer of obfuscation I just don't need. I have a hard enought time grokking those bits as it is. :p
22:03 jberger sri: that's exactly how I feel, Vue components are very readable and sensical while jsx is ... really scary
22:03 jberger the nice thing is that Vue 2.0 can do jsx if you want it to and even per-component so if you are in the rare case where it is useful you could do it and not need to deal with it otherwise
22:04 jberger http://vuejs.org/v2/guide/render-function.html
22:05 suede http://appleinsider.com/articles/16/11/02/phil-schiller-new-macbook-pro-has-more-orders-from-apple-than-any-other-pro-model-ever
22:05 suede wat
22:06 jberger so many people have been putting off getting a new pro
22:06 jberger I know Phil21 is in dire need of a new lappy
22:06 suede i am as well
22:06 jberger and was waiting until the new one was announced
22:07 jberger I'm still very happy with my late 2014
22:09 * jberger reads that jsx example and thinks he would never do that
22:09 jberger I'd stick with the render functions approach if/when basic templates fail (as in their example)
22:11 jberger I thought we were done intermixing code and templates
22:12 jberger and now they are tempting fate by not even having template marks at the boundaries
22:12 * jberger shakes head
22:15 Adurah joined #mojo
22:17 Phil21 if my 2012 15" wasn't crashing I'd have zero interest in this latest piece of trash
22:18 Phil21 just don't want to deal with windows, and linux is horribad on laptops still (battery life)
22:18 Phil21 worst case if I hate this thing, I'll trade it in for a tricked out 2015 rMBP and just wait a few years to upgrade - hopefully windows/linux has caught up by then for dev station
22:19 bpmedley Phil21 : You could get a refurbished older laptop that still has applecare warranty..
22:20 * sri is still very sad he got his 2015 rMBP with only 8gb ram :(
22:21 sri with 16 it would last years
22:23 sri that said, the 2015 13" is a damn fine laptop
22:23 sri i'm so going to miss the keyboard in the future
22:31 polettix joined #mojo
22:51 Grinnz sri: sure. if jberger can come up with a more generic solution that i could mention that would be great too
22:51 Grinnz that's my only reservation, that people are actually using it for testing a Mojo::Pg pubsub app
23:28 sri Grinnz: seems rather unlikely from my experience
23:28 sri in fact, one of the big reasons i can't convince coolo to use Mojo::Pg more is how hard it is to write SQL that works in Mojo::SQLite and Mojo::Pg at the same time ;p
23:29 sri is it just me or does booking.com get more silly as an employer every year?
23:29 sri now they use hackerrank.com as a cornerstone of their hiring process
23:30 * sri sighs in disbelief
23:49 Grinnz opened https://github.com/Grinnz/Mojo-SQLite/issues/8 in case anyone has a comment, though not many are watching that repo

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