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

IRC log for #mojo, 2014-07-20

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

All times shown according to UTC.

Time Nick Message
00:41 condor66 joined #mojo
01:36 mattastrophe joined #mojo
01:43 D4RK-PH0ENiX joined #mojo
01:52 klapperl_ joined #mojo
02:01 D4RK-PH0ENiX joined #mojo
02:23 hengaini joined #mojo
02:34 hengaini I want to develop a database-heavy web application ( not web sites), and I want to use SPA. What about SPA in mojo?
02:35 preaction what is spa?
02:35 purl hmmm... spa is doing a good job against piracy I think.  They seem to be gaining momentum or Single Page Application or at http://www.spa-info.be/
02:36 preaction single-page application then? what about it in mojo? mojo will work just fine. i've done a few already
02:37 hengaini oo.  Is this belong to web client programm such as dojo toolkit not mojo?
02:37 preaction i've used angular, but probably, considering most of mojo's involvement is going to be web services calls or websockets
02:40 hengaini Is angular  also a js framework? Why do you choose this? and not jquery or dojo toolkit?
02:41 preaction jquery isn't much of a framework, it's just a library. it's not scaffolding to build an app, it's convenience functions. i've not used dojo, but as far as i know it's similar to jquery in that respect
02:42 preaction i chose angular because YUI and ExtJS are too big and take over too much of the application
02:42 preaction i don't remember why i chose angular over ember and knockout, but i think i didn't go with ember because it was too much like sproutcore (i believe was its predecessor)
02:44 hengaini thanks! Do you have a blog
02:45 preaction i do, though i'm not sure any of this information is on it
02:46 preaction ah, looks like my "tabbed interfaces with angular" is in my drafts section, because long story
02:46 hengaini How can I find your blog
02:47 preaction http://preaction.github.io (developed using my own static blogging thing, http://github.com/preaction/Statocles)
02:47 preaction some fire-suppression system bastards have really SEO's my name, unfortunately. makes me want to give them my twitter back when they asked for it
02:48 hengaini :)
02:48 preaction so, perhaps the blog is an attempt to SEO them back... maybe...
02:52 hengaini I will read your blog, see you later.
02:56 bayashi joined #mojo
02:59 neyasov_____ joined #mojo
03:03 basic6_ joined #mojo
03:21 jnbek joined #mojo
03:51 jnbek joined #mojo
03:57 sujithm joined #mojo
04:14 Lingo joined #mojo
04:52 KCL_ joined #mojo
06:18 Lingo joined #mojo
06:38 _WildPikachu_ joined #mojo
06:43 dod joined #mojo
08:09 dotan joined #mojo
08:24 Lingo joined #mojo
08:26 sujithm joined #mojo
08:33 Vandal joined #mojo
08:53 condor66 hi all, I'm using steps in two delays and I'd like to run subs from delay2 only when all the ioloops from delay1 finished, for some reason delay2 gets executed before I have all the results from ioloop back: http://pastebin.com/KevpTTaJ
09:06 sujithm joined #mojo
09:07 mire joined #mojo
09:08 firnsy condor66: http://mojolicio.us/perldoc/Mojo/IOLoop/Delay#SYNOPSIS
09:08 firnsy if you're sequentialising steps, why not keep them to the same delay?
09:27 basiliscos joined #mojo
09:59 dpetrov_ joined #mojo
10:19 condor66 I've tried that first, didn't work
10:19 condor66 I'll reread that again
10:43 denis_boyun joined #mojo
10:51 jberger_ condor66: why are you passing delay objects?
10:52 jberger_ That is rarely what you mean
10:52 condor66 jberger_: just wanted to ask whether that I'm calling sub and delay from within it makes a difference: http://pastebin.com/ASVUHLjT
10:53 condor66 get_orderb, get_bal, get_openord are subs that loop certain amouf of times on their own delay_*->begin
10:54 condor66 I have 3 diff delays, to have some clarity when processing results from UA requests, is that possible to sequence couple of delays together at all?
11:00 kyshtynbai joined #mojo
11:01 denis_boyun joined #mojo
11:21 jberger_ You sequence them by calling a dummy callback as the last step
11:25 jberger_ Here is an example app
11:26 jberger_ https://github.com/jberger/MojoForum/blob/master/lib/MojoForum/Helpers.pm
12:26 D4RK-PH0ENiX joined #mojo
12:45 basic6_ i want to schedule a method (sub) to be run every few minutes in the background in my mojolicious application, is there a cron-like feature in mojolicious?
13:08 Lingo joined #mojo
13:29 D4RK-PH0ENiX joined #mojo
13:33 Lingo joined #mojo
13:35 dpetrov_ joined #mojo
13:51 dod joined #mojo
13:58 denis_boyun joined #mojo
14:14 basic6_ i'm trying to remove a route in the before-dispatch hook but the content behind that route is still being served. however adding routes seems to work fine. i'm using $route->remove, is that wrong?
14:40 jberger_ basic6_: the router is fixed after the first request is processed
14:41 jberger_ Use a bridge to disable the route instead
14:41 jberger_ Also, you might want to check out Minion
14:43 basic6_ jberger_: oh so $route->remove() doesn't work after the first request, that should be mentioned in the documentation. using a bridge to disable the route will work fine for me, thanks for the tip
14:44 basic6_ this means that dynamically added routes will accumulate (for the runtime of the mojo server) and can only be hidden, not removed
14:45 basic6_ Minion Job Queue that sounds nice, i might be able to use it, i'll check it out. thanks
14:54 sri people complaining about lack of documentation for something without actually checking beforehand is something that bugs me so much, from now on everyone who does it goes on a list and will not receive any help from me in the future
15:01 basic6_ sri: sorry i should assume this behavior is already documented but i can't seem to find where. http://mojolicio.us/perldoc/Mojolicious/Routes/Route#remove only says "Remove route from parent" and i thought if removing is not supported after the first request, it should be mentioned in that section. i'm going thought that page again right now, looking for hints
15:01 sri http://mojolicio.us/perldoc/Mojolicious/Guides/Routing#Rearranging_routes
15:02 basic6_ sri: thanks that's what i was looking for
15:02 sri and that was the last help you get from me
15:05 basic6_ i'm sorry if i pissed you off by not finding this part of the documentation. i thought i was doing something good, suggesting an improvement (of the already great) documentation but i was wrong. i will try to refrain from suggesting changes to the mojo docs
15:06 sri i have no problem with "i was unable to find that in the documentation"... but "that should be documented" as a way of fishing for answers of deflect blame is not acceptable to me
15:06 sri there's been too many cases recently... and it's burning me out
15:07 basic6_ i see my wording was bad. i wasn't fishing for answers since jberger_ had already answered my question at that point.
15:09 basic6_ but if there are many cases of people simply compaining about the docs, i can understand that it can get annoying after some time
15:11 Lingo joined #mojo
15:17 sri i might even have removed the limitation on changing routes at runtime if this had gone a little more constructive
15:20 basic6_ sri: i would like to know how i could have made my question more constructive. i don't know the inner details of mojolicious, so other than asking about the implications of the removal of this limiation, i can't really provide technical suggestions on how to remove it at this point.
15:22 sri a) you could have just said that you were unable to find the appropriate documentation, b) you could present a use case for changing routes at runtime and make a feature request
15:29 basic6_ a) well i thought http://mojolicio.us/perldoc/Mojolicious/Routes/Route#remove is the appropriate documentation but i failed to check other parts of the docs that could mention the limitation i'm running into. b) in my case, there's an admin interfaces that allows admins to create (very limited) and remove dynamic pages (like /events/event1, where /events is a fixed root). adding an event page should work (and it does) but removing an event
15:29 basic6_ page (event obsolete) should also be possible. if this qualifies for a feature req, i'll make one
15:29 purl Sorry, I don't know (event's email address.
15:31 basic6_ so i'd want to be able to create a page for "event2" (/events/event2) but remove "event1" (/events/event1) (different name)
15:33 sri the limitation is not on ->remove... it works just fine... the caches which are in an entirely different part of the framework just don't get cleared
15:34 sri features that span over multiple parts of the framework can only be explained in the guides
15:37 basic6_ oh i didnt realize there is a global route cache (?) that still holds the route info even after i've removed it. so this sounds like what i would want to do is clear this cache after removing a route...
15:38 sri you can't
15:39 basic6_ of course this explains why the limitation is not mentioned at the remove method, as it affects many different methods, not just remove. that's some interesting insight
15:42 basic6_ i'm assuming the main reason for this route cache is a performance (serving requests using a cache should be faster than re-calculating routes everytime). but i'm wondering if other parts of the framework might fail if the cache is suddenly cleared after at least 1 request has been served.
15:42 sri there are many questions that need to be answered... so if you really want to move ahead with the feature request i suggest you find another core dev to champion your proposal
15:45 sri should you succeed, i actually have working code
15:46 basic6_ ok, i'm not entirely sure if there is anyone else who might like to be able to clear this cache (maybe i'm the only one) but maybe i will create a feature request on github.
15:47 sri http://mojolicio.us/perldoc/Mojolicious/Guides/Contributing#Feature_requests
15:47 basic6_ sri: thanks for the insight about the route cache
15:51 basic6_ "Please do not open GitHub issues for feature requests, ... write the code yourself" ah ok so i won't create a feature request now because i don't know the mojolicious internals to be able to write that code myself. but since you say you do, maybe i can discuss this with another dev here on this channel some time
15:52 sh4 joined #mojo
15:55 PerlMonkey joined #mojo
16:11 sri yes, normally i would now ask all the important questions and talk the feature through with you... but since i want to discourage that kind of documentation complaint from earlier...
16:12 btyler joined #mojo
16:14 basiliscos joined #mojo
16:18 sri soooo... everyone, please be nice in this channel... thoughtless complaints do have consequences
16:18 PerlMonkey lol
16:19 PerlMonkey wut
16:19 basic6_ well you have certainly discouraged me from suggesting any changes to the mojo docs (until i one day forget that i did).
16:19 sri that's not my intention at all
16:20 basic6_ PerlMonkey: as it turns out, i insulted the mojo documentation by saying "that should be mentioned in the documentation". it is already mentioned, i was jsut looking in the wrong place
16:20 sri all i'm asking for is you being nice about it
16:20 PerlMonkey I don't see how that's not nice sri..
16:20 PerlMonkey He's being polite and nice about it
16:21 basic6_ PerlMonkey: basically i want to remove routes, didn't know they're cached and since this is a global cache, it's not mentioned in the routes method but in the general guide.
16:21 sri if you want to go there...
16:21 sri "oh so $route->remove() doesn't work after the first request, that should be mentioned in the documentation."
16:21 sri that is thoughtless deflection of blame
16:21 PerlMonkey He was not being rude though.
16:22 sri which happened multiple times in the last few weeks
16:22 PerlMonkey He simply stated that it didn't work after after the first request
16:22 basic6_ exactly thats what i said. and i wasn't careful with my wording, it came across as a complaint (although it was meant as suggestion).
16:22 sri and every time burns me out a little more
16:22 PerlMonkey if you seriously take that as a complaint, well you wouldn't survive in the Java world.
16:23 PerlMonkey "You use a 2D array instead of a hashmap? WHAT KIND OF A IDIOT ARE YOU, ARE YOU 12?! MORON!"
16:23 sri i'm setting the bar a little higher here, deal with it
16:23 basic6_ i can imagine a developer like sri would receive a ton of requests and many of them probably stupid complaints, so i can understand that it can get frusrating
16:23 PerlMonkey Welcome to Java
16:23 sri this is not java, if you're not being nice you don't belong here
16:23 PerlMonkey He wasn't being mean though
16:24 sri and with that we are ending the argument
16:24 PerlMonkey See how easy that was?
16:24 PerlMonkey I wish Java arguments ended this way :(
16:24 basic6_ i dont want to offend sri (or anyone else really) so i will try to be nicer (esp. concering the mojo docs) in the future. i dont want no trouble.
16:25 btyler PerlMonkey: people have a right to protect themselves from burnout -- little things can add up, even if any individual incident looks pretty innocuous. pretty sure I'm also guilty of accidentally saying "could X be documented" rather than "have I missed this?"
16:25 btyler anyways, woo for maturity
16:26 PerlMonkey btyler, people do, but if you're going to call anything you don't like a complaint then you're burning yourself out
16:26 PerlMonkey It's like saying that when someone doesn't agree with your opinion they're opressing you.
16:26 PerlMonkey And denying your rights to a opinion
16:27 thowe joined #mojo
16:27 btyler eh, personally I don't see a problem with it if there's a clear standard expressed
16:28 btyler but this is unconstructive meta discussion imo. back to making stuff with mojo :)
16:28 thowe I'm reading this blog post: Non-blocking Mojolicious apps are even easier now!  And, I'm hoping someone can explain to me what's going on here...
16:28 PerlMonkey I'm not even sure what Mojo is
16:28 PerlMonkey I joined this channel as it sounded interesting
16:29 thowe I've been casually programming Perl for a couple of years now, and I've made a couple of Catalyst apps.  Reading about Mojo gets me excited, but it stuff like this makes me think I'm never going to understand Perl well enough to use it correctly.
16:30 PerlMonkey oh a real time web framework in perl
16:30 basic6_ PerlMonkey: Mojolicious is a great (well i really like it) web framework in perl. take a look at http://mojolicio.us/ and also check out some example applications. it lets you write complicated and powerful webapps in almost no time
16:30 PerlMonkey I like it
16:30 PerlMonkey Well I like the idea
16:31 thowe Why does this leak memory?  I don't really understand the debug output, either...
16:31 thowe http://blogs.perl.org/users/joel_berger/2014/07/non-blocking-mojolicious-apps-are-even-easier-now.html
16:31 jkg I'm pretty impressed, I looked at mojo properly yesterday for the first time, and now I have a (admittedly fairly trivial) web app which Just Works and was super-easy to write.
16:31 PerlMonkey sri, so you're the lead developer right?
16:31 sri basic6_: i do recognize that you've been pretty nice in our conversation, and i think i made my point, so i think we can just let it go now
16:32 basic6_ PerlMonkey: it has a thorough documentation and i've learned that if you can't find something in its documentation that should be there, it probably *is* there but youre looking at the wrong part of the documentation
16:32 PerlMonkey sri, so you
16:32 basic6_ sri: thanks very much i appreciate it
16:32 PerlMonkey you're the lead developer, correct?
16:33 sri i am
16:33 PerlMonkey sri, I think I have a use for your web framework.
16:34 PerlMonkey I hate how chromebooks have a lack of a proper IDE and giving PHP unrestricted access to a shell is a horrible idea. I could see me using your framework to make a web ide
16:34 sri basic6_: allright.... there are a couple of problems with your use case
16:34 sri basic6_: you mentioned dynamic routes configured through a web admin... what about preforking?
16:35 btyler PerlMonkey: like this: https://metacpan.org/pod/Farabi ? (also made with mojo)
16:35 PerlMonkey let me take a look
16:35 PerlMonkey Sorta
16:35 PerlMonkey but for multiple different languages
16:36 PerlMonkey I'm actively working on a C kernel, my end goal is to have it interpret lua
16:36 PerlMonkey I'm thinking about adding perl as well.
16:38 basic6_ sri: yes, the dynamic routes are configured in the web admin gui. i have to say i haven't thought about the preforking aspect at all yet. at first, i created those routes at the end of the startup method (never updated them). now, i'm updating the routes in the before_dispatch hook. i'm thinking this should work for any number of forks?
16:38 sri are your routes stored in a database?
16:39 basic6_ yes
16:40 sri http://pastie.org/9407405 # this partial patch allows routes to be changed at any time
16:41 sri problems are a) i don't like the method name unfreeze, and b) i'm not convinced yet there is much value in allowing routes to be changed at runtime
16:42 sri every now and then someone pops up asking for it... but they never present a compelling use case
16:43 * thowe wonders when jberger hangs out in here
16:44 go|dfish thowe: he was here a little earlier
16:44 go|dfish well, he's here now, but he was active a couple of hours ago
16:46 thowe I need an ELI5 on that post of his...  Hopefully he has a few minutes.
16:46 basic6_ i'm looking at this patch right now (not knowing any other mojolicious source code). as i understand it, the unfreeze sub will allow routes to be changed again by clearing (resetting) the cache if there is one (there probably isn't one before the first request)...
16:47 PerlMonkey basic6_, do you know C?
16:47 basic6_ PerlMonkey: my C is a bit rusty but yes i do
16:47 PerlMonkey basic6_, want to help me with my kernel?
16:48 basic6_ sri: my first though for a name was clear_routes_cache but thats just a clumsy name. maybe i'll come up with an alternative name
16:49 basic6_ PerlMonkey: you can send me a private message, let
16:49 basic6_ lets talk privately rather than getting way off-topic here
16:51 basic6_ sri: as for my use case... so at this point i'm still dynamically adding routes but since i can't remove obsolete routes, i've added a simple check to the controller method which looks if this route should still be active or not. if not, it does return $self->render_not_found
16:51 PerlMonkey basic6_, done
16:52 KCL joined #mojo
16:53 basic6_ so obsolete routes accumulate the londer the application is running (until restart). now while i would rather remove routes that arent needed anymore, i have to say that the number of routes that are removed isn't so high that i would run into any kind of memory shortage (duh)
16:54 basic6_ which means the workaround i am using now (which just "hides" routes) will probably work for me. in my specific case i can even add that i don't need to change such a route to point to a different controller method. i just remove it or re-enable it later  by adding it again
16:56 basic6_ so i am thinkiing that my workaround probablly wont cause any issues for me even if the app server is running for months without  a restart. if this is correct, my conclusion is that it would still be nice to be able to "properly remove" routes, but at least in my specific case i can live without it
16:57 PerlMonkey basic6_, I'm just going to message you a few notes
16:59 basic6_ the only thing is if these dynamic routes point to different controller methods (or supply different stash values) (which is not what's happening in my case, just to be clear) - then i'm wondering if this would require to actually restart the server since the workaround bridge doesnt know these page-specific values and the controller-target
17:01 basic6_ again this last thing is just a theoretical statement. to sum it up, for my case it would be easier (no additional check needed) and also cleaner (no dead routes lying around) to be able to clear the cache, but a workaround is possible. and at this moment i'm not sure if there is a different use case where this workaround wouldnt help, like i said the changing of routes thing might be one but i'm not sure
17:16 sri thowe: did you read all parts of that series?
17:17 thowe As I was going through there, I noticed that there seemed to be more.  I haven't yet, but if that is meant to give me the proper context, I will do that.
17:18 thowe I need to work up a basic app that intelligently connects to my DBIC stuff and has bcrypt authentication.  That would be a good starting template for half of the things I need to do for work...
17:19 basiliscos joined #mojo
17:20 denis_boyun joined #mojo
17:20 thowe jberger's MojoExample app seems to be kept up to date, that dude helping me a lot
17:22 * thowe really needs to figure out something like Distzilla
17:26 thowe The trouble I have shifting my head between lite and not-lite syntax makes me doubt my Perl-foo
17:38 thowe can you recurse through template chunks?  For example, to build embedded unordered lists.
17:38 thowe embedded isn't the right word...  but a tree
17:41 thowe so, call a block within itself, I guess is what I'm asking....
17:50 sri basic6_: your use case doesn't really convince me
17:50 sri but i suppose testing might be another one... like being able to shuffle around routes to test stuff
17:57 bowtie joined #mojo
17:58 basic6_ sri: i understand, my case is probably not a reason to clear the route cache. testing sounds like it could be one. if an application adds/removes/changes routes dynamically, its tests should also create a route, check if the content behind it looks ok, remove it and check if calling it leads to a 404 (this would fail now)
18:01 basic6_ but would it be a bad decision to provide the clear-cache feature anyway? sure, clearing the cache will result in bad performance. but if that's all, why not offer a "use-at-your-own-risk" kind of advanced feature?
18:03 mr-foobar joined #mojo
18:05 sri http://pastie.org/9407524 # mostly a full patch
18:05 * sri pokes basic6_, jberger, tempire, marcus and crab
18:06 denis_boyun_ joined #mojo
18:10 denis_boyun joined #mojo
18:16 basic6_ i'm still here but if the other devs are busy i can come back some other day
18:21 basic6_ maybe flush instead of unfreeze/release
18:23 sri hmm, not bad
18:25 sri http://pastie.org/9407540 # updated
18:26 thowe So, either I can't call a block from within a block, or I don't know how.  https://gist.github.com/thowe/21ecffa08c203248a40e   How does one make a recursive template chuck?
18:26 thowe er chunk
18:32 basic6_ sri: i like it. are there other methods that would also have to call the flush method in the same way remove does in line 107?
18:32 BinGOs joined #mojo
18:33 thowe or do I need to use TT for something like that?
18:36 sri basic6_: no, or i would have changed those
18:37 BinGOs joined #mojo
18:40 basic6_ sri: ok, so this patch looks perfect to me. it would be great if this would be available in a future version. i'll come back here the next few days to see if theres an update on this from you or one of the other coredevs
18:40 BinGOs joined #mojo
18:52 mire joined #mojo
19:16 denis_boyun joined #mojo
19:20 good_news_everyon joined #mojo
19:20 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/leuqfg
19:20 good_news_everyon mojo/master 4b9fc41 Sebastian Riedel: allow routes to be rearranged at runtime
19:20 good_news_everyon left #mojo
19:23 denis_boyun_ joined #mojo
19:27 thowe sri, That example for $walk in development.html.ep was exacty what I needed once I figured out what it was doing.  Honestly, I shouldn't be that generous...  I figured out how to copy it, I don't claim full comprehension.
19:32 sugar joined #mojo
19:33 denis_boyun joined #mojo
19:36 good_news_everyon joined #mojo
19:36 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/2TPgtA
19:36 good_news_everyon mojo/master 40a3368 Sebastian Riedel: custom names have a higher precedence
19:36 good_news_everyon left #mojo
19:39 thowe updated my gist with working code and some html to make it obvious (if anyone was interested): https://gist.github.com/thowe/21ecffa08c203248a40e
19:42 sri ah right, passing the block reference as first argument to avoid memory leaks
19:44 lipizzan joined #mojo
19:46 thowe Hmm.  OK.  I didn't realize you could get that dangerous.
19:57 denis_boyun joined #mojo
20:11 jberger_ Actually MojoExample is tempire's
20:12 jberger_ sri i think that the patch is good
20:12 jberger_ I just all have never needed it in practice
20:13 jberger_ s/all//
20:13 jberger_ usually when someone thinks they Bex
20:13 jberger_ Grrrr
20:14 good_news_everyon joined #mojo
20:14 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/rnziBA
20:14 good_news_everyon mojo/master 1bca6bb Sebastian Riedel: small optimizations
20:14 good_news_everyon left #mojo
20:14 jberger_ Thinks they need a dynamic router, they really just need placeholders and maybe a bridge
20:15 sri yea
20:15 sri i'm also a little worried it might encourage bad design
20:17 jberger_ It's not a usually style of the doc, but you can probably say something like "if you think you need this, you probably are doing it wrong"  😋
20:17 jberger_ My swipe type is being especially bad today
20:17 purl joined #mojo
20:17 jberger_ I'm in bright sun
20:18 jberger_ My wife just finished a half ironman
20:18 * jberger_ sets purl on fire
20:18 * purl basks in the warm glow HEY WAIT A MINUTE
20:20 jberger_ thowe thought
20:21 jberger_ Though it is and advanced topic, you could use the __sub__ token to recurse
20:21 jberger_ Need perl 5.16
20:22 jberger_ Or else make a helper which renders to a string, which can recuse like any normal method
20:23 meshl joined #mojo
20:23 * jberger_ is proud he remembered that it is not called partial rendering anymore
20:26 thowe What I have in that example feels very clean and discoverable (it must be if even I can find it given only a file name).
20:28 jberger_ Do you you about the sub token?
20:30 thowe No.  It is safe to assume that I don't know just about everything.
20:30 thowe I'm willing to play with it
20:30 thowe can you give me a link?  What is is a "token" of?
20:32 meshl joined #mojo
20:32 marcus I know about everything \o/
20:33 jberger_ http://modernperlbooks.com/mt/2012/05/the-current-sub-in-perl-516.html
20:33 marcus Everything is awesome when you're part of a team
20:33 * jberger_ hums
20:35 * preaction looks for his pants
20:35 jberger_ Hahahaha
20:35 purl LOLCON 4 reached.
20:36 thowe Oh, it's not a Mojo thing, but a Perl thing.  OK.  I keep meaning to buy the latest of that book and give my 2012 version to a friend.
20:40 thowe It is already pure hubris that I attempt to play with all these functional-y things as it is; I ought to avoid hurting myself...
20:40 thowe But that looks exactly like what I was thinking to begin with.
20:40 thowe I'm a danger to myself and others.
20:44 thowe As an aside....  I can't help thinking, as someone who has played with Catalyst, that Mojolicious is sort of the "next level" in programming foo from Catalyst.  It's like that was OO and kind of complex and but grokable by someone who has played with OOP a bit, and Mojo is all in the realm of scheme and Haskell and the deep parts of javascript.  But my viewpoint is probably hopelessly ignorant.
20:53 thowe When I write something, and nothing else happens, I always imagine I said something that made an entire room suddenly go quite while everyone stares at me in disgust.
20:53 thowe I don't understnand how these memory leaks are happening in these anon functions.
20:59 mishantil thowe: Do you do any messing around with the reference counters?
21:09 berov joined #mojo
22:29 punter joined #mojo
22:35 WildPIkachu_ joined #mojo
22:49 dvinciguerra joined #mojo
22:58 jberger thowe: the truth is, few people do understand memory leaks and that's why it has been nice to improve things to where you almost don't have to know
22:59 jberger In my blog post, i don't expect people to understand that debug output
23:00 jberger The point was that doing it right was hard, and so we made it easier
23:00 dwierenga joined #mojo
23:00 jberger 😊
23:02 meshl joined #mojo
23:20 thowe jberger, I didn't even know they could be an issue until I read this stuff this weekend.
23:48 jberger ah, but now they are not, so go about your weekend :-P
23:49 jberger off topic, does anyone know of a good interface to the keyword api?
23:49 jberger I want to try to implement a pythonic "with" keyword, but I sure as heck am not doing it in XS

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