Camelia, the Perl 6 bug

IRC log for #mojo, 2013-06-11

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

All times shown according to UTC.

Time Nick Message
00:23 HtbaaPi joined #mojo
00:33 mattastrophe joined #mojo
00:35 marty Behold, sri is more than just a Supervillain in the Perl universe.....  http://new.livestream.com/artofliving/face2face
00:42 hesperaux_ joined #mojo
00:48 egopro joined #mojo
00:51 shmuel joined #mojo
01:05 whitebook joined #mojo
01:20 hesperaux_ joined #mojo
02:00 d4rkie joined #mojo
02:05 amirite joined #mojo
02:28 arpadszasz joined #mojo
02:42 amirite my sessions don't seem to behave properly
02:43 amirite i set a session variable in a POST request and the next request should reflect that saved data but it's hit or miss
02:43 amirite i don't quite know how best to trace the problem or if i should use some kind of session extension to save session data in something other than cookies or what
02:43 egopro joined #mojo
02:49 inokenty joined #mojo
03:03 preflex_ joined #mojo
03:16 mattastrophe joined #mojo
03:17 whitebook joined #mojo
03:25 zivester joined #mojo
03:34 buu amirite: Can you track the data in the cookies?
03:37 zivester joined #mojo
03:46 amirite no, it's all combobulated
03:50 mattastrophe joined #mojo
04:00 preflex_ joined #mojo
04:11 amirite maybe session handling is broken
04:13 amirite i really don't get it, if i call the request that sets the session variable once it doesn't set
04:13 amirite if i call it twice or three or quadrice it sets
04:13 amirite wtf is this
04:14 popl quadrice?
04:14 amirite i.e. four times
04:14 arpadszasz_ joined #mojo
04:14 popl what's more likely, that session handling is broken or you're just not sure how to do it right?
04:14 popl :)
04:14 amirite session handling is brokenm
04:16 amirite how do i force session to flush
04:21 popl amirite: have you showed any code yet?
04:30 amirite here you go
04:30 amirite http://bpaste.net/show/106085/
04:31 amirite if the ajax action/method gets called once, it doesn't save but if it gets called twice it saves
04:31 amirite it's really ridiculous
04:31 amirite it's as if somewhere in code it says "$self->sessions->flush if $req->num % 2"
04:33 amirite i'll try chrome, maybe firefox is stupid
04:33 whitebook joined #mojo
04:34 amirite nope, it's not a browser issue
04:34 * amirite bangs head against wall
04:34 popl hit it harder, maybe it will jog something loose
04:34 popl ;P
04:40 Meiermann joined #mojo
04:41 popl amirite: where are you calling ajax?
04:43 amirite dynamically on page scroll via javascript
04:44 egopro joined #mojo
04:46 Mike-PerlRecruiter_ joined #mojo
04:54 * sri wonders how long it will take for trashcans that look like macpros to appear on amazon
04:56 arpadszasz_ joined #mojo
05:00 omega sri: http://www.amazon.co.jp/dp/B0018​NRM9U/ref=tsm_1_tw_s_azjp_mo7ilt close but not completly I guess :)
05:01 sri hahaha
05:01 omega (miyagawa retweeted it just now from some other japaneese dude :P)
05:03 amirite i knew it was a bug
05:03 amirite return unless keys %$session || $stash->{'mojo.active_session'};
05:04 amirite session has an extra key, "expires" which makes this bool false on the first call
05:04 popl where's that at amirite?
05:04 amirite then on the second call it matches because the key is deleted
05:04 amirite https://metacpan.org/source/SRI/Mojoli​cious-4.14/lib/Mojolicious/Sessions.pm
05:05 amirite i don't write buggy code
05:05 popl bullshit
05:05 popl with all due respect :)
05:18 basiliscos joined #mojo
05:22 egopro joined #mojo
05:22 sri hmm, perhaps the dots pragma has a chance after all :o
05:23 popl yep
05:25 denisboyun_ joined #mojo
05:29 arpadszasz_ joined #mojo
05:41 hesperaux joined #mojo
05:47 tomi-ru joined #mojo
05:50 Britzel_ joined #mojo
06:02 suy joined #mojo
06:04 arpadszasz_ joined #mojo
06:18 amirite weird, very weird
06:18 amirite if my route is more than one path deep then my session gets erased
06:18 amirite i.e. /foo/bar erases the session
06:18 amirite and /foo doesn't
06:26 amirite fuckit i've had enough of this non-sense i'm going to move to db storage
06:26 amirite attention #mojo mojolicious sessions are fucked up
06:26 mattastrophe joined #mojo
06:26 * amirite adds an additional 10ms to each request because mojo sessions don't work
06:29 popl amirite: at least you're being constructive.
06:33 Vandal joined #mojo
06:33 Mike-PerlRecruiter_ joined #mojo
06:34 yakudza joined #mojo
06:44 mattastrophe joined #mojo
06:59 dod joined #mojo
07:01 mattastrophe joined #mojo
07:18 Akron joined #mojo
07:25 Dandre left #mojo
07:30 popl joined #mojo
07:34 dod joined #mojo
08:04 jzawodn joined #mojo
08:17 batman amirite: is there any redirect involved?
08:18 bpmedley joined #mojo
08:18 batman popl ;)
08:18 * batman had some issues with sessions, but simply because browsers misbehave on 302
08:20 amirite nope, it remains a mystery
08:20 kwa If you can write a test to replicate it, I'm sure the guys would appreciate it.
08:20 amirite however i was able to construct a hackified MojoX::Session::Store that only costs an additional 1 ms to select from the db
08:20 amirite hopefully the ms cost doesn't increase drastically when the session.store table gets a million rows
08:28 dbr joined #mojo
08:39 Dandre joined #mojo
08:44 ghandi|work joined #mojo
08:47 ghandi|work Hi there. I'm trying to run a Mojo-App behind Apache mod_proxy as reverse proxy. But i don't go the HostName-Aproach instead it's located under /myapp. The Trouble: accessing /myapp brings up the Index-Page of the  app. But nothing else. Accsing /myapp/$FOO/index.html delivers the public/index.html file. Without $Foo i get nothing.
08:48 ghandi|work I tried the rewirting from the Cookbok, no change
08:48 ghandi|work Any why i need this extra directory-level?
08:48 ghandi|work *idea
08:56 Dandre Hello,
08:59 Dandre Is there any way to 'mount' a controler to a path or to set a prefix. For instance, if Foo  controler has: get 'bar' => ... I could mount this controler to /foo so that GET /foo/bar would be handled by this router
09:00 Dandre or may be some sort of prefix (ie setPrefix('foo')
09:01 kwa Dandre: Check out bridges.
09:04 Dandre kwa: nice, thanks
09:05 egopro joined #mojo
09:06 fhelmber_ joined #mojo
09:14 arpadszasz_ joined #mojo
09:15 maxhq joined #mojo
09:26 trone joined #mojo
09:33 trone I would add the action elaboration time (from route call to end of rendering) in the page footer. Is there any plugin? mojolicious api to call?
09:51 ghandi|work Ok guys...i found it myself...when deploying under a <Location /myapp> with mod_proxy as reverse proxy you must not add trailing Slashes to the proxy directives...otherwise apache will swallow the first Path-Parts and all your routes go to hell...
09:52 ghandi|work Reading carefully the mod_proxy doc tells you this: If the first argument ends with a trailing /, the second argument should also end with a trailing / and vice versa. Otherwise the resulting requests to the backend may miss some needed slashes and do not deliver the expected results.
09:52 ghandi|work :D
09:53 ghandi|work Maybe this advice is also worth noticing in the Mojolicious Cookbook about deploying?
09:55 popl ghandi|work: did you use the ScriptAlias directive as noted in Mojolicious::Guides::Cookbook?
09:56 ghandi|work I'm not deploying as CGI-Script. I'm using the ReverseProxy Approach. so i don't need ScriptAlias, right?
09:56 popl I don't know. I was just clarifying what you said before. :)
09:58 ghandi|work The "problem" occurs in Apache-Config. If you say "<Location /myapp>" your ProxyPass and ProxyReversePass Directives must not end with a slash. Because your location-directive has no ending slash. If the Locatin-directive has an ending slash, your Proxy* Directives need an ending slash too.
09:58 ghandi|work If you don't respect this little rule, you get messed up routes and missing path-parts :D
10:03 ver joined #mojo
10:27 trone joined #mojo
10:37 whitebook1 joined #mojo
11:16 moltar joined #mojo
11:21 iankent joined #mojo
11:22 iankent hi all, qq - not sure if I've found a bug in session handling
11:23 iankent if I have a session variable which is a hashref, if I do something like "if ($self->session('key')->{subkey} == 1)", if the value in subkey is a string (so we get numerical == warning), the value seems to get set to the numeric eqivalent (i.e., "A123" = 0, "00005" = 5), but its not an assignment, just an equality test
11:24 iankent it also only happens if I chain session('key') and ->{subkey} - if I split them up or use references etc, it doesn't have the same problem
11:26 iankent if I correct it to use 'eq' instead of '==' it works fine, but '==' shouldn't be causing a variable reassignment anyway
11:26 iankent golfed example: http://pastebin.com/8qPLUH6U
11:27 iankent sorry, correct example: http://pastebin.com/KC9BfvaM
11:36 nicolaas joined #mojo
11:56 navi joined #mojo
11:59 denisboyun joined #mojo
12:02 jast joined #mojo
12:07 libsysguy joined #mojo
12:25 ryozi joined #mojo
12:29 nicomen iankent: when I run your golfed example, first /set, then /get, then /redir => /get I get 00500 as the value every time
12:37 mire_ joined #mojo
12:42 asarch joined #mojo
12:43 rem_lex joined #mojo
12:53 iankent nicomen: what version of mojolicious are you running it on?
12:54 iankent and I had to do /set, /get, /redir then another /redir - it breaks on the second (which admittedly is even weirder)
13:05 jberger amirite, I'm not going to step into a firestorm, but I have one question: are you attempting to store a million rows in the session?
13:05 amirite_ joined #mojo
13:06 nicomen 2.98 it seems
13:06 jberger repeating since it seemed possibly necessary:
13:06 jberger amirite, I'm not going to step into a firestorm, but I have one question: are you attempting to store a million rows in the session?
13:07 nicomen iankent: same with 3.89
13:08 nicomen iankent: also after several redirs
13:08 mire__ joined #mojo
13:11 mire joined #mojo
13:11 iankent nicomen: we're on 4.11 or 4.12 and it still happens there
13:11 iankent its just very odd that if you break up the line to separate session from the inner hash value it doesn't happen, only when chained... I know the == is wrong syntax, but it feels wrong that something is breaking expected perl behaviour
13:19 ghandi|work iankent: i could reproduce your problem once...but currently no second time...
13:19 mire joined #mojo
13:20 ghandi|work Ahh no, sorry, can reproduce it alle the time with "==" while "eq" works correct
13:21 ghandi|work Mojolicious 4.12, Perl 5.14.2
13:24 nic iankent: In your pastebin example the "==" coerces both sides to numeric
13:25 nic so '000500' --> 500
13:28 nic ah, no, there's more to it than that; I tried a smaller example and it remained string
13:29 arpadszasz_ joined #mojo
13:30 abstract joined #mojo
13:34 russum joined #mojo
13:34 ghandi|work But perl -e '$ref = { id => "00500" }; if($ref->{id} == '500') { print $ref->{id} . "\n" }' doesn't have this effect? Why on the Session-Object then?
13:37 nic (Thinking aloud) Could it be object->json->object is checking whether it has been used as a numeric?
13:37 alester joined #mojo
13:38 jberger try using Devel::Peek on the result to see if there are differences
13:39 ghandi|work It's indeed a perl->json->perl issue
13:40 ghandi|work http://pastebin.com/cGxu8cpA breaks it for me
13:40 jberger yeah, these have been talked about before (recently even)
13:40 jberger getting that right 100% of the time is something like impossible
13:40 nic ghandi|work: There you go, so if you didn't use Mojo json marshalling you might not spot your code had a bug :)
13:41 ghandi|work It's not my bug. I was just curious for the problem ;)
13:41 jberger oh wait, don't trust Data::Dumper either though
13:41 jberger Data::Dumper tends to stringify things too
13:42 bluescreen joined #mojo
13:42 ghandi|work say $newref->{id} returns 500 too
13:42 mire_ joined #mojo
13:43 ghandi|work But nevertheless, $variable ==  '00500' is somehow wrong anyway :D
13:43 marty The Data::Dumper stringify issue bit me a few weeks back.   Was very perplexing.
13:44 jberger Devel::Peek is the only way (I know of) to really see the current state of a variable
13:44 jberger I think Data::Printer does its best to try too
13:45 nic ghandi|work: Using "==" is a bug in your case... and Mojo helped you find it
13:45 ghandi|work again: It wasn't my bug in the firstplace. iankent had discovered this behavior when missusing ==  ;)
13:47 mire__ joined #mojo
13:47 nic ghandi|work: There's no shame; we all have bugs from time to time
13:47 * marty is covered in bugs atm
13:48 nic marty: Are you in your luxury cabin?
13:52 inokenty joined #mojo
14:04 mire_ joined #mojo
14:11 dod joined #mojo
14:12 btyler joined #mojo
14:17 marty nic:  I wish... would prefer those kinds of bugs atm.  Working from home today.   But this weekend... wooo hooooo.  :Q
14:18 * nic is envious as always
14:23 * jberger too
14:26 * libsysguy prefers to work from the office
14:26 libsysguy I worked from home for like 6 months, it wasn't all its cracked up to be, imo
14:28 jberger libsysguy, we're jealous of his cabin I think :-)
14:28 libsysguy oops
14:29 jberger hehe
14:29 libsysguy well in that case, yes I'll jump on the boat
14:29 * libsysguy is jealous of cabins :p
14:29 * jberger is currently working from home, though as he has no employer, thats kinda a misnomer
14:30 jberger things are looking up though ;-)
14:30 libsysguy If I was working on my own stuff from home I would guess I'd have a very unhappy girlfriend
14:30 libsysguy are you looking for employment or going it on your own?
14:31 jberger I'm looking, though I have a very promising lead
14:32 jberger I just defended my Ph.D. recently, so this hiccup in employment was expected
14:32 libsysguy shoot, I was about to say the company I work for is hiring :p
14:32 jberger I do wish my lead was a little closer to my field of interest
14:32 libsysguy congratulations on your Ph.D.
14:33 jberger thanks
14:33 libsysguy color me jealous
14:33 jberger http://ultrafast.phy.uic.edu/berger_thesis.pdf
14:33 jberger I haven't signed anything yet
14:33 libsysguy heh I was just about to ask for a link to your thesis
14:33 * jberger is quick with that link
14:34 jberger once it gets final approval I'm putting it on my personal site
14:34 jberger libsysguy, my big employment problem is that I'm unlikely to leave Chicago atm
14:34 libsysguy ah, yes well the climate is slightly different here in Houston
14:34 jberger so its either local businesses or ones for which I can work remotely
14:35 jberger its not climate, its mostly home ownership
14:35 jberger we have a home that we would lose quite a bit of money on if we tried to sell now, that plus we are near family and its not a recipe for leaving easily
14:36 libsysguy well I don't think cPanel deals much in complex mathematical modeling for Electron Microscopy
14:37 libsysguy is the market in Chicago not recovering well?
14:38 abstract joined #mojo
14:38 jberger its coming around, my problem is convincing businesses that having someone who knows programming and science would be useful when the employ both engineers and programmers
14:39 jberger worse than that, its finding a human to make that case to
14:39 jberger giant databases of applicants don't understand nuance
14:39 jberger :-)
14:39 libsysguy there isn't much that is worse than the automated application process
14:40 libsysguy actaully, my girlfriend just applied for dental school in Illinois
14:40 libsysguy Midwestern i think
14:41 jberger Midwestern? I haven't heard of that
14:41 jberger Nothwestern?
14:42 libsysguy http://goo.gl/QFhFz
14:42 libsysguy I think its new
14:42 jberger libsysguy, re: automated application, I am learning that lesson the hard way, every actual lead I have is because I knew someone
14:42 btyler jberger++ on virtual yapc::na. I had a great experience too, although I guess I should have looked for #yapc. #perl6 was also a pretty enjoyable co-watching crowd
14:43 btyler also, as a writer of automated application processing tools, apologies for any evil usage of said tools
14:43 libsysguy jberger I've always been told its not what you know, but who you know
14:43 jberger #perl6 was probably the right place for you if you spent most of your time in the perl6 track
14:43 jberger it didn't get discussed much in #yapc
14:44 jberger libsysguy, yes I've heard that too, but I guess I didn't believe it was so true :-)
14:44 libsysguy blames btyler :p
14:44 jberger btyler, its not the automated tool's fault, its the fault of those who rely on it to exclusion
14:45 jberger I'm not the average candidate, I wish there was an easy way for me to indicate that
14:45 jberger but I'm sure when they need to fill a post they enter certain search terms, and in that case I doubt I would ever pop out
14:46 btyler I was entirely luckless until I started specifically hunting for small companies, since my education is not in programming/cs. same issue with keywords, I imagine
14:46 jberger yeah, the automated process is going to help find a specialist, not someone with a broad range of knowledge
14:46 btyler for some reason russian language majors don't score highly on resume searches for web programmers
14:47 jberger hahaha
14:47 jberger well, you're writing the software now, you can add that "feature"
14:47 jberger :D
14:48 ajmrch joined #mojo
14:48 btyler hah, good point. +20 search rank positions for 'liberal arts'
14:48 jberger can you add +100 for Ph.D. in Physics ?? that would be great
14:49 KindOne joined #mojo
14:50 btyler php, though, so it'll be something gross like "preg_match(/something like but not quite perl RE/)". can you stand that being done in the name of physics PhDs?
14:50 jberger libsysguy, then again, rurban liked my talk last year, so you never know :-P
14:50 jberger (re: cPanel)
14:51 jberger and also, Downers Grove, IL isn't all that far from me
14:53 gryphon joined #mojo
14:56 KindOne joined #mojo
14:58 d4rkie joined #mojo
15:03 whitebook joined #mojo
15:12 whitebook joined #mojo
15:16 libsysguy heh jberger well if you decide to apply just put me down as your refer-er :p
15:17 libsysguy referer*
15:18 KindOne joined #mojo
15:24 whitebook joined #mojo
15:26 btyler joined #mojo
15:34 tmirchev joined #mojo
15:43 sri amirite_: be nice, final warning
15:45 KindOne joined #mojo
15:50 garu jberger, marty: is this what you're talking about? https://github.com/garu/Data-Printer/pull/48
15:51 KindTwo joined #mojo
15:52 jberger garu, yeah, that looks right
15:52 jberger its amazing how hard it is to get some of those right!
15:56 marty garu: Yes, here a discussion on p5p.  http://www.nntp.perl.org/group/perl.​perl5.porters/2013/05/msg201351.html
15:59 libsysguy http://perlmonks.org/?node_id=1038191
15:59 KindOne joined #mojo
16:00 sri and it looks bad again for the dots pragma
16:01 libsysguy indeed it does
16:01 libsysguy I have been defending it at $work
16:01 libsysguy but there is a lot of hating on it
16:03 sri the "no change, ever!" faction is just too strong
16:04 maxhq1 joined #mojo
16:05 denisboyun joined #mojo
16:07 sri it's moments like this where i feel like i might be a ruby programmer in denial :S
16:07 garu marty: thanks, I'll push the new DDP version with those changes as soon as I get some time off from $work. Hopefully the workaround will help people if/until the issue is solved
16:09 garu sri: I wouldn't use it myself, and I'm not sure I'd like it on core, but if there's a way to change to core to support something like that through CPAN, I would totally support it
16:09 garu s/to core/the core/
16:10 sri it's pretty simple, the dot is more perlish than the arrow
16:10 sri there are more languages in the perl family using it
16:10 btyler jeez. frustrating reading p5p with arguments like "you're trying to make perl look like cool langauges like java or C!" ... personally java and C don't appear near the top of langauges that strike me as being considered 'cool'
16:10 jberger garu, I also probably wouldn't use it, but I'm all for it if it bring new perlers / retains them
16:10 KindOne joined #mojo
16:10 jberger but it seems like it is probably one of those things that would only work in the core, so we'll see
16:11 marty Implementing it as a pragma (with default off) seems relatively safe imo
16:12 jberger marty, agreed
16:12 sri when moe started people were asking a very interesting question... "what is the essence of perl?"
16:12 libsysguy I would be willing to wager that the haters of the .'s also dislike the Moose MOP
16:12 libsysguy perl purist :p
16:12 garu I'm not sure about that. I mean, I love the Moose MOP but I also like the arrow syntax :)
16:13 marty Just for fun I did a search and replace on one of my mojo controllers to replace all arrows with dots.  It looked kinda nice and (imo) was pretty easy to read.  I might use it.
16:13 garu it "points" to what you want to do
16:13 garu but I take it YMMV
16:15 garu marty: the issue with a core pragma is that (a) someone needs to build it; and (b) someone needs to support it, make the changes, fix bugs and make sure it doesn't break anything when it's activated... I haven't read the thread yet but it just might be that the price is too high for the canonical perl5 implementation. This is probably one of the places where 'moe' or other implementations can help
16:15 garu but I'll just shut up now :)
16:16 sri garu: same can be said about pretty much every new core feature, including sugnatures
16:16 marty sugnatures!
16:16 sri welcome to the "no change, ever!" faction
16:16 sri haha
16:18 garu sri: oh, absolutely! and in fact signatures have been a part of this for a long time, right? I mean, everybody knew what to do, but nobody had the time/knowledge/courage to do it. When someone did tackle the issue, it became something the pumpking itself said it *would* land in core :)
16:18 garu s/itself/himself/
16:18 sri moe is a dead end too though
16:18 sri only hope for perl going forward seems to be rakudo+moarvm now
16:19 garu sri: so who knows, maybe all it takes is a brave and willing person to JFDI with the dots pragma and it might be implemented
16:19 sri garu: the patch is ready, chip wrote it
16:19 garu oh, cool
16:20 sri http://www.xray.mpe.mpg.de/mailing-lis​ts/perl5-porters/2013-06/msg00277.html
16:20 garu I stand corrected then :)
16:20 sri it's tiny even
16:20 basiliscos joined #mojo
16:20 libsysguy wow that is small
16:21 sri but once the "no change, ever!" faction mentions the future backcompat boogeyman... it's pretty much over before it started
16:22 sri and it's always ribasushi, nicholas and jvromans :S
16:24 sri often demerhq and vincent too
16:25 garu sri: this seems pretty reasonable though: http://www.xray.mpe.mpg.de/mailing-lis​ts/perl5-porters/2013-06/msg00316.html
16:26 dod joined #mojo
16:26 sri that was before though
16:28 sri same pattern every time, once those people put future backcompat on the table the patch is toast
16:28 libsysguy joined #mojo
16:29 btyler I haven't been reading p5p all that much, but why is future backcompat such an issue when so many distros/organizations stick to old perl versions anyways?
16:30 btyler I understand that that is a Bad Thing, but it seems that plenty of groups have already made a decision regarding newer versions of perl, fretting about backcompat on p5p or no
16:33 jberger joined #mojo
16:33 sri and me ranting about it here is also becoming a pattern :/
16:35 amirite_ how do i setup a route that looks like /:foo/:bar-:baz
16:35 amirite_ i think the hyphen messes up the route
16:35 amirite_ it doesn't catch
16:35 garu escape it ("\-") maybe?
16:36 amirite_ i just tried that
16:37 marty amirite_: did you mean /:foo/:bar-/baz
16:39 amirite_ why would i mean that?
16:39 amirite_ three variables
16:40 amirite_ is this a bug?
16:40 garu sri: "my current inclination is to say that this can land, because there seems to be great popular acclaim" --rjbs's latest post on that thread
16:41 amirite_ is there a way to control the route match on a low level?
16:42 marty amirite_:  I see nowhere in the docs where it says you can have multiple generic placeholders between path delimiters.  Maybe I am missing something
16:43 amirite_ what? why should it say that
16:44 amirite_ imo the route should see :text and separate it from any other character based on whether it's \w
16:44 amirite_ if docs were supposed to document every unsupported exception then we'd be tax accounts
16:44 amirite_ accountants
16:45 amirite_ i consider this a bug, /me plays with mojo:routes:pattern in hope
16:50 r0b3rt get '/:foo/:bar-/:baz'   matches  http://localhost:3000/a/b-/c
16:51 amirite_ i don't want that, i want /a/b-c
16:51 amirite_ http://quran.com/2/3-4
16:52 good_news_everyone joined #mojo
16:52 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/5EhwSw
16:52 good_news_everyone mojo/master 758cb7c Sebastian Riedel: more unicode examples and tests
16:52 good_news_everyone left #mojo
16:54 marty thank you sri:  very informative.
16:54 sri marty++
16:58 amirite_ found a solution
16:59 amirite_ $r->route( '/(surah)/(ayah)-(last)'
16:59 amirite_ :)
17:17 trone joined #mojo
17:24 libsysguy I am trying to make the route / go to a route /admin without having to redefine the route (use the variable)
17:24 libsysguy what I have so far http://pastebin.com/8Reav6jc
17:25 libsysguy I don't see anything in the Routes::Routes docs on how to do that
17:30 tba libsysguy: any reason you need to do it that way? theres other ways you can achieve the same thing (placeholder regex, bridge, etc)
17:31 libsysguy maybe a bridge
17:32 libsysguy timtowtdi…too many :p
17:32 tba libsysguy: but thats why you use perl, right? ;)
17:33 libsysguy yup, so I can find all the non-optimal solutions first :p
17:34 libsysguy what is that edison quote about lightbulbs
17:34 tba you mean 'I have not failed. I've just found 10,000 ways that won't work.'
17:35 libsysguy yeah that one
17:35 tba sadly in perl its not 10,000 - its ∞ :p
17:36 libsysguy ^^ this one speaks the truth
17:48 libsysguy I've noticed when I screw something up a lot of the time I'll get the "page not found" error with a list of routes.  what is the best way to find out why its erroring out?
17:54 tba log/development.log should tell you, steps through the actions it takes
17:54 libsysguy ah
17:54 libsysguy I should have known that, I added that to my gitignore the other day too
17:54 * libsysguy smh
18:02 btyler joined #mojo
18:19 bluescreen joined #mojo
18:23 kitt_vl joined #mojo
18:32 Mike-PerlRecruiter_ joined #mojo
18:42 basic6 joined #mojo
18:48 Adura joined #mojo
18:49 abstract joined #mojo
18:52 libsysguy sri, does data insertion in mango work the same as MongoDB?
18:52 libsysguy err the MongoDB module
18:59 whitebook joined #mojo
19:02 sri libsysguy: i don't follow the MongoDB module, it's too broken, Mango is closer to the ruby and python drivers
19:02 libsysguy it appears to work similarly in documentation
19:02 libsysguy I am wrote a script to move a few million files into mongo using Mango, so we'll see how she holds up
19:05 sri Mango inserts are pretty slow, since they are always safe
19:05 libsysguy it seems like they were still a lot faster than what MySQL was working
19:06 libsysguy somebody had the bright idea to poke tarballs into a mysql db
19:06 libsysguy so I am moving them out and into mongo :-/
19:08 * sri starts to get the feeling that mongodb will go all mysql in the next release and replace the whole storage engine
19:08 sri (like mysql did with myisam vs innodb)
19:08 libsysguy I was wondering if they would come up with a more tightly coupled relational engine
19:09 sri going by the hints here and there that they are going to skip collection level locks and go straight for document level
19:09 sri apparently this company has already done it :) http://www.tokutek.com/pro​ducts/tokumx-for-mongodb/
19:10 sri mongodb with acid and everything
19:10 libsysguy is it really mongo anymore then
19:11 sri mongodb is still in its early mysql stage, pretty sure the internals will change completely in the next few years
19:12 libsysguy maybe they will change their .'s to ->'s
19:12 sri the new book also talks about pluggable storage engines
19:12 * libsysguy ducks
19:13 * sri puts libsysguy on the blacklist
19:13 libsysguy no no, I'm too young to be on the black list
19:13 libsysguy career = over
19:18 libsysguy well sri it worked pretty well
19:18 libsysguy till I ran out of disc space
19:18 sri which reminds me... everybody vote for this issue! https://jira.mongodb.org/browse/SERVER-1594
19:26 r0b3rt anyone played with rethinkdb?
19:27 sri yea, it's nowhere near ready
19:27 sri still needs a lot of work to get to the point where mongodb is now
19:29 gtodd I never got beyond redis ....
19:39 zege joined #mojo
19:47 bjoernfan joined #mojo
19:48 libsysguy sri would it be recommended to use the respond_to function in the controller with Mango or have a separate function for the mango api call to reside in?
20:01 trone joined #mojo
20:13 whitebook joined #mojo
20:40 SmokeMachine joined #mojo
20:40 gryphon joined #mojo
20:55 mire_ joined #mojo
21:01 MeierM joined #mojo
21:06 denisboyun_ joined #mojo
21:12 mire__ joined #mojo
21:15 amirite joined #mojo
21:22 mire_ joined #mojo
21:34 basic6 joined #mojo
21:35 lukep joined #mojo
21:57 basiliscos joined #mojo
22:36 wesal joined #mojo
22:43 sri neat, zsh-syntax-highlighting in homebrew
22:47 wesal left #mojo
22:49 moltar joined #mojo
22:58 rem_lex|pivo joined #mojo
23:06 jnbek joined #mojo
23:10 mire_ joined #mojo
23:23 ryozi joined #mojo

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