<gryfft>
I am attempting to remove a string from an array. I have the string, I have the array, I know the string is in the array. I want it out of the array. I forget how.
<nataraj>
used to code on fx 2 years back, on angstrom linux on an Arm handheld
yxhuvud has joined #ruby-lang
<nataraj>
but as of now, we dont use Xwindows on the handheld, just the framebuffer device for Qt Apps
<nataraj>
need ruby + Gui that works on fbdev
<erikh>
isn't there a qt toolkit for ruby as well?
<erikh>
dunno if it's still maintained; ruby generally doesn't lend itself well to gui applications
<erikh>
there's also libgosu, if banisterfiend is around he can tell you more.
<nataraj>
fx was too good for words
<erikh>
but that's more for gaming-style drawing
<shevy>
when ruby-gtk is available it is very nice, but on things like Arm handhelds I dunno, I can understand why you want to use fxruby :D
<nataraj>
rapid prototyping i suppose is what you guessed
<shevy>
why, that would work in ruby-gtk or ruby-qt just as well
<nataraj>
fx with X was damn cool, got a major order using it for prototyping
<nataraj>
Qt-embedded works well on framebuffer
<nataraj>
"Found qtruby4 (2.1.0), but was for platform mswin32"
<nataraj>
means, dont work on linux?
<shevy>
I dont know what that works. Qtruby however works on linux.
<shevy>
*means, not works
havenn has joined #ruby-lang
<drbrain>
nataraj: it means that the author didn't release a C-extension-only version
<drbrain>
just precompiled versions for windows ☹
reeze_xia has joined #ruby-lang
<nataraj>
drbrain, any way to get it fixed ?
<drbrain>
nataraj: contact the author and ask them to release just the source
<drbrain>
nataraj: reading the scroll back, it appears your fxruby error is due to fxtwitter not loading twitter4r properly
<nataraj>
so reinstall fxtwitter after installing twitter4r gem?
tenderlove has joined #ruby-lang
<erikh>
haha so DateTime\#- subtracts days when integer
<erikh>
man I just optimized the hell out of a sql query
francisfish has joined #ruby-lang
|Vargas| has joined #ruby-lang
<tenderlove>
zenspider: where did you get the syntax.css file for coderay?
looopy has joined #ruby-lang
<zenspider>
tenderlove: you know I'm on IM, right?
<tenderlove>
:-P
<zenspider>
I just _happened_ to look at IRC... prolly wouldn't have seen this otherwise
H2H has joined #ruby-lang
x0F__ has joined #ruby-lang
<drbrain>
nataraj: you'll probably have to fix fxtwitter, I doubt reinstalling will help if it didn't work the first time
burgestrand has joined #ruby-lang
kitallis has joined #ruby-lang
rue has joined #ruby-lang
<shevy>
in other words, you are quite doomed nataraj :)
<shevy>
a dead toolkit, an application that no longer works
<shevy>
but as long as you are not out of beer, things can still be good
darkf_ has joined #ruby-lang
JohnBat26 has joined #ruby-lang
<nataraj>
well, twitter aint so important
<nataraj>
but a ruby extension for a gui toolkit is
RORgasm has joined #ruby-lang
<nataraj>
interesting, there's a thing called ruby/fltk
solars has joined #ruby-lang
gryfft has joined #ruby-lang
nofxx has joined #ruby-lang
darkf_ has joined #ruby-lang
<erikh>
#<Sequel::SQLite::Dataset: "SELECT 'distinct(nodes.id)', 'nodes.*' FROM `nodes` INNER JOIN `reports` ON (`reports`.`node_id` = `nodes`.`id`) WHERE (('reports.created_at' >= '2012-03-26 22:57:49.086504-0700') AND ('reports.created_at' <= '2012-03-26 23:57:49.086561-0700')) GROUP BY 'nodes.id'">
* erikh
cries
<erikh>
it won't stop quoting those column names.
olesu has joined #ruby-lang
zmack has joined #ruby-lang
<erikh>
or the distincT()
dhruvasagar_ has joined #ruby-lang
wubino has joined #ruby-lang
srbartlett has joined #ruby-lang
looopy has joined #ruby-lang
tjadc has joined #ruby-lang
takaokou_ has joined #ruby-lang
gryfft has joined #ruby-lang
lperry65 has joined #ruby-lang
darkf has joined #ruby-lang
olesu has joined #ruby-lang
apeiros_ has joined #ruby-lang
Guest92496 has joined #ruby-lang
dhruvasagar has joined #ruby-lang
gokul has joined #ruby-lang
thrcka has joined #ruby-lang
havenn has joined #ruby-lang
kain has joined #ruby-lang
tekin has joined #ruby-lang
francisfish has joined #ruby-lang
gnufied has joined #ruby-lang
gianlucadv has joined #ruby-lang
judofyr has joined #ruby-lang
dc5ala has joined #ruby-lang
rue has joined #ruby-lang
heftig has joined #ruby-lang
futurechimp has joined #ruby-lang
woollyams has joined #ruby-lang
qwerxy has joined #ruby-lang
adambeynon has joined #ruby-lang
adambeynon has joined #ruby-lang
takaokouji has joined #ruby-lang
ando has joined #ruby-lang
pygmael has joined #ruby-lang
<Tasser>
erikh, distinct? Does it return sets?
akamike has joined #ruby-lang
gnufied has joined #ruby-lang
heftig_ has joined #ruby-lang
olesu has joined #ruby-lang
brushbox has joined #ruby-lang
mytrile has joined #ruby-lang
<judofyr>
erikh: how's the Sequel going?
heftig has joined #ruby-lang
<manveru>
wow... wc has some really crazy source
<chris2>
heh
beiter has joined #ruby-lang
<Defusal>
what is the best way to unescape a string like phps stripslashes()?
rue has joined #ruby-lang
<Defusal>
i have done a fair amount of searching and have not found a satisfactory answer, which seems strange for such a trivial question
<manveru>
Defusal: example?
<zenspider>
chris2: maybe it's just me... but when you put a link to "New Condoms" next to a link for "one tiny hand"... well...
<Defusal>
manveru, it just unescapes one level, so "\\" becomes "\", "\\n" becomes "\n"
<Defusal>
im embedding installation of sql files into my platform and the normal php install script reads the sql file, substitutes template stuff, splits on ";" and then runs each query
morozovm has joined #ruby-lang
<Defusal>
err, i missed the "stripslashes()" step
<Defusal>
ideally i'd like to do it as one query, but im not sure if thats possible with the ruby mysql driver, but either way, i need to unescape the sql first
morozovm has joined #ruby-lang
<manveru>
>> s = '\\\\ and \\\\n'; puts [s, s.gsub(/\\\\/, '\\')].join(' => ')
<manveru>
\\ and \\n => \ and \n
rohit has joined #ruby-lang
<Defusal>
hmm i guess i was overthinking it
<Defusal>
simple gsub should do the job, thanks
pygmael has joined #ruby-lang
<Defusal>
any idea on bulk querying with the MySQL driver manveru?
<manveru>
Defusal: uh, no
<manveru>
i use sequel, usually with sqlite or postgres
morozovm has joined #ruby-lang
daglees has joined #ruby-lang
daglees has joined #ruby-lang
<Defusal>
yea, i'm using mongodb for my platform, but i need to create and configure the mysql databases for the daemons it installs
RORgasm has joined #ruby-lang
<Defusal>
it seems terribly inefficient to split on ";" and make each query seperately to create the schema
<apeiros_>
load data infile
<apeiros_>
must be enabled serverside iirc, or requires a special grant
<apeiros_>
been a while since mysql
<apeiros_>
@Defusal
<manveru>
mysql and mongodb... what is the world coming to?
<Defusal>
hmm, i'll look into that, thanks apeiros_
<Defusal>
mongodb is awesome manveru
<Defusal>
faster and more convenient than sql
shtirlic_ has joined #ruby-lang
<manveru>
as long as you don't need persistence :P
<andrewvos>
Aww can we not do the mongo argument? It's to early for me to get unhappy about something.
<andrewvos>
too*
<manveru>
heh
fayimora_ has joined #ruby-lang
shtirlic_ has joined #ruby-lang
morozovm has quit [#ruby-lang]
morozovm has joined #ruby-lang
<Defusal>
oh right apeiros_
<Defusal>
i would use the mysql command line application if i wanted to use a sql file
<Defusal>
but i dont really want to write a file
<Defusal>
i'd rather do each query seperately :/
morozovm has quit [#ruby-lang]
<Defusal>
oh well, this will have to do
fayimora has joined #ruby-lang
morozovm has joined #ruby-lang
morozovm has quit [#ruby-lang]
morozovm has joined #ruby-lang
morozovm has joined #ruby-lang
morozovm has quit [#ruby-lang]
morozovm has joined #ruby-lang
morozovm has joined #ruby-lang
<zenspider>
Defusal: popen4
* zenspider
goes to bed
morozovm has joined #ruby-lang
burgestrand has joined #ruby-lang
dhruvasagar has joined #ruby-lang
dfr|mac has joined #ruby-lang
fayimora_ has joined #ruby-lang
gianlucadv has joined #ruby-lang
srbartlett has joined #ruby-lang
srbartle_ has joined #ruby-lang
<Defusal>
zenspider, i have no problem spawing stuff...
rue has joined #ruby-lang
<Defusal>
manveru, while i was out i just realized that isnt the same as stripslashes
<Defusal>
strip slashes will also turn "\'" into "'"
<apeiros_>
stripslashes… that's something from before the time of sane database interfaces, right?
<Defusal>
apeiros_, it's still used in php today
<Defusal>
i havnt needed to even read php code in years
<Defusal>
but now i need to rewrite some in ruby
<apeiros_>
Defusal: I think I said *sane* already… :-p
<Defusal>
and one of the things this does before executing the sql is stripslashes()
<Defusal>
so i need a way to do the same from ruby apeiros_
<apeiros_>
Defusal: if you want to replicate the php code bug for bug, you'll use something like stripslashes
<Defusal>
apeiros_, which is why i am here, asking what the best way to do the same in ruby is...
<judofyr>
Defusal: oh? :S
<apeiros_>
using stripslashes before performing a query is weird, though. usually, you added slashes (in order to escape data)
<Defusal>
judofyr, i need to simulate php's stripslashes(), and i sure as hell am not going to create a hash with every possible character in it
<apeiros_>
later people realized, that db drivers had escape mechanisms of their own, and started using those (like mysql_blabla_quote_blabla or something like that)
<apeiros_>
and even later, people realized, that there's bind variables, which obviates the need for escaping altogether
<judofyr>
Defusal: you don't need "every possible character". it falls back to the string if it's not found
<Defusal>
apeiros_, sure... but that doesnt chance the fact that this is how this is done
<Defusal>
judofyr, sigh.
<apeiros_>
"it is done this way, however stupid, lets do it that stupid again"
<apeiros_>
great
<apeiros_>
meh
<apeiros_>
not my problem
<Defusal>
judofyr, php does not use a hash of every char possible, so neither will i
<judofyr>
Defusal: then read stripslashes' source :)
<Defusal>
apeiros_, i don't have time to waste with this...
<Defusal>
i just need it to work, its for installing the database of a php application
<Defusal>
i couldnt care what it does, as long as it works
<Defusal>
i wouldn't have fcgi php installed on my server at all if i didn't need to provide these php applications
<judofyr>
Defusal: popen then
<Defusal>
popen what?
<Defusal>
spawn php to unescape a string?
<Defusal>
i have never heard of something so rediculous
<judofyr>
Defusal: write a PHP script and runs stripslashes on ARVG[0] :)
<Defusal>
good god
<Defusal>
i am not quite *that* lazy
<judofyr>
…
<Defusal>
it's bad enough i have to have php installed on the server at all
<Defusal>
im not going to call php code from ruby to remove some slashes from a string
<judofyr>
well, you don't want to figure out how stripslashes works either, so…
<Defusal>
well i've got my answer
<Defusal>
ruby has no way to do such a thing
<Defusal>
so now i will have to waste time reimplementing phps function
<judofyr>
true
<judofyr>
Ruby doesn't ship with crappy PHP functions :)
<Defusal>
that i do not expect
<Defusal>
i do however find it strange that out of every php developer thats ever moved to ruby, not one has ever needed to do this
<apeiros_>
weird prioritization… you prefer wasting time over porting/reimplementing the wrong way, rather than do it right…
dhruvasagar has joined #ruby-lang
<Defusal>
apeiros_, and the right way is what?
<apeiros_>
already told you
<apeiros_>
if you've data to be escaped, use bind variables
<apeiros_>
s/data to be escaped/variable data in queries/
<Defusal>
i have sql files
<apeiros_>
use load data infile for sql files
sora_h has joined #ruby-lang
<Defusal>
which have template variables subbed
<apeiros_>
well, that's stupid
<apeiros_>
again, bind variables
<apeiros_>
but I'm repeating myself
<Defusal>
well, thats how the php install script does it
<apeiros_>
well, that's the stupid way
<Defusal>
you're missing the point
<Defusal>
im not going to bother again
<apeiros_>
but again, I'm repeating myself
<apeiros_>
Defusal: make an example of one such file…
<Defusal>
it's fine
<Defusal>
i'll hack it together to make it work
<Defusal>
i don't care anymore
<apeiros_>
*shrug* train wreck it is, fine
<Defusal>
i have far better thing to do
dous has joined #ruby-lang
<Defusal>
and i have weeks of work to do before the end of the month
dous has joined #ruby-lang
<Defusal>
which is in 5 days time
* apeiros_
bets it'd probably just be replacing those "template vars" with proper bind variable syntax
<apeiros_>
but ignorance rules the planet, rock on
<Defusal>
dude.
<Defusal>
they are stupid bloody template variables
<Defusal>
they are not even dynamic
<Defusal>
stuff like {prefix}
<Defusal>
you're making a big deal about nothing at all
<apeiros_>
Defusal: how about you waste 5 more minutes on google and find out how bind vars work?
<apeiros_>
and you're behaving stupid. I'm off. good luck.
<Defusal>
this doesn't even have anything to do with my issue -.-
<Defusal>
theres a reason the slashes are stripped
<Defusal>
it has nothing to do with the templating
<Defusal>
i can't perform the queries without stripping the slashes, regardless of how the templating is done
<Defusal>
i don't know why that is so hard to understand.
<Defusal>
...and you tell me im ignorant, pfft
tla has joined #ruby-lang
zmack has joined #ruby-lang
olesu has joined #ruby-lang
sora_h has joined #ruby-lang
kokozedman has joined #ruby-lang
<kokozedman>
hey guys... i'm trying to develop a Ruby C extension, but somehow, i can't get the error "[BUG] cfp consistency error - call0" fixed
<kokozedman>
what i'm doing, is that i create a C threads (pthread), and call a Ruby function
<kokozedman>
the threads call the Ruby function... and causes this "[BUG] cfp consistency error - call0"
<kokozedman>
but the strange thing is: the error happens mostly when the call is completed
fayimora has joined #ruby-lang
batmanian has joined #ruby-lang
<kokozedman>
anyone with some knowledge on this?
zz_wasnotrice has joined #ruby-lang
<kokozedman>
this error happens only on 1.9.x
qwerxy has quit [#ruby-lang]
<rue>
I haven't tried to use native threads with a Ruby extension, so I dunno
<judofyr>
kokozedman: you should probably only call Ruby methods from one thread
<kokozedman>
another info: i'm developing the extension using RubyInline... and the method I call is a RubyInline one
<kokozedman>
so, i think, it is somehow, still something for the C context...
<rue>
RubyInline ≠ extension
<kokozedman>
but when i call rb_funcall(...), something about control frame is involved
<kokozedman>
and when the call ends, some kind of check that should be incremented doesn't get incremented, and seems to cause the error
<kokozedman>
rue: i agree... but it's somehow generating a C code that is really abiding to the way how C extensions are made
fayimora has joined #ruby-lang
lsegal has joined #ruby-lang
<rue>
I'm not sure what that means, but what I mean is that if you're using Inline, don't say you're writing an extension :)
<rue>
There's FFI, there's Inline, and then there're extensions
tommyvyo has joined #ruby-lang
<judofyr>
there's FFI, there's extensions, and then there's madness
<judofyr>
kokozedman: if you just need to call some methods in a dylib
<Defusal>
this is gonna be a very long day :/
mssola has joined #ruby-lang
<kokozedman>
hmm... good to know... thanks judofyr
<judofyr>
kokozedman: well supported in both MRI, JRuby and Rubinius
<kokozedman>
rue: have you looked at a C code generated by RubyInline? it IS an extension... exactly like you would write if you are to create an extension...
<rue>
kokozedman: Don't argue, please, these are all distinct.
<rue>
Is the code generated similar? Yes, it does call into the same codebase, after all
heftig has joined #ruby-lang
ramonmaruko has joined #ruby-lang
<kokozedman>
rue: alright... speaking of codebase... where can i find the API list of Ruby C functions (like rb_protect, rb_xxxx)
<rue>
README.ext has some, for more you'll probably need to go into ruby.h and possibly other headers
<manveru>
oh well, at least the normal bbc page is still fine
<Defusal>
manveru, im gonna switch to sequel for these maintenance queries, it looks a lot nicer than rolling my own solution with the mysql driver. thanks for mentioning it
<manveru>
good to hear :)
<manveru>
sequel rocks
RORgasm has joined #ruby-lang
ankurgel has joined #ruby-lang
<judofyr>
sequel++
<yorickpeterse>
^
<lianj>
|
Guest92496 has joined #ruby-lang
Guest92496 has joined #ruby-lang
<chris2>
zenspider: heh
<ankurgel>
shortest way to tell if a number is a prime or not? I ended up writing http://pastebin.com/tMA8g4C5 to find first 10001 prime numbers.
<ankurgel>
Also, very hoggy. Took 1 minute in execution.
<matled>
ankurgel: efficient or short?
<manveru>
well, for one you can use += 2 if you start with a prime
<matled>
short I'd go with: (2...n-1).all? { |i| n.divmod(i).last != 0 }
<f3ew>
ruby 1.9.2, erb is being consumed via Puppet, if that matters
stiang has joined #ruby-lang
<rue>
It'd probably be better to take this out of your template anyway
stiang has quit [#ruby-lang]
<rue>
But your problem is that you're not seeing changes in entry?
<andrewvos>
rue: Funny
<f3ew>
rue: right
<f3ew>
This is work in progress, so it will be cleaned up
JEG2 has joined #ruby-lang
Stalkr_ has joined #ruby-lang
stiang has joined #ruby-lang
<f3ew>
Suggestions on debugging?
wmoxam has joined #ruby-lang
<Silex>
use your debugger? breakpoints etc
jxie has joined #ruby-lang
fayimora has joined #ruby-lang
<rue>
I'd move it out of the erb first.
<f3ew>
rue: the function?
<rue>
Everything but the presentation
wyhaines has joined #ruby-lang
arkonova has joined #ruby-lang
pygmael has joined #ruby-lang
RORgasm has joined #ruby-lang
dejongge has joined #ruby-lang
nofxx has joined #ruby-lang
selvakumaran has joined #ruby-lang
heftig has joined #ruby-lang
<selvakumaran>
Hello there?? I wanna install QTRuby on my opensuse machine, Any help?
futurechimp has joined #ruby-lang
wallerdev has joined #ruby-lang
<andrewvos>
!tumbleweed
<andrewvos>
@ @ @ @
wyhaines has joined #ruby-lang
<matti>
lol
pgpkeys has joined #ruby-lang
stiang has joined #ruby-lang
fayimora has joined #ruby-lang
msisk has joined #ruby-lang
sandbags has joined #ruby-lang
rolfb has joined #ruby-lang
<ankurgel>
arr=[200,100,50,20,10,5,2,1]. I want to find groups such as total_sum_value=400.. these values can be grouped by [200,200] (total=400), [1.. 400 times] (total=400), [50,50,100,200] etc
PsiliPharm has joined #ruby-lang
<ankurgel>
guess Arr#combination can't help me out here.
<shevy>
dont even know of an algorithm to do that
<ankurgel>
:D guess then will have to formalize an adequate algorithm for it.
<ankurgel>
there is a related problem on ProjectEuler I read some months ago. Just remembered some part of it and modified the problem statement to practice.
<andrewvos>
yeah I did that at some point
<andrewvos>
ankurgel: If you can find the number, I'll give you the code ;)
<judofyr>
ankurgel: good luck. if you can solve that in polynomial time, you'll be famous
<erikh>
oh, nothing complicated, just sequel's aggressive quoting is getting in my way
<erikh>
anyhow; I need to bail, work beckons
<shevy>
BEND OVER!!!
selvakumaran has joined #ruby-lang
<shevy>
oops, sorry. wrong channel
<andrewvos>
/whois shevy
<andrewvos>
You're only in ruby-lang shevy :/
<judofyr>
wrong server?
slyphon has joined #ruby-lang
workmad3 has joined #ruby-lang
RORgasm has joined #ruby-lang
<erikh>
andrewvos: it's called +i
<erikh>
you can only see the channels you're both in
<andrewvos>
erikh: +i?
<erikh>
invisible mode
<erikh>
try /mode andrewvos +i
Austin__ has joined #ruby-lang
<andrewvos>
erikh: You lost me man
mrchrisadams has joined #ruby-lang
florentg2 has joined #ruby-lang
<rue>
I can't see you
florentg2 has joined #ruby-lang
dejongge has joined #ruby-lang
<Mon_Ouie>
IRC ninjas
<matled>
I don't like that classes get indented more and more if they are within multiple modules. is there any typical way to solve this? module A; module B; module C; \n class Foo? or class A::B::C::Foo (the modules have to be defined already for this..)?
<judofyr>
matled: that's pretty much your options
<shevy>
andrewvos I am spawned on the whole of freenode man
<shevy>
even on #machomebrew without having a mac
senthil has joined #ruby-lang
gsav has joined #ruby-lang
<andrewvos>
hehe
<shevy>
but I am learning
<shevy>
they use pure ruby recipes for every program
<shevy>
but my approach would be different, I'd just embed pry :)
<matti>
Um.
<Mon_Ouie>
I'd do it the other way around, have a convenient syntax to use Ruby from a regular shell
<Mon_Ouie>
(and running said code in the same VM)
<matti>
I was using rush for two weeks.
<matti>
Its nice.
<shevy>
yeah
<shevy>
why did you stop after 2 weeks?
<matti>
I am more used to bash with vi mode.
<frem>
bash has vi mode?
<matti>
And sometimes something easy takes too long.
<matti>
frem: Yes!
chimkan has joined #ruby-lang
SuperTaz_work has joined #ruby-lang
Stalkr_ has joined #ruby-lang
<frem>
matti: that is awesome, and now i need to customize my prompt to look like vim-powerline. :3
g0bl1n has joined #ruby-lang
apeiros_ has joined #ruby-lang
PsiliPharm has joined #ruby-lang
<matti>
frem: Haha ;-)
ascarter has joined #ruby-lang
ilyam has joined #ruby-lang
wyhaines has joined #ruby-lang
g0bl1n has joined #ruby-lang
<rue>
Isn't there a powerline for bash?
davpoind has joined #ruby-lang
hagabaka has joined #ruby-lang
hagabaka has joined #ruby-lang
<shevy>
what is a powerline?
Indian has joined #ruby-lang
H2H has joined #ruby-lang
solars has joined #ruby-lang
io_syl has joined #ruby-lang
morozovm_ has joined #ruby-lang
<andrewvos>
shevy: It's like a normal line, but more power.
butchanton has joined #ruby-lang
Banistergalaxy has joined #ruby-lang
singpolyma has joined #ruby-lang
darkf has joined #ruby-lang
<shevy>
lol
ddfreyne has joined #ruby-lang
burgestrand has joined #ruby-lang
kyrylo has joined #ruby-lang
kyrylo has joined #ruby-lang
ascarter has joined #ruby-lang
morozovm_ has joined #ruby-lang
achiu1 has joined #ruby-lang
ruskie has joined #ruby-lang
takaokouji has joined #ruby-lang
slyphon has joined #ruby-lang
Kero has joined #ruby-lang
brownies has joined #ruby-lang
Carnage\ has joined #ruby-lang
achiu has joined #ruby-lang
dejongge1 has joined #ruby-lang
ilyam has joined #ruby-lang
Oloryn_lt2 has joined #ruby-lang
havenn has joined #ruby-lang
<kyrylo>
What I'm supposed to do, if my ~/.gemrc doesn't work? I mean, I write stuff there, but nothing happens. But /etc/gemrc works flawlessly. I don't use RVM.
jmontross has joined #ruby-lang
<burgestrand>
Anybody knows about the caveats when installing C extensions on Mac OS that uses frameworks instead of just regular library linking? Have a user of one of my gems where the compiled extension can’t find a symbol that *should* exist on Mac OS.
<burgestrand>
O
<Mon_Ouie>
That framework is bundled with Mac OS?
<burgestrand>
Mon_Ouie: OpenAL, as far as I know it is
<Mon_Ouie>
Yep, I didn't have any problem with that one
<kyrylo>
Hum, I guess, I know the answer. Sorry for the false alarm.
<burgestrand>
Mon_Ouie: you ever written an extconf.rb that links against it?
<Mon_Ouie>
Maybe the user has another version that takes priority over the system one?
<Mon_Ouie>
So pretty much the same thing as far as OpenAL is concerned
<Mon_Ouie>
btw I think have_library changes $LDFLAGS itself
<burgestrand>
Oh, have to look out for that one
<burgestrand>
Ah, yes, it might actually add -llibraryname itself
olesu has joined #ruby-lang
gabeanze1ini has joined #ruby-lang
skyjumper has quit [#ruby-lang]
fayimora has joined #ruby-lang
brianpWins has joined #ruby-lang
RORgasm has joined #ruby-lang
morozovm_ has joined #ruby-lang
jmfcool has joined #ruby-lang
qpingu has joined #ruby-lang
<burgestrand>
Mon_Ouie: the user had openal installed in addition to the bundled one, thanks for the tip!
butchanton has joined #ruby-lang
crudson has joined #ruby-lang
olesu has joined #ruby-lang
segy has joined #ruby-lang
morozovm_ has joined #ruby-lang
kitallis has joined #ruby-lang
gnufied has joined #ruby-lang
gnufied has joined #ruby-lang
tbuehlmann has joined #ruby-lang
Jake232 has joined #ruby-lang
<Jake232>
Randomish question, but I figure somebody here probably knows. What's the opposite of a tree transversal? eg: Putting something "flat" into a tree structure
<Asher>
hierarchizing?
<Jake232>
I'm not even sure if that's a word, but I'll go with it
tonesfrommars has joined #ruby-lang
<drbrain>
building a tree
<outoftime>
Jake232: arborization
vpamulap has joined #ruby-lang
<Asher>
haha
<shevy>
woodcutting
<apeiros_>
transmogrification
morozovm has quit [#ruby-lang]
varreli has joined #ruby-lang
morozovm has joined #ruby-lang
scampbell has joined #ruby-lang
varreli has joined #ruby-lang
ascarter has joined #ruby-lang
olesu has joined #ruby-lang
fayimora has joined #ruby-lang
workmad3 has joined #ruby-lang
<TTilus>
botanization
gregf has joined #ruby-lang
mrsolo has joined #ruby-lang
chimkan_ has joined #ruby-lang
Oloryn_lt2 has joined #ruby-lang
slyphon has joined #ruby-lang
<jmontross>
comiditization
fayimora has joined #ruby-lang
Okasu has joined #ruby-lang
<Okasu>
hello
<Okasu>
can you explain the difference btween net/http in 1.8 and 1.9?
ilyam has joined #ruby-lang
<Okasu>
i can't post with 1.8 because of bot detection or something
<Okasu>
but with 1.9 post work fine
olesu has joined #ruby-lang
<rue>
Okasu: A more precise error and reproduction might be helpful.
<rue>
Though I don't remember, maybe there was a change to cookie handling or summit.
<Okasu>
with 1.8 i got "<html><title>No bot</title></html><body>No published bot with.</body>"
<ftravers>
can someone recommend the best way to auto t s
<ftravers>
...to auto start a sinatra/thin app on ubuntu
woollyams has joined #ruby-lang
<ftravers>
i've tried daemonizing it, which works fine running from the command line as a logged in user but i want it auto started (like production) by root on system startup...
<imperator>
Defusal, do you see anything with -W ?
<Defusal>
workmad3, EventMachine, so no
<workmad3>
Defusal: oh, just you said you were hitting Process.exists?
<Defusal>
yup
<Defusal>
i use Process.getpgid(pid)
<Defusal>
which performs extremely well
<Defusal>
i benchmarked it not long ago
<imperator>
what platform?
<Defusal>
linux
<Defusal>
ok, thats definitely it workmad3
<Defusal>
i renamed the method to does_exist? and it behaves correctly
<imperator>
got a short list of gems that could be it?
<Defusal>
and the EPERM error is because the pid has since been recycled, so even though it does exist, it's not the process spawned by this application
y3llow_ has joined #ruby-lang
<Defusal>
that is gonna be tricky to solve
s0ber_ has joined #ruby-lang
<Defusal>
im not sure of a reliable way to check if a pid is the expected process
<Defusal>
imperator, its not all that short, but i can pastebin it if you like?
<Defusal>
but heres a more important issue right now
<Defusal>
Process.getpgid says this pid exists, but i can't ps ax | grep it
<Defusal>
how is that possible?
<Defusal>
it kills the application because of a EPERM error trying to kill the pid, because it thinks its running
<imperator>
well, i thought linux showed everything, but i know some platforms only show the processes you own
<Defusal>
i tried sudo
<Defusal>
it does not exist to root or anything
<imperator>
wait, EPERM?
<imperator>
one sec, let me check something
<Defusal>
EPERM killing a non-existant pid
<Defusal>
well, non-existant according to px ax
<Defusal>
ps ax*
<Defusal>
but then Process.getpgid is faulty?
<imperator>
if it didn't exist, i would expect ESRCH
<Defusal>
yeah
<Defusal>
well
<Defusal>
thats what Process.getpgid gives when it doesn't exist
<imperator>
fire up irb, and try a Process.kill(0, 99999999), what does it return?
<Defusal>
but Process.getpgid thinks it does exist
rock-n-rolla has joined #ruby-lang
<Defusal>
-e:1:in `kill': No such process (Errno::ESRCH)
rock-n-rolla has quit [#ruby-lang]
<Defusal>
$ ruby -e 'p Process.kill(0, 4652)'
<Defusal>
-e:1:in `kill': Operation not permitted (Errno::EPERM)
<imperator>
got sys-proctable installed?
<Defusal>
this pid gives the same EPERM error, yet it doesn't appear to exist :|
<Defusal>
nope
<imperator>
install it if you don't mind
<imperator>
if you do mind, then look under /proc for 4652
robbyoconnor has joined #ruby-lang
<slyphon>
gah fuck
<slyphon>
the new bundler behavior is totally fucking me
<imperator>
slyphon, hello there
<slyphon>
it used to just say "Oh well, can't reach your-internal-repo" now it pukes
<slyphon>
imperator: oh hai!
<imperator>
Defusal, any luck?
<Defusal>
imperator, finally figured it out now
heftig has joined #ruby-lang
<Defusal>
its a thread, not a process
<Defusal>
so then my next question is, how do i test if the Process.getpgid is for a thread or a process
<imperator>
well, let's go back to what your method is trying to do
<imperator>
Process.exists? i mean
<Defusal>
basically: (Process.getpgid(pid) and true) rescue false
<imperator>
the way i was taught was to use Process.kill(0, pid) + rescue if you want to check for its existence
<Defusal>
but it rescues ESRCH
<Defusal>
getpgid is supposed to be a less destructive method to do it
<imperator>
what does "less destructive" mean?
<Defusal>
i havnt benchmarked Process.kill(0.. but it produces the same EPERM error
<Defusal>
so that doesn't help anything
<imperator>
which is odd, actually
<slyphon>
wow, the bundler code is...
* slyphon
chokes down some vomit
<slyphon>
...pragmatic
<imperator>
slyphon, something that looks like it was hacked together in a weekend, and pushed into production so that users would become beta testers?
<slyphon>
yyyyyyyyeah
<imperator>
just a guess
<slyphon>
that
<Defusal>
ok then
<slyphon>
"here's some code! we'll make it work, y'know...soon"
<Defusal>
i guess my only choice is to assume all EPERM errors are threads
<slyphon>
like everything _why ever wrote
<Defusal>
of course that means it'll assume that any process owned by someone else is a thread
<Defusal>
but it doesnt look like theres any other option
<imperator>
Defusal, i guess my question is, how are you getting a thread id in the first place?
alexkane has joined #ruby-lang
<Defusal>
imperator, reused process id
<Defusal>
stale pid in a text file
<imperator>
a....text file?
<slyphon>
warning! race conditions!
dcorbin has joined #ruby-lang
<Defusal>
simplest way to track spawned pids and kill them on startup if there was not a clean shutdown
<slyphon>
er
<Defusal>
i tried a few methods with pipes back when i implemented this, but no method worked correctly
<whitequark>
kill -0 is a proper way to check for existence of a process
<Defusal>
so i implemented this, and it has worked fine for over a year
<whitequark>
there's nothing "destructive" in it
<Defusal>
whitequark, getpgid works just the same
<whitequark>
and `kill' should in fact has been called "sendsignal"
<imperator>
whitequark, right, but why would he get EPERM errors on a thread id?
<Defusal>
i'll benchmark kill -0 sometime to compare
<Defusal>
whitequark, i need to test if the pid is a pid or a thread id
<imperator>
i mean, if the pid doesn't exist, it should give him ESRCH
<whitequark>
do you mean thread IDs on Linux?
<Defusal>
imperator, to linux threads are processes too :/
<whitequark>
they're like in some other namespace or something like that
<Defusal>
whitequark, pid has been reused as a thread id and now causes a EPERM when trying to kill
<Defusal>
so i need to tell if its a thread id or not
<imperator>
whitequark, oic
<whitequark>
oic?
<imperator>
Oh, I see
<whitequark>
ah ok
<Defusal>
any ideas?
<imperator>
hm
<imperator>
Defusal, so, if it *is* a thread id, you still need to kill it, or just ignore it?
<whitequark>
Defusal: it kills the process for me
<Defusal>
imperator, ignore it and remove from the pid list
<Defusal>
whitequark, the process that the thread belongs to is owned by a different user
<Defusal>
but i wouldn't want to kill it either way
<Defusal>
killing random processes threads would not be good at all
<Defusal>
ruby should have a way to test if the pid is a thread if all its methods of checking a pid return true for a thread :/
<whitequark>
it shouldn't
<Defusal>
then im not sure how you can do safe process management
<rue>
Processes aren't threads
<Defusal>
rue, they are in this case
<whitequark>
ruby is not linux-specific, and posix does not mention a way to do this
<Defusal>
yeah
<Defusal>
ok, next best workaround...
<Defusal>
whats the best way to get the process name from a pid?
<whitequark>
I'm pretty sure you are doing something else wrong
<Defusal>
then i can check if it is the correct process
<rue>
This doesn't make any sense
<Defusal>
whitequark, i already explained the entire situation to you?
<rue>
If you're in a situation where your pids and thread ids are getting mixed
<Defusal>
rue, old pid has been reused by a thread and now appars to be alive
<Defusal>
appears*
<slyphon>
anyone know if bundler with recursively follow 'git' references?
<Defusal>
rue, since threads use the same ids close to the parent pid, it can easily happen
havenn has joined #ruby-lang
<Defusal>
keep in mind
<imperator>
slyphon, i'm scared to find out
<Defusal>
i only found out yesterday for the first time that threads even have pids
futurechimp has joined #ruby-lang
<slyphon>
imperator: it would provide a way around the new behavior of breaking
<slyphon>
which is kind of chafing my nether parts
<slyphon>
y'know
<slyphon>
"down there"
<Defusal>
to htop they appear the same as processes, unless you hide userland threads, which confused the hell out of me before i spoke to some experienced linux professionals
<Defusal>
and i've been using linux for a long time :/
sdeobald has joined #ruby-lang
<whitequark>
Defusal: I still don't understand what's the overall task you're trying to accomplish (no, "check if pid is running" is not an overall task)
Okasu has quit [#ruby-lang]
Aaaarg has joined #ruby-lang
<rue>
All threads of a given process have the same PID
<Defusal>
whitequark, ignoring the fact that this applications modules do a lot of process management for installing, configuring and spawning external processes... its core spawns a process which is required before any modules can be loaded and the pid of that process is stored in a text file. the pid is killed on clean shutdown, but also on startup incase on a crash
<Defusal>
rue, incorrect
<Defusal>
well, not completely
PsiliPharm has joined #ruby-lang
<Defusal>
you are right, but threads have their own thread id too
<Defusal>
which is close to the parents pid
<rue>
Yes, but that's nothing to do with process handling
robbyoconnor has joined #ruby-lang
looopy has joined #ruby-lang
<rue>
If you send a signal to a pid, it's not going to go look for tids, it looks for the process
Aaaarg has joined #ruby-lang
<outoftime>
anyone run into an SSL_connect ECONNRESET when trying to talk to the Facebook API?
<Defusal>
rue, tell that to Process.getpgid, Process.kill and kill -0
<whitequark>
rue: technically threads do have distinct pids
<whitequark>
rue: and when a signal is sent to any thread of a process, it is delivered to a random (sic) thread of that process
<Defusal>
rue, why don't you test yourself ;)
<whitequark>
according to a manpage
<rue>
whitequark: Sure, any thread might end up handling a signal
<Defusal>
ok, i guess the best i can do then is read /proc/<pid>/cmdline to see if the process is the one intended to be killed
<Defusal>
since there does not appear to be any way to actually tell if its a thread or a read process
<Defusal>
and /proc/<thread id>/cmdline returns the command line of the parent process
<Defusal>
but this will work for the most part
heftig has joined #ruby-lang
melter has joined #ruby-lang
Oloryn_lt2 has joined #ruby-lang
pygmael has joined #ruby-lang
<erikh>
kill(0, pid) is the way to roll
<Defusal>
erikh, you kinda missed the whole discussion
<Defusal>
erikh, the issue is telling whether pid is a thread or a process
<rue>
erikh: Bizarro world :)
<Defusal>
because to kill and getpgid, the pid is alive whether its a parent process or a thread belonging to a process
<erikh>
getpgid does something totally different from kill.
<erikh>
it gets the process group
<erikh>
which is not necessarily the pid you're after
<erikh>
rue: totally.
<Defusal>
erikh, getpgid does a good job of testing if the pid exists or not, i havn't benchmarked kill -0, but getpgid performs extremely well
<Defusal>
erikh, but either way, both can't tell if its really a thread id, or a process id
<erikh>
dude
<erikh>
they're syscalls
<erikh>
you don't need to benchmark them, especially under ruby.
H2H has joined #ruby-lang
<erikh>
getpgid also doesn't test the pid
<erikh>
it tests the process group
<erikh>
go read manpages.
robbyoconnor has joined #ruby-lang
<imperator>
slyphon, there's powders and such for that
<slyphon>
haha
<Defusal>
erikh
<Defusal>
getpgid takes a pid
<erikh>
yes
<Defusal>
i have been using it to test process existance for a long time
<erikh>
GO READ THE FUCKING MANUAL.
<Defusal>
it is recommended over kill -0 in many places
<Defusal>
wtf?
<imperator>
let's settle down
<Defusal>
are you on crack?
<Defusal>
theres nothing at all wrong with using getpgid to check a pids existance
<bougyman>
is this all on linux?
<wmoxam>
loldrama
<Defusal>
its no better or worse than kill -0
<imperator>
bougyman, yeah
<bougyman>
wtf, then
<bougyman>
/proc/PID
<Defusal>
bougyman, theres nothing wrong with Process.getpgid(pid) or Process.kill(0, pid) for existance checking, but you still don't know if the "pid" is a process or a thread id
robbyoconnor has joined #ruby-lang
<bougyman>
Defusal: you do if you check /proc
<bougyman>
which is why I wtf-ed
<imperator>
sys-proctable wraps it, so: require 'sys/proctable'; if Sys::ProcTable.ps(pid); do_stuff; else; do_nothing; end
<Defusal>
bougyman, ok then where exactly must i check in /proc/pid/...
<imperator>
there won't be an entry under /proc if it's a thread id
<bougyman>
imperator just gave you a rubyish way to do so.
<imperator>
afaik
<bougyman>
there won't, imperator
futurechimp has joined #ruby-lang
<rue>
Because it's a THREAD :D
<Defusal>
i'll have to benchmark proctable, but i guess if it uses /proc it shouldn't be too bad
<Defusal>
imperator, bougyman: wrong
<bougyman>
benchmarking system calls?
<slyphon>
wow
<bougyman>
what do you mean: wrong?
<imperator>
yep, on linux it's just reading /proc (and maybe getting some info from other /proc dirs)
<slyphon>
ding-dong
<Defusal>
ok, two answers:
<Defusal>
benchmarking, because i plan on polling many pids at least twice a second forever
<Defusal>
because there is no way to watch a detached pid without a netlink implementation
<Defusal>
and i can cat /proc/<thread id>/cmdline and it will show me the parent processes command line
<bougyman>
a detached pid is a misbehaving process.
<Defusal>
so that does not test if its a real process
<Defusal>
bougyman, detached intentionally, with a lot of effort, from the spawning daemon
thone_ has joined #ruby-lang
<bougyman>
/proc/PID, not /proc/thread id
<bougyman>
stat /proc/$$
<Defusal>
cat: /proc/PID: No such file or directory
<Defusal>
theres no such thing
<bougyman>
i meant, replace with a PID, sir.
<bougyman>
like $$ for your current one.
* drbrain
palms face
<Defusal>
bougyman
<Defusal>
you are really not following are you?
<Defusal>
i will paste you an example, so we are on the same page
<bougyman>
i'm not the one who looked for /proc/PID.
<imperator>
let's wait for the example
<Defusal>
bougyman, because that is the only thing that made sense
<imperator>
i would think these detached pids would be zombies
<Defusal>
beacuse you did not listen to me
<Defusal>
because*
<bougyman>
you have a list of pids and want to see if they're running, right?
<bougyman>
so, other than how gross that is, you just need to see if the directory exists of that PID in /proc/
<imperator>
hm....if you try to kill a zombie (process), what happens? would it give you EPERM?
<Defusal>
$ kill -0 4652 # 4652 is a *thread* id
<Defusal>
bash: kill: (4652) - Operation not permitted
<bougyman>
if you awnt to find more data about it, it exists inside that /proc/1234 (for PID 1234) dir.
<Defusal>
there you go, if you want more proof, do your own damn testing
<Defusal>
-.-
io_syl has joined #ruby-lang
<Defusal>
imperator, what kind of zombie process?
<Defusal>
a defunct process just ignores even kill signals until its parent dies
<imperator>
right, but i'm not sure what ruby does
<Defusal>
bougyman, you should really not just assume stuff and ignore what people say :(
<Defusal>
imperator, well this is really not related to this issue at least
<Defusal>
i solved all my defunct process issues last week
<Defusal>
truely detaching a spawned process from ruby is not very easy, the only way is to spawn a proxy ruby process which catches signals, detaches from the child, prints the real processes id to stdout and then exits
divoxx has joined #ruby-lang
<bougyman>
which kernel are you running?
robbyoconnor has joined #ruby-lang
<bougyman>
just checked on 3.3, there are no /proc/thread dirs.
<Defusal>
anyway, i guess using the /proc/<id>/cmdline to check if the process is at least similar to what is expected is the best way to be safe
<bougyman>
there is no such thing, accd to the manpage.
<bougyman>
can you show me how you got a thread id in /proc/ ?
<Defusal>
2.6.32-29-server
S2kx has joined #ruby-lang
<bougyman>
see man proc on that.
<Defusal>
err wait
<Defusal>
that is not the right version
<bougyman>
eevry dir in /proc/ that's numeric is a PID.
<Defusal>
i used the wrong ssh window :P
<Defusal>
3.0.0-16-server
<bougyman>
yeah, even on 3.3
<Defusal>
bougyman, as i said
<Defusal>
i can simply cat the id
<Defusal>
bougyman, you realize threads have their own "pid"
<bougyman>
03/27-16:19 < Defusal> bougyman, you should really not just assume stuff and ignore what people say :(
<Defusal>
htop shows them unless you hide userland threads
<bougyman>
i was trying not to assume, but I can't find anything about it in my manpages.
<rue>
Threads do NOT have process IDs.
<bougyman>
Defusal: inside each pid is a num_threads.
<bougyman>
THOSE are the threads of the process.
<Defusal>
bougyman, well, where does htop get the pids of threads from
<Defusal>
and how come i can cat them from /proc?
<Defusal>
:P
<bougyman>
Defusal: you're mistaking something for threads that aren't threads, but this is getting silly.
<Defusal>
bougyman, sigh.
<rue>
If you ask a thread for its pid (i.e. getpid()), it'll give you the PID of the process to which the thread belongs
<bougyman>
i'm driving home, feel free to research (with man proc)
<Defusal>
no comment
<Defusal>
i have given you all the evidence i have
<rue>
XY
S1kx has joined #ruby-lang
<Defusal>
if you still don't believe me, then good luck when you one day come across such an issue
<Defusal>
but as i said, fire up htop, show and color userland threads, and copy one of their ids
<Defusal>
then cat /proc/id/cmdline
<Defusal>
if it doesn't work for you, something in newer kernels has changed
<Defusal>
but chances are if 3.0.0 does it, other do to
<Defusal>
too*
<Defusal>
either way, im out for the night, cheers
<rue>
Alright, well, I'm just going to invite you to reassess your situation. Threads are not processes, and they're in no way interchangeable.
<Defusal>
rue: you sir, are wrong.
<rue>
I don't typically use the term, but whatever.
<Defusal>
i made it very clear that these ids show in htop as userland threads, give a EPERM error when you try kill -0 them from a different user, and do not show in ps ax at all, ever
<Defusal>
if you are so sure they are not threads, please for the love of god tell me what they are
<Defusal>
but until you can do that, you are wrong
* Defusal
out.
<rue>
I'd be more inclined to think that htop is busted.
<rue>
Or, you know, you're interpreting it wrong.
burgestrand1 has joined #ruby-lang
dan00 has joined #ruby-lang
S2kx has joined #ruby-lang
<erikh>
are you on freebsd?
pygmael has joined #ruby-lang
<erikh>
oh nevermind. I don't even want to talk about this.
publicvoid__ has joined #ruby-lang
<imperator>
yes, let us discuss other things
florentg2 has joined #ruby-lang
<imperator>
happy things....as if bob ross were writing code....happy little ruby code....in pthalo blue
<imperator>
no bugs, just happy accidents
<rue>
Happy little hashes over here
<Defusal>
rue, erikh and bougyman -- i apologise my edgyness today.. i am rather heavily sleep deprived at the moment
<rue>
Sleep would probably help.
<Defusal>
long story, but i get little sleep as is because of my hectic schedule and insane workload
<Defusal>
but i ran out of my prescription med over the weekend, and underestimated the effect of the full dose after a few days break
pygmael has joined #ruby-lang
<Defusal>
so on top of my sleep deprevation, i have had insomnia for a little more than 24 hours
<imperator>
coding on little sleep == bad idea
<Defusal>
luckily the drug negates almost all fatigue, so cognition is ok.. but edgyness remains :P
<Defusal>
yeah, if it were not for the drug, i would not be productive at all imperator
<imperator>
or writing code for too long; i quit at 10 hours - i'm counterproductive after that
<Defusal>
i have gotten a lot done today though
<Defusal>
on a normal day with my current schedule, i work from 8am till 11pm
<erikh>
I take breaks throughout the day
<imperator>
i can't work that long
<Defusal>
with a few brief breaks for snacking and lunch
<imperator>
that's not really very healthy
<erikh>
^
<Defusal>
but i have virtually no fatigue
<Defusal>
that will change in a few hours time though
<Defusal>
indeed it is not
<Defusal>
but with my current workload, i have little choice
<imperator>
i'd refuse to work at that place if that was a regular occurrence
<Defusal>
i have had to put all my courses on hold and will have to catch up and complete them late
<Defusal>
because i have to somehow finish this platform before the end of the month
<Defusal>
imperator, yeah well, in life, sh*t happens
<Defusal>
sometimes stuff just has very bad timing
<RickHull>
also, i just enabled JS so maybe links were hidden
<ged>
No, I mean that I haven't linked to them yet, but I should.
<ged>
:P
slyphon has joined #ruby-lang
<RickHull>
ah, righto. thanks. unitialized constant PG is def. an issue for me though, right?
<RickHull>
i put in a support ticket
<RickHull>
to my host, i mean ;)
<RickHull>
i do have Object::PGError
IPGlider has joined #ruby-lang
<RickHull>
maybe it's not hooked into the underlying postgres lib correctly?
<RickHull>
i know this is not "your" issue, just trying to make sure i inform joyent correctly
wyhaines has joined #ruby-lang
<ged>
RickHull: The older API had three toplevel constants: PGconn, PGresult, and PGError.
<ged>
RickHull: I recently put everything into a PG namespace with backward-compatibility aliases.
<ged>
So if you're running pg < 0.13.0, it's PGconn and PGresult.
<ged>
And I've added links to the API docs and the sample/ directory to the overview page. Thanks for pointing that out.
<imperator>
RickHull, dbi!
<ged>
imperator: :P
<imperator>
although....probably won't work with pg any more if pg changed, now that i think about it
jkyle_ has joined #ruby-lang
<ged>
imperator: It should, and I'd consider it a bug if it doesn't.
<jkyle_>
how do I get help for a method/class in the repl?
<imperator>
ged, well, i believe in using an abstract interface when possible, which is usually
<ged>
jkyle: Which repl?
<jkyle>
irb or pry
<RickHull>
imperator: i have used ruby dbi in the past. erikh has worked on it
<ged>
imperator: Me too. I use Sequel way more than 'pg' directly.
<RickHull>
and he recommended pg ;)
<imperator>
oh, hah
<RickHull>
ged: the gem is 0.10.0
<ged>
RickHull: Yeah, so you'll need to use PGconn.
<RickHull>
ok, rereading, got it
<imperator>
just saw that erikh was using sequel and was getting frustrated with it; i encouraged him to revive rdbi
<RickHull>
checking
<RickHull>
cool, works
<ged>
Method API is mostly the same minus a few newer methods, though. PG::Connection -> PGconn.
mikeric has joined #ruby-lang
deryl has joined #ruby-lang
<ged>
jkyle: For method docs in 'pry', you can do: show-doc <method>
<jkyle>
ah, there's not a built in method? like help(foo_obj)
<ged>
jkyle: No.
<jkyle>
ged: that would explain my inability to find it hehe. thanks!
rue has joined #ruby-lang
<ged>
imperator: Oh, interesting. I absolutely love Sequel. It's one of my favorite Ruby libraries.
<imperator>
i haven't tried it; i've either been using AR or DBI
<rue>
Hrm. Flaky net
<ged>
imperator: Did he post something about his frustrations?
<imperator>
tweeted a brief note
<imperator>
erikh, comments?
<ged>
Ah, okay.
<jkyle>
doesn't seme to work for classes. So is ruby-doc is where everyone goes for documentation?
ilyam_ has joined #ruby-lang
<RickHull>
imperator: regarding dbi type stuff. unless you anticipate actually switching, it seems like a bad tradeoff. you are handicapped from using db-specific features
<injekt>
I use ri more than online documentation
<RickHull>
and if you accidentally manage to use db-specific features, switching off may not be so straightforward
<ged>
I use 'gem server' for in-browser docs.
<jkyle>
ri Array
<jkyle>
Nothing known about Array
<injekt>
so generate it
uniqanomaly_ has joined #ruby-lang
<deryl>
injekt: i got verbally looked at like iw as nuts for preferring ri over gem server or the online docs.
<slyphon>
i think i'm gonna get an extra cpu core put in just for jruby startups
<injekt>
deryl: I'm with you, bro
<slyphon>
THE FIVE CORE INTEL! <monkey-grunt>
<ged>
Hahaha, "verbally looked at"?
* deryl
plays slyphonfreeswitch's Screaming Monkeys
<imperator>
slyphon, what are you trying to say?
<slyphon>
hah
<injekt>
ged: like.. when I say I'm looking at you
<injekt>
.. or something
<RickHull>
is there no way to treat the JVM like a normal machine? i mean, a normal machine has to boot and stuff. but once its up, any number of resident programs can use the machine "instantly"
<deryl>
ged: hehe only way i could put it :)
<slyphon>
imperator: oh, i'm just bellyaching
<injekt>
:D
* slyphon
hugs jruby
<slyphon>
jruby, i wish i could quit you
<corsican>
haha
* corsican
pictures slyphon and jruby having gay sex in a tent
<ged>
deryl: Synaesthetic! :P
<slyphon>
oh ffs
<slyphon>
corsican: i'd hit it
<corsican>
me too
<corsican>
actually I admit I don't understand jruby
<slyphon>
Gemfile.lock records your *sources
<deryl>
lol
<slyphon>
*
<slyphon>
that SUCKS
<corsican>
that's basically a ruby interpreter that runs in java?
<slyphon>
GAH
<slyphon>
it's a ruby interpreter that runs on top of the JVM, and is AWESOME
<corsican>
oh cool
<deryl>
initial startup time sucks, but after that its pretty decent
<slyphon>
it's the most performant ruby in some benchmarks, has grown-up GC, multiple threads for real
<slyphon>
yeah
<slyphon>
it's great for servers
<imperator>
so, uh, spin up a jruby machine in drb/rinda and use that? is that possible?
<slyphon>
nah
<injekt>
it's amazing for long running processes, as is the jvm in general
<corsican>
nice :)
<corsican>
thanks
<slyphon>
there's nailgun
<corsican>
I will have to look into it
<slyphon>
but i never figured out how to use it
<slyphon>
i'm just horseshittin' though, the startup time only really sucks when it comes to specs
<slyphon>
god *dammit*
<slyphon>
why does Gemfile.lock record your sources?
<slyphon>
that's so goddamn stupid
<slyphon>
grrrrrrr!!!!
<ged>
I think I still have some leftover distrust of 'ri' from back when it went through a period of bitrot.
<slyphon>
yeah
<ged>
I should get over it, though. It's pretty good nowadays.
<slyphon>
i got it to work for a while, then it sucked pretty bad
<drbrain>
ged: I fixed many of the things
* slyphon
waves to drbrain
<ged>
drbrain: Yes, I know! Many thanks!
<ged>
Like I said, I should get back into the habit of using it.
<drbrain>
I committed markdown support to rdoc recently
<slyphon>
woah, nice
<ged>
Heh, I saw.
<drbrain>
… it needs battle testing still
<RickHull>
i am a markdown True Believer
<RickHull>
markdown all the things!
<ged>
With a new parser, right?
<slyphon>
programming is not religion
<drbrain>
it's a port of peg-markdown
<slyphon>
i'm still waiting for matz to sell out
<slyphon>
AND THEN YOU'LL ALL SEE!!
<RickHull>
good thing neither markdown nor markup are programming
* slyphon
adusts his tinfoil hat
<drbrain>
oddly, the README and ruby is not in any particular text markup format
<slyphon>
RickHull: code is data! data is code! the lamb lies down with the lion!
<ged>
Ah, right. I saw the .kpeg files.
<imperator>
death to markdown!
<slyphon>
hah!
<RickHull>
the files are *IN* the computer
<slyphon>
:D
<drbrain>
I half expected it to be in rdtool format, but it likely predates even rd (let alone rd2)
<ged>
imperator: Hehe.
<ged>
Long live rd2!
<ged>
Hehe.
<imperator>
you know what works as a pretty good markup language?
<imperator>
HTML THAT'S WHAT
<deryl>
something i think *might* be nice is something like pinfo for ri. pinfo did colorized links to related info pages, made traversal to those pages easy, and was definitely a better info.
<ged>
LateX?
<slyphon>
imperator: eh
<drbrain>
HTrollML
<RickHull>
XHTML is the new hotness, i heard
<ged>
Errr... LaTeX>
<mahlon>
imperator: Oh, whew. I thought you were going to say 'nroff'.
<imperator>
ged, yes, well, if i'm typesetting, sure
<slyphon>
i find html to be pretty tiring
<deryl>
doubt ri would go that direction, but for info vs. pinfo, pinfo definitely won
<drbrain>
deryl: see rdoc-browser
<drbrain>
(on github)
<slyphon>
i had a friend-of-the-family that used to call them 'hotmail' files
<deryl>
will do
<slyphon>
that still makes me chuckle
<RickHull>
ha
<imperator>
slyphon, having to figure out how to get markdown et al to do things i can easily do in plain html is very tiring
<drbrain>
… I already wrote it as a proof of concept
<slyphon>
imperator: agreed
<RickHull>
aitch-titty-pee
<slyphon>
hahahah
<ged>
There was an HTML editor for a while called "HoTmeTaL". Still makes me chuckle.
<mahlon>
HoTMaiL. Not coincidence.
<slyphon>
i'm such a dunce when it comes to HTML, i find if markdown can't do it, i probably didn't need it anyway
<drbrain>
"huh-tup" for "HTTP" (courtesy zenspider)
<slyphon>
dunce is probably the wrong word, i'm completely apathetic
postmodern has joined #ruby-lang
<RickHull>
markdown nails the 80/20 pretty well for documents
<drbrain>
or maybe "hut-tup"
<slyphon>
yeah
<slyphon>
we need an 'uh-duh' protocol
<slyphon>
uddh or something
<ged>
And so many markdown converters to choose from! *snerk*
<slyphon>
oh gawd
<slyphon>
yeah, seems like YARD is always picking a different one
<slyphon>
AND NOT INSTALLING IT
* slyphon
shakes his fist
<slyphon>
"If you want to parse markdown..."
<slyphon>
no, i'd rather just have you bitch at me
<slyphon>
ok, bbl
* ged
giggles.
Joeysomo has joined #ruby-lang
<musl>
My new daemon 'cloudmemd' speaks TNHP (the new hottnes protocol)!
Joeysomo has joined #ruby-lang
<drbrain>
I had to write a new one because there wasn't one that was BSD/MIT license, maintainable and could produce a syntax tree ☹
imperator has quit ["Leaving"]
<drbrain>
well, port
<deryl>
drbrain: ok, so i'll wait for 2.0 to be released :)
<injekt>
lol
<drbrain>
deryl: if I have enough spare time I'll finish keyboard navigation in rdoc-browser and release a curses-newpad for non-ruby-2.0 users
<drbrain>
with ruby 2.0 only a year away, I don't know if I'll have that spare time, though
<deryl>
doh!
postmodern has joined #ruby-lang
ilyam has joined #ruby-lang
<deryl>
actually suprised that this even exists. my info/pinfo idea just came off the top of my head thinking off the cuff about what other doc viewers i liked for console users.
<deryl>
nice to see something along that vein
<drbrain>
I wrote it as an experiment in curses
<deryl>
got ya
<drbrain>
lack of complete support for modifier keys was a very annoying discovery
<drbrain>
so if/when I release it I'll probably get many complaints over the "clumsy" keyboard commands
<deryl>
hehe usually the way. get the blame for things not really your failing.
<Defusal>
i hope ruby is optimized for ARM at some stage
<Defusal>
that will be great
<drbrain>
Defusal: like mruby?
<Defusal>
yeah
<Defusal>
i didn't even know about mruby until recently
ascarter has joined #ruby-lang
chimkan has joined #ruby-lang
<Defusal>
but apparently MRIs performance on ARM is not even as good as 1.8 on x86
fayimora has joined #ruby-lang
<Defusal>
with the ability to run standard linux distros on inexpensive ARM based hardware, it would be awesome to be able to use ruby
<deryl>
linux ARM-based router, using nginx+unicorn+rails+ruby1.9.x for the interface? or a ruby based control infrastructure accessed via ssh? :)
<Defusal>
oh there are almost unlimited use cases :P
<deryl>
Defusal: yep. friend of mine bought a usb stick looking thing thats basically a little system. miniHDMI on one end, usb on the other. slightly bigger than a U3 thumbdrive but not by much at all. probably based on something like this.
<deryl>
never saw it hooked up to anything but he says its pretty much a little system. the hdmi output, usb for attaching a keyboard (he says he's using a usb hub to attach both a keyboard and mouse), has like an 8GB flash built into it. haven't seen it work, but if its tue, damn!
<Defusal>
you can do a lot with ruby and a $35 micro pc that can play blueray videos and quake 3, has usb and ethernet and an array of General Perpose IO pins
<lianj>
if the rpi was actually available ;)
<Defusal>
it got delayed again due to an issue with the ethernet, but it'll be available eventually :P
<drbrain>
I found it appalling that my old G4 PowerMac could play 720p video but not a 320x240 youtube stream
<deryl>
Defusal: nice :)
<drbrain>
was it really that hard for flash to use the video hardware?
<Defusal>
still a lot sooner than MRI being optimized for ARM :P
<jmontross>
so you can get ruby on that little raspberrypi ?
<lianj>
sure
<deryl>
i'd spend the 35 bucks to play with that
<Defusal>
jmontross, if ubuntu, redhat and gentoo are already supported for it by the company producing it, why not
<Defusal>
it may be easier to optimize rubinius for ARM though, since its core is small