<shevy>
but remember, you are on debian so DO NOT TRY TO USE THE SOURCE
<manhunter>
libyaml is already installed
<manhunter>
shevy: why not installing from source?
<shevy>
because this is debian. if you keep the old ruby, you are in a world of pain
<shevy>
if you remove the old ruby, perhaps ruby will never work again
<shevy>
hmm
<manhunter>
can i install ruby-1.9.3 to a seperate directory to see if it's installed successfully or not? NOTE: I have already installed system libruby1.8 and libruby1.9.1 and i have other ruby installed with rvm
<shevy>
do you have a libruby.so* ?
<shevy>
you can install ruby-1.9.3 to a separate directory by using --prefix
<shevy>
that's a lot of ruby you got there
<shevy>
when you have installed 10 rubys you will get a prize
<manhunter>
./configure --prefix /path/to/directory; make ; make install , is that ok?
<shevy>
yes
<shevy>
well
<shevy>
./configure --prefix=/path/to/directory
<manhunter>
shevy: prize? what prize?
<shevy>
the prize of "I have the most rubys installed!!!!"
<shevy>
and this prize is coupled with
<shevy>
"I have no idea which ruby I am using right now"
woollyams has joined #ruby-lang
<manhunter>
why not?
<manhunter>
if i install with ./configure --prefix=/path/to/dir then can't i use that ruby from /path/to/dir?
macmartine has joined #ruby-lang
<manhunter>
i just ran, ./configure --prefix=/home/myaccount/sepdir ; make ;
<manhunter>
it's compiling
<shevy>
you can use that ruby
<shevy>
you'd either have to call the binary directly (with the full path to it) or modify your $PATH variable so that it is listed first
<shevy>
or
<shevy>
you can put a symlink at /usr/bin/ruby too
<shevy>
but this one probably points to your debian one
<shevy>
the /usr/bin/ruby1.8
<manhunter>
ok, i understand that
<manhunter>
during compiling, i got a line like, Failed to configure fiddle. It will not be installed.
<manhunter>
what is that for? is that a problem/error?
<manhunter>
during make
<manhunter>
^
<shevy>
fiddle?
<shevy>
I dont know what is fiddle
<shevy>
never had such an error
<shevy>
it's best to paste what the error is
<manhunter>
during make, i got that, Failed to configure fiddle. It will not be installed.
<manhunter>
but it's compiling other files successfully
<shevy>
did it continue?
<shevy>
I really have no idea what is fiddle
<manhunter>
yes, it continues
<shevy>
that is usually a good sign
<manhunter>
~MRI
<manhunter>
!MRI
<manhunter>
MRI = what ruby interpreter?
<manhunter>
manveru: MRI
<shevy>
matz ruby
<shevy>
the main ruby
<manhunter>
for the first time i compiled ruby by hand doing, ./configure --prefix=/home/myaccount/sepdir
<shevy>
good
<manhunter>
Total: 15386 (5195 undocumented)
<manhunter>
66.24% documented
<manhunter>
Elapsed: 175.2s
<manhunter>
do i have to rung sudo make installl --prefix=/home/myaccount/sepdir too?
<manhunter>
or only sudo make install
<shevy>
the latter
<shevy>
"sudo make install"
<manhunter>
ok
<manhunter>
it's done
<manhunter>
i have installed ruby-1.9.3 by hand for the first time. yahoo
<manhunter>
this time i didn't use RVM
<shevy>
try to run the irb
<shevy>
./home/myaccount/sepdir/bin/irb
<manhunter>
i got these directories there, "bin include lib share"
<shevy>
if irb works, chances are high that your compiled ruby will work too
<shevy>
I did not say anything about using = 1.9.3
<manhunter>
1.9.2-p180 :001 > puts RUBY_VERSION
<manhunter>
1.9.3
<manhunter>
=> nil
<shevy>
????
<erikh>
do you have a ~/.irbrc
<erikh>
are you using rvm?
<shevy>
well, whatever it is, I think it knows that it is the right version to use
<manhunter>
erikh: no
<erikh>
you aren't using rvm at all
<erikh>
it isn't loaded on your system at all.
<manhunter>
but i have another ruby installed by RVM
<erikh>
right.
<erikh>
ok
<erikh>
fun time with environment variables
<cout>
manhunter: do you have a $RUBYOPT set?
<erikh>
you need to get your rubies managed better.
<shevy>
hehe
<shevy>
I told him!
<shevy>
shevy> "I have no idea which ruby I am using right now"
<shevy>
that will be his famous quote
<erikh>
well beating him about isn't going to help
<erikh>
anyhow -- stick to one thing. either you use individual compiles, rvm, rbenv, whatever
<erikh>
just pick one thing and stick to that.
<erikh>
they all piss and moan in their own unique ways and you're creating a mess having them all mixed in like that.
<erikh>
rvm injects its own .irbrc's
<manhunter>
cout: echo $RUBYOPT displays nothing
<erikh>
to add the prompt with the version in it.
<manhunter>
but , cd /home/myaccount/sepdir/bin; ./ruby --version = ruby 1.9.3p125 (2012-02-16 revision 34643) [i686-linux]
<manhunter>
that is ok
<manhunter>
if i do , ruby --version = ruby 1.9.2p180 (2011-02-18 revision 30909) [i686-linux] , this is the RVM ruby
<erikh>
you're missing hte point. rvm messes with more than your rubies
<erikh>
right
<manhunter>
yes
<erikh>
that's what's showing up in your prompt, right?
<erikh>
that's because rvm messes with a part of irb.
<erikh>
hence my comment about sticking to one thing -- you're making this harder on yourself.
<manhunter>
erikh: but ./ruby --version show the correct version of ruby
<cout>
oh that makes sense
<manhunter>
erikh: problem is with only irb
<erikh>
arguing in circles doesn't change anything
<manhunter>
so what to do to get the irb i just installed from source?
<erikh>
if you really want to fix it, you'll either have to remove rvm or have rvm install that ruby for you
<erikh>
I'm not giving you the command to do that because I don't have children for a reason
<manhunter>
i can't remove RVM,
<cout>
erikh: so it sounds like rvm is slightly confused as to which version of ruby manhunter is running, and rvm's modifications to the .irbrc file result in an incorrect prompt?
<manhunter>
i need that RVM too
<erikh>
cout: exactly
seanstickle has joined #ruby-lang
<manhunter>
where is the .irbrc?
wmoxam has joined #ruby-lang
<erikh>
manhunter: it swaps them out depending on what version of ruby you're using
<erikh>
that rvm thinks you're using, that is
<erikh>
this *will not* work the way you want it to
<erikh>
so, rvm implode or rvm install 1.9.3 and use that copy of 1.9.3
<cout>
that sounds like a bug IMO
<erikh>
cout: it's not, it's rvm
<manhunter>
cout: bug for what?
<erikh>
rvm takes over your whole shell environment
<cout>
surely there's a better way to inject settings than modifying a user's config files
<manhunter>
cout: is it RVM bug?
<erikh>
cout: you can modify them them per-ruby
<erikh>
IIRC, it just shuffles a symlink.
<manhunter>
erikh: how?
<erikh>
manhunter: rvm.beginrescueend.com
<erikh>
go read.
<rue>
cout: There isn't a better way to inject *permanent* settings
<cout>
erikh: I can already do that without using rvm
<cout>
if RUBY_VERSION == '...' then ...
<rue>
I think there was some talk about letting the user add the requires if they want (but since pretty much everyone does…)
<erikh>
there is the capability to do it
<cout>
rue: maybe I just don't like it because it's not the way I would have implemented it :)
<rue>
How would you have done it?
zmack has joined #ruby-lang
Karmaon has joined #ruby-lang
<erikh>
cout: that's all fine and dandy until you deal with something that doesn't parse in 1.8
<cout>
rue: (and I always implement it the best way, of course)
<erikh>
heheheh
<cout>
erikh: if RUBY_VERSION =~ /^1\.9/ then require '.irbrc-1.9' end
<erikh>
ok, fair
<cout>
erikh: basically I would have solved the problem by monkey patching irb to load a different (or both) config files
<cout>
which I guess carries its own set of problems
<erikh>
yep
<erikh>
rvm gets enough shit for what it does to ruby already
<manhunter>
how can i get the irb-1.9.3?
<erikh>
did you read about rvm yet?
<manhunter>
i installed ruby from source and have RVM ruby
<cout>
(actually I would never have thought to write rvm in the first place, because ./configure --program-suffix= or in some cases ./configure --prefix= is sufficient for what I need to do)
Karmaon_ has joined #ruby-lang
brownies has joined #ruby-lang
<shevy>
I told him he is in a world of pain
<erikh>
yeah I don't believe he's going to listen to anything we say
wmoxam has joined #ruby-lang
<erikh>
but he keeps making my irc client jingle so it's hard to ignore
<shevy>
lol
<manhunter>
erikh: whereis irb = irb:
<cout>
jingle?
<erikh>
yep
<cout>
erikh: like this?
<erikh>
pretty much
<manhunter>
erikh: but if if use this, IRB.conf[:PROMPT_MODE] = :DEFAULT in .irbrc
<erikh>
probably will do what you want
<manhunter>
erikh: then i see , whereis irb = /home/myaccount/.rvm/rubies/ruby-1.9.2-p180/bin/irb
<cout>
yeah
<erikh>
that's becaues
<erikh>
for the last time
<erikh>
rvm messes with your shell
<manhunter>
erikh: but if if use this, IRB.conf[:PROMPT_MODE] = :DEFAULT in .irbrc , then i see, whereis irb = /home/user1/.rvm/rubies/ruby-1.9.2-p180/bin/irb
<erikh>
that has nothing to do with IRB
andkerosine has joined #ruby-lang
<erikh>
that has to do with your system path
<erikh>
which rvm manipulates
<manhunter>
erikh: if i dont use ~/.irbrc which irb does the system use?
<erikh>
you're confusing two topics
<erikh>
how long have you been using unix?
<manhunter>
shy
<cout>
my guess it not long if he didn't know what a shebang is
<manhunter>
erikh: if i use .irbrc and then if if type irb, i get , irb(main):001:0>
<erikh>
yes
<erikh>
but...
<erikh>
answer the question please
<shevy>
haha
<manhunter>
erikh: which question?
<erikh>
you have some learning to do before you get to this point
<erikh>
how long have you been using unix:
<rue>
erikh: bells
<erikh>
rue: die
<rue>
erikh: all the way
<manhunter>
erikh: how long have you?
<erikh>
a while.
<cout>
lol rue
<erikh>
long enough to know what your problem is
<shevy>
stop jingling with erikh!
<manhunter>
erikh: i know you are pointing me to the path problem , right?
* cout
was borne with Bourne shell syntax as a birthmark on his right buttcheek
<erikh>
point is, you need some time to learn your system before this will make sense to you
<shevy>
some things are best kept a secret
<erikh>
are you more familiar with windows?
<erikh>
ruby has windows versions.
<manhunter>
erikh: i'm also using linux for a while
<manhunter>
using ruby on windows is more painfull
<erikh>
you are intentionally making this harder on yourself by insisting you use a separate compile along with rvm.
<erikh>
now, I can't really help you anymore -- either kill rvm or your compile. you pick.
<cout>
I agree
<cout>
manhunter: can you use two different user accounts, one for rvm and one for this other install?
<manhunter>
cout: ah, that's another option,
<manhunter>
cout: good
QaDeS_ has joined #ruby-lang
<twittard>
Test::Unit fans, question: It seems that "_wrap_assertion" was deprecated as of ruby 1.8.7. What was it replaced with?
<manhunter>
when would we get ruby-2.0 ?
<shevy>
you have to ask matz
<cout>
manhunter asked the rite question :)
SuperTaz_work has joined #ruby-lang
<twittard>
Seems like it'll happen around Christmas, either this year or next.
<twittard>
I'm sure ruby-core is furiously writing C. lulz.
andrewhl has joined #ruby-lang
<cout>
twittard: we thought that 8 years ago too
<shevy>
hehehehe
<cout>
personally I like the Duke Nukem Forever strategy
<twittard>
DNF is out.
<twittard>
Ruby 2 and Perl 6 are in furious competition
<manhunter>
cout: :)
<manhunter>
twittard: yes
slimfit has joined #ruby-lang
<manhunter>
cout: Duke Nuke??
<twittard>
Anyone know much about Test::Unit? I'm trying to fix up a library that was written for Ruby 1.8 .. to run in 1.9.
<twittard>
It's using "_wrap_assertion" but that method was deprecated as of 1.8. I don't even know why it exists, tbh.
<twittard>
Hmm, I just commented them out and I have more tests passing than before.
<twittard>
Looks like it was just decoration. Moving along.
manhunter has quit ["WeeChat 0.3.6"]
<twittard>
Sweet. Who knew that deleting stuff could make tests pass.
<shevy>
lol
<shevy>
I should really start a collection of quotes from #ruby-lang
<twittard>
@shitrubyistssay
<rue>
chris2 had a collection at some point, or maybe that was just on Anarchaia
nedbat has joined #ruby-lang
twelvechairs has joined #ruby-lang
Roujo has joined #ruby-lang
<Harzilein>
hmm
<Harzilein>
does pp not include a way to print the object name as well?
<twittard>
object name? you mean, class?
<twittard>
If it's a class, it'll respond to .name.
SuperTaz_work has joined #ruby-lang
<twittard>
If it's an instance, call .class.
<twelvechairs>
object_id?
<Harzilein>
twittard: i mean that i want pp to print the instance name as well (it's a simple hash)
<Harzilein>
twittard: not how to determine the instance name
<heftig>
instances don't have names
<twittard>
right
<twittard>
You can dump out the object_id if you care
<heftig>
a = Object.new; a is just the name of a variable referencing the object
<heftig>
it's not a property of the object itself
<deryldoucette>
there was a way someone showed me to show that difference in irb, and i forget how he did that
<Harzilein>
i want to set pp up so it when i run foo = { fooblah: "bar" } ; pp foo it prints foo = {:bar=>"baz"} where it now prints just {:bar=>"baz"}
<Harzilein>
should be easy to understand what i want
<heftig>
puts "foo = #{foo.inspect}"
<Harzilein>
-it
<twittard>
You'll have to do that yourself.
<Harzilein>
heftig: -.-
<deryldoucette>
Harzilein: pp doesn't do that
<twittard>
puts "foo = #{foo.pretty_inspect}"
<twittard>
I know what you're asking, but it's just not common for languages to support this.
<Harzilein>
deryldoucette: that's why i want to set my object up so pp _does_ print the name.
andrewhl has joined #ruby-lang
<deryldoucette>
Harzilein: umm, see what twittard said
<Harzilein>
twittard: not common?:
<deryldoucette>
and heftig
<twittard>
Harzilein: Perl, Python, Ruby, etc, do not bind the name of an object to the instance of the object.
<Harzilein>
well, i would accept references
<twittard>
You can go `a = b = "foo"` etc. You can pass the object around all over the place. 'a' and 'b' are only for your reference.
<heftig>
def xp(var, &blk); puts "#{var} = #{blk.binding.eval(var.to_s).pretty_inspect}"; end
<twittard>
accept references?
<heftig>
xp(:foo){}
<heftig>
still ew.
<twittard>
What you are asking for just doesn't have any value in a programming language. It'd be a complicated mess to implement, and would end up slowing down the language.
<twittard>
Besides that, when you call "pp" on something, you know what the name is anyway.
takaokouji has joined #ruby-lang
<twittard>
Programming is hard, let's go shopping.
<h4y4shi>
whats the difference between a class variable @@foo and an instance variable @foo
<rippa>
h4y4shi: @@foo is class variable
<rippa>
and @foo is instance variable
<rippa>
don't use class variables
<Mon_Ouie>
Class variables are shared with a class, its instances, and its subclasses
<Mon_Ouie>
And the scoping rules from them is kind of similar to those that apply with constants
<Mon_Ouie>
for them*
<h4y4shi>
Thanks
pp01bit has joined #ruby-lang
hexo has joined #ruby-lang
macmartine has joined #ruby-lang
davidw has joined #ruby-lang
bjensen has joined #ruby-lang
nanothief has joined #ruby-lang
<nanothief>
If I have a string like "one two \"word three\" 'word four'", what would be the easiest way to convert that to ['one', 'two', 'word three', 'word four']?
mytrile has joined #ruby-lang
macmartine has joined #ruby-lang
vdrab has joined #ruby-lang
cmatheson has joined #ruby-lang
<TTilus>
nanothief: .gsub(/["']/, '').split
<Mon_Ouie>
I'd use delete '"\'' over gsub
<Mon_Ouie>
But it doesn't look like your solution would work anyway
<Mon_Ouie>
I'd use Shellwords.split (after require 'shellwords'), because I'd be too lazy to do it myself
<TTilus>
nanothief: uh, sorry, better use .scan /\S+|"[^"]+"/
<epitron>
hmm.. is there an easy way to make a Singleton class automatically redirect class methods to instance methods?
<Mon_Ouie>
That works, as long as you don't have escaped quotes inside the quoted words
<TTilus>
nanothief: shellwords would be best as Mon_Ouie said
<epitron>
i guess that would work tho :)
<Mon_Ouie>
Other solutions I can think of would involve changing the code that defines the method
<Mon_Ouie>
Like using a method that you pass a block to and automatically defines the proxy method, or calling def_delegator (or similar) after each method
<Asher>
would would you have instances of a singlteton class
<Asher>
do you just mean redirect class methods to instance methods?
<Asher>
how does it know which instance?
<epitron>
so ruby's stdlib has a mixin called Singleton
<epitron>
and MyClass.instance always returns one instance of that class
<epitron>
the same one
<epitron>
it's magic :D
<Asher>
yea why would you use that lib instead of just a module
<epitron>
that question doesn't make sense to me
<Asher>
a module is already a singleton
<Asher>
that you can't create additional instances of
<epitron>
i don't want additional instances
<Asher>
exactly
<epitron>
oops, misread :)
<Asher>
so what does the singleton mixin do
<Asher>
that using a module doesn't
<epitron>
but how would i use the module?
<Asher>
Module.method
<epitron>
but then i'd have to make class methods for everything
<Asher>
MyModule.method
<Asher>
so?
<epitron>
and how would i store ivars?
<Asher>
in the obj
<epitron>
well, it's ugly!
<Asher>
why's it ugly
<Asher>
use a mixin to define your methods
<epitron>
def self.everything
<Asher>
so use a module to define methods and extend your singleton
<Asher>
that's better practice anyway
<epitron>
o_O
<Mon_Ouie>
And in that case, do *not* use module_function
<Mon_Ouie>
Because presumably your methods will rely on self if you thought about using Singleton in the first place
<Mon_Ouie>
Which means "include Foo; bar" wouldn't do the same thing as "Foo.bar"
<Asher>
doesn't the singleton mixin create an instance of your singleton class
<Asher>
so it actually ends up with a doubleton?:P
<epitron>
i assume .instance is like: @instance ||= self.new
<epitron>
i didn't look at the code
<Mon_Ouie>
More like @instance ||= new, they make new a private method
<epitron>
ok :)
<epitron>
i still don't see how using a module is nicer
<epitron>
like, one nice thing about the singleton is that i have an initialize method that gets run after the whole app has loaded
<epitron>
when the singleton is first accessed
<epitron>
instead of a module, where i'd need an .init method or something
<epitron>
lemme make it more concrete
<epitron>
i have a global Settings class that loads a config
<epitron>
for testing, it can be instantiated manually
<epitron>
with special parameters
<Mon_Ouie>
You don't need an init method
<Mon_Ouie>
Well, that depends, actually
<epitron>
i need to load the configfile :)
<Mon_Ouie>
Sometimes you can just call it in the module body
<epitron>
if i did that, then i couldn't test it
<epitron>
the configfile would be hardwired
hexo has joined #ruby-lang
<epitron>
i suppose i could be like, Settings.load
<epitron>
or Settings.load("otherfile")
<epitron>
that's a bit weird though, because if something tried to access it before it was loaded.. it would break
<epitron>
it's a global class, see
<nanothief>
Mon_Ouie: Shellwords.split is exactly what I wanted, thanks!
Para has joined #ruby-lang
wmoxam has joined #ruby-lang
Banistergalaxy has joined #ruby-lang
<Asher>
you can always test stuff in the module/class body by using #clone
apeiros_ has joined #ruby-lang
Jade has joined #ruby-lang
rue has joined #ruby-lang
<rue>
Testability is probably better
<epitron>
is testability a gem? :)
tbuehlmann has joined #ruby-lang
Aaaarg has joined #ruby-lang
rohit has joined #ruby-lang
<epitron>
rue: i'm not sure what you meant there
mssola has joined #ruby-lang
workmad3 has joined #ruby-lang
Aaaarg has joined #ruby-lang
ged_ has joined #ruby-lang
mtkd has joined #ruby-lang
Para has joined #ruby-lang
solars has joined #ruby-lang
Aaaarg has joined #ruby-lang
chendo has joined #ruby-lang
gokul has joined #ruby-lang
bubonicpestilenc has joined #ruby-lang
Para has joined #ruby-lang
Aaaarg has joined #ruby-lang
Banistergalaxy has joined #ruby-lang
davidw has joined #ruby-lang
vdrab has joined #ruby-lang
Aaaarg has joined #ruby-lang
shajahan has joined #ruby-lang
dreinull has joined #ruby-lang
pp01bit has joined #ruby-lang
jtoy has joined #ruby-lang
workmad3 has joined #ruby-lang
ged__ has joined #ruby-lang
<jtoy>
what simple way is there do to get a new hash from an old hash with keys you selected
yugui_zzz has joined #ruby-lang
<jtoy>
like hash.pick(:foo,:bar)
<apeiros_>
jtoy: if you use activesupport: Hash#slice
<apeiros_>
if you don't: Hash[keys.zip(hash.values_at(*keys)]
<jtoy>
apeiros_: currently I am, although I wish i wasn't, thx
<apeiros_>
adding the missing ) is left as an exercise to the reader…
<jtoy>
Hash[keys.zip(hash.values_at(*keys))]
<apeiros_>
exercise completed
<apeiros_>
advance to level 4
<Asher>
is there a good way to start iterating a hash or array from a key/index in the middle?
<apeiros_>
which of the questions do you want answered?
<apeiros_>
(they're unrelated)
rohit has joined #ruby-lang
<Asher>
then both i suppose
<apeiros_>
Asher: what's the middle index of an array with 4 items?
<apeiros_>
also: there's no efficient way to do it with hashes, also before ruby 1.9, there technically is no *middle* of a hash (as they're unordered)
<Asher>
oh i just meant
<Asher>
items not first or last
<Asher>
and 1.9 is current ruby fuck 1.8 :P
<Asher>
so hashes are ordered
<Asher>
so like say i want to iterate from index #2 to the end of array
<apeiros_>
ary[2..-1].each
<Asher>
should i just slice it?
<Asher>
ok
<apeiros_>
hash.each.drop(hash.size>>1).each { …
<Asher>
does using the range like that avoid duplicatin gthe array?
<apeiros_>
oh, you don't need the first .each
<apeiros_>
hash.drop(N).each …
<Asher>
but drop will remove them right?
<apeiros_>
no
<manveru>
moin folks
<apeiros_>
drop will drop them from the iteration
<Asher>
by creating a second array
<Asher>
ok
<apeiros_>
no
<apeiros_>
by not yielding them during the iteration
<Asher>
yeah ary[range] creates a new array too
<apeiros_>
.drop(N) has O(1) memory consumption, unlike hash.to_a which uses O(n) memory
<Asher>
ok thanks
<apeiros_>
there's no "too"
<apeiros_>
the two work differently. there is *no* temporary array being created with Enumerable#drop
<Asher>
docs for Array#drop suggests there is
<Asher>
that's wrong?
<apeiros_>
…moment
<apeiros_>
wow… ok, I stand corrected
<apeiros_>
that's fucked.
<apeiros_>
you're correct. drop creates an array. I expected it to return an Enumerator
<Asher>
i guess it creates a new array obj but can create a slice without copying anything until it changes
* apeiros_
thinks that was a stupid choice…
<manveru>
it's non-destructive?
<apeiros_>
returning an enumerator wouldn't be destructive either
<apeiros_>
it'd also scale quite a bit better…
<manveru>
returning a slice doesn't take any more memory, i think
<apeiros_>
only true if you already have an array
<apeiros_>
which ain't a given for Enumerable#drop
<manveru>
oh, it's on Enumerable?
<apeiros_>
yes
<manveru>
well then :)
<manveru>
dunno, i never use that method
<manveru>
but you're right of course, there's no need to return an array
* apeiros_
must finish his stupid language
<apeiros_>
where I can avoid those stupid things and make other stupid mistakes instead!
fayimora has joined #ruby-lang
<apeiros_>
*sob* but I'm still looking for a good object-literal there :-S
<apeiros_>
s/looking/searching/
<apeiros_>
trying to come up with?
|Vargas| has joined #ruby-lang
knu has joined #ruby-lang
<Asher>
object literal?
<manveru>
hm
<manveru>
this is weird
<apeiros_>
a literal for Object instances
<apeiros_>
like javascripts {}
<Mon_Ouie>
Like class, and a set of ivars-value pairs?
<apeiros_>
Mon_Ouie: correct
<manveru>
e = [1,2,3].to_enum; e.next; e.each{|v| p v } # => 1 2 3
<dominikh>
he's looking for syntax for his own language :P
<apeiros_>
atm. I'm pondering [@foo => val] for Object and Classname[@foo => val] for everything else
<manveru>
i would've thunk this would continue iterating from the position after next
<manveru>
apeiros_: as long as you don't use => :)
<cout>
apeiros: why do you think this is a good feature for a language?
<apeiros_>
manveru: d'oh, actually I use :, true
<manveru>
apeiros_: you looked at fancy yet?
<apeiros_>
cout: it makes a couple of things pretty easy.
<cout>
apeiros: but at what expense?
<apeiros_>
manveru: no. I usually click the links you send me… must have missed that one
<cout>
IMO it's too pythonic (everything is a dictionary vs. everything is an object)
<apeiros_>
cout: potential issues from not invoking the initialization, also it means some characters are reserved for that syntax
<apeiros_>
something being "pythonic", "rubyish" or "lispy" or whatever is no measure I care about.
Spooner has joined #ruby-lang
<apeiros_>
either it has a use, makes sense etc., or it doesn't. so the question "at what expense" is spot on. being "pythonic" ain't.
<cout>
give me a better adjective then
<manveru>
well, it's been all tried before
<apeiros_>
it's not a matter of the adjective used
<cout>
I couldn't think of one, which is why I qualified "pythonic" with what followed in parens
<manveru>
everything is an (object|string|table|dict|array|list)
<apeiros_>
being similar to X is not a rationale in favor or against something.
<manveru>
what's left?
<cout>
apeiros: again, you're not understanding what I attempted to say
<apeiros_>
manveru: actually I only distinguish between 2 things: code & data
<apeiros_>
so everything is either data, or code
<manveru>
oh, and haskell with everything is a strange mathematical formula nobody really understands
<cout>
apeiros: being able to initialize an object like that seems like a feature that is best suited for a dictionary-based language (of which python is the best example I know of)
<Asher>
i like the idea that everything is an object and a function
<Asher>
from _why
dreinull has joined #ruby-lang
<manveru>
potion and io?
<cout>
apeiros: I haven't thought it through to the end, but that's my first impression when I think about it
<Asher>
didn't realize io did that too
<Asher>
but potion yea
<manveru>
i like agora, with everything is a message
<manveru>
but that language went extinct
<apeiros_>
cout: you can look at rubys objects as dicts too
<apeiros_>
ivar names are the keys, and their values are, well, their values
<apeiros_>
the distinction between scalars and lists is sometimes difficult
<apeiros_>
you can look at an rgb value as a single integer, or as a structure of 3 values, which could be either an array or a dict… etc.
<cout>
apeiros: you can look at ruby that way, but ruby doesn't expose the dictionary as bluntly as python
<apeiros_>
it's not hard to get there, though
<apeiros_>
cout: anyway, so far I'm missing the rationale - you seem to imply that it's a bad thing. but why is it?
<cout>
apeiros: not bad in terms of pragmatism, just in terms of style
<apeiros_>
ok, what makes it bad style then?
* cout
scratches head
<cout>
I don't mean to say that it is "bad style"
<cout>
I just believe in coherent/consistent style
<cout>
and since I don't know exactly what you are building, I can't say whether that feature is coherent/consistent with regard to everything else you've constructed
<cout>
so I'm not saying it's "bad", just that it needs thought (and I believe you've already thought about it)
<apeiros_>
ok
<apeiros_>
some of that has to show itself when I can finally use the language. may well be that I drop/change a lot of features then :)
<cout>
:)
<apeiros_>
but so far it'd be consistent with the language being very rich in literals already
<cout>
hmm.
<apeiros_>
but granted, I actually only wanted it for "Object" (the base object class), and the "for other classes" came as an afterthought and may very well be feature-bloat
<andkerosine>
Hm... where to put a comment explaining the next few methods that doesn't get set as the description for the first method encountered by rdoc?
m3nd3s has joined #ruby-lang
<whitequark>
andkerosine: in the class description
<shevy>
andkerosine I never allowed rdoc to decide how I comment and describe ruby code
<shevy>
unfortunately as my code changes, documentation also has to change, and it's not a lot of fun to maintain a changing documentation :(
* manveru
doesn't write docs
<manveru>
ppl gotta learn how to read code
<shevy>
hehe
<shevy>
well
<shevy>
I always try to have a good tutorial
<andkerosine>
^
<andkerosine>
I think I might be commenting this current project a bit excessively, though.
<shevy>
and I mean also with some explanations that go over it, like design choices here or there, or conventions for a project (and why)
<Asher>
manveru - and this is why you refuse to learn C?:P
<manveru>
Asher: i refuse to _write_ C
<Asher>
oic
<manveru>
i can read it well enough for my uses
<Asher>
fair enough :)
<Asher>
i was just giving you shit
<manveru>
just don't wanna inflict even more C on this world
<whitequark>
^ that
<whitequark>
there is a reason libcurses was called like that
<whitequark>
C should have been given a similar name
<shevy>
man
<shevy>
C code can be beautiful
<shevy>
but *curses has the worst API ever
<manveru>
> The name is a pun on the term “cursor optimization”. <
<manveru>
well, or that :)
<whitequark>
shevy: "beautiful" as in "burning people to make funny ash patterns"? yes.
<whitequark>
you also think that autotools are sane and reasonable, don't you?
<shevy>
Asher, that you don't get any real control over compilation as end-user
<Asher>
what sort of control do you want
<shevy>
--disable-this --enable-that
<Asher>
err.. doesn't cmake make that super easy
<shevy>
how? When you are in a project with a CMake* file, how do you do this specifically?
<whitequark>
shevy: err, that's C++
<shevy>
whitequark, is it???? :(
<Asher>
well how do you do it at all unless you know the internals of the source?
<Asher>
presumably it's up to the package maintainer to make it possible to disable portions
<Asher>
cmake is intended for facilitating that
<shevy>
damn
<shevy>
.cc files
<whitequark>
shevy: if C should have been called "A" from "Abomination", C++ is "EA" from "Evil Abomination".
<cout>
cmake feels obscure and awkward to me
<whitequark>
cout: compared with autotools?
<shevy>
Asher well but how do you do it? I mean with configure based systems, I can do "./configure --help" and get lots of options, how can I do the same in cmake, and how can I then use any of these options?
QaDeS has joined #ruby-lang
<cout>
whitequark: yes, even compared to autotools
<Asher>
whitequark - like EA sports?
<shevy>
the only thing I found out so far was to use prefix in cmake :P
<cout>
at least with autotools you get some decent documentation
benanne has joined #ruby-lang
* whitequark
slowly moves aside from cout
<shevy>
whitequark man ... this is really C++ :(
<shevy>
then I no longer know of any cool C project
<shevy>
stupid C
<cout>
MRI is written in C :)
<whitequark>
that doesn't make it any better
<whitequark>
more like the opposite
<cout>
whitequark: you don't like MRI?
<whitequark>
no
<cout>
whitequark: then you don't like Ruby :P
<Asher>
shevy - i don't know offhand but everything i've done w/cmake has been easy and that would shock me not to be there
<whitequark>
bzzt wrong
<whitequark>
MRI is slow and obscure at least
<whitequark>
you can't argue with "slow"
<cout>
whitequark: I can argue with a stop sign.
seanstickle has joined #ruby-lang
<shevy>
well, slow
<whitequark>
and for "obscure", let me ask you for a feature and we'll see how long would it take you to find the necessary lines in MRI sources
<cout>
whitequark: that would be a faulty argument given how familiar I already am with the sources :(
<shevy>
not that many alternatives to MRI
<whitequark>
cout: then you can't say anything about their quality
<whitequark>
I am unfortunately quite familiar
<whitequark>
shevy: jruby
<cout>
whitequark: I can say anything I want.
<shevy>
:)
<whitequark>
cout: I can (and will) silently disregard anything I don't like and continue to argue with you
<whitequark>
:D
<shevy>
I hope there is minimalist java
* Asher
senses a sequel to RubyPanther
<shevy>
then I could go into jruby
<shevy>
but I can't overcome my barrier AGAINST learning java ...
<shevy>
can I write ruby-looking java code?
<whitequark>
shevy: yes. google for `mirah'
<shevy>
hmm ok, this is the second time I heard that
<whitequark>
it's basically ruby parser which generates java bytecode from something ruby-like.
<whitequark>
Asher: wtf is RubyPanther?
<seanstickle>
One of the plentiful rubytrolls
<cout>
asher: which one of us were you associating with trolls? :)
<shevy>
:)
<shevy>
only RubyPanther I suppose ;)
<Asher>
whitequark's assertions about the implications of MRI being C
<shevy>
oh!
* shevy
fetches popcorn
<whitequark>
Asher: err, for the record, I wasn't serious about that
<whitequark>
that _still_ does not make MRI any better. but it's unrelated to the language it's written in.
<shevy>
gaaaaah man ... it had such a good start ... :(
<Asher>
haha ok noted for the record
<whitequark>
you can write bad cobol
<whitequark>
bad fortran
<whitequark>
bad java
<whitequark>
bad c
<whitequark>
you presumably cannot write bad brainfuck because well. can you define what "good brainfuck" is?
<andkerosine>
Efficient brainfuck?
<whitequark>
any brainfuck code is an efficient brainfuck
<Asher>
how about malbolge??
<rue>
Mehbolge amirite
* whitequark
surrenders
<Asher>
i love that for years the only hello world code in malbolge printed out HellO World
<shevy>
lol
<shevy>
I dont even want to ask why ...
<seanstickle>
Some of these languages have limited source code available for reading
<cout>
Hell 'O World
<seanstickle>
Try finding a substantive program written in APL to study, for instance. Devilishly hard.
<andkerosine>
Some of the APL solutions to Project Euler problems are pretty enlightening, though.
<seanstickle>
True.
<cout>
seanstickle: KDB's query language is based on APL.
<seanstickle>
cout: well, based on J, which was based on APL.
<andkerosine>
Which was based on trolling the programming world.
Stalkr_ has joined #ruby-lang
<seanstickle>
Oh, don't be mean. :P
<whitequark>
cout: I try to resist the urge to /nick cin
<whitequark>
but hardly can
<cout>
how cinful of you.
<rue>
Hm, amirite isn't taken as a programming language name. I'm going to make a prolog.
<cout>
seanstickle: I didn't realize K was based on J. The wikipedia article forces you to read between the lines to figure that out.
<nedbat>
what is the distinction between #ruby and #ruby-lang? Why are there two channels?
<seanstickle>
cout: well, written by Arthur Whitney, who wrote A+ and J.
<seanstickle>
cout: so presumably, he took his insights with those when he wrote K
<seanstickle>
nedbat: ancient enmity
<cout>
seanstickle: wikipedia says J was Ken Iverson
<seanstickle>
cout: Iverson designed the language.
<nedbat>
seanstickle: over what?
<cout>
but I wouldn't be surprised if Whitney worked on J too
<seanstickle>
cout: Whitney was the wacko genius who wrote a lot of the source code
<rue>
cout: Houston?
<cout>
rue: who?
<rue>
Hm, we have a problem.
<seanstickle>
nedbat: opinions on how to run a channel
<cout>
lol
<seanstickle>
nedbat: and also because people just kept typing in /join #ruby
<cout>
I didn't know there was a #ruby
<cout>
(except on efnet)
<rue>
Stupid #ruby.
<seanstickle>
Well, now you know
<cout>
who hangs out on #ruby?
<seanstickle>
There is also #philosophy, but I'd recommend staying away
<nedbat>
cout: more ppl than here....
<andkerosine>
Quality > quantity.
ramonmaruko has joined #ruby-lang
<seanstickle>
Quantity has a quality of its own
<andkerosine>
And it is a lesser one.
<cout>
great scott
<seanstickle>
Except when it's larger
<andkerosine>
I... hm... touche.
<rue>
nedbat: Yeah, unfortunately the douchebags who run #ruby broke the agreement that they'd have a /topic forward to here so newbs tend to end up there. (#ruby-lang being designated in ruby-lang.org)
<andkerosine>
#ruby: the #ruby-lang for people who can't read.
rohit_1 has joined #ruby-lang
<nedbat>
can you explain the disagreement that led to the split? If you'd rather not, no big deal, just interested in these dynamics...
<seanstickle>
Man, when people start hanging their self-respect on the name of an IRC channel, something has gone very wrong.
<seanstickle>
nedbat: this isn't 19th century European history
<andkerosine>
It used to be.
<seanstickle>
nedbat: two channels popped up, people go to both
<seanstickle>
nedbat: end of story
<shevy>
people tend to go to #python people tend to go to #lua people logically tend to go to #ruby first too
<seanstickle>
There is no dramatic reveal of a deep-seated feud
<seanstickle>
We are not always at war with Oceania
<nedbat>
seanstickle: ok, you introduced the word "disagreement" into the discussion, but ok.
<cout>
I see a few nicks in there I recognize... ixx, MetaCosm... mostly C++ programmers
<seanstickle>
nedbat: actually, I said ancient enmity
<cout>
so the channel must be pretty irrelevant
<nedbat>
seanstickle: true, "opinion" was what you said.
<shevy>
seanstickle hahah Oceania :D
<seanstickle>
nedbat: which is so over the top, I figured you'd pick up on it
<rue>
nedbat: The original #ruby crowd was mostly people who got banned, probably not many of them left.
<shevy>
big matz is watching over you
<andkerosine>
#ruby is a human face being stomped on.
<andkerosine>
Forever.
<seanstickle>
andkerosine: no, no, no, that's #php
<shevy>
#php is just sadness 2.0
<rue>
But now, Omloop Het Nieuwsblad
<rue>
nedbat: (Also, per Freenode TOS, #ruby should be renamed ##ruby, but they've not done so yet.)
<nedbat>
rue: that would be a clear indicator to me, maybe not others.
<Asher>
most of the people active in #ruby were either banned from #ruby-lang or also dwell here
<cout>
I wonder if there's an active ruby community on undernet.
<cout>
I miss undernet.
<andkerosine>
Oh. My. God.
<shevy>
what is undernet?
<andkerosine>
'phag'.rot13 == 'cunt'
<cout>
my first irc experience *sniff sniff*
<seanstickle>
undernet, dalnet, and efnet, the trinity of anger
<andkerosine>
Mind === blown.
<shevy>
ah ok
<Asher>
trinity of anger eh? nice
<cout>
I wouldn't go that far, more like...
<cout>
ok well yes I would go that far.
<andkerosine>
But undernet has #bookz.
<cout>
what's that?
<andkerosine>
All the books you could possibly want... gratis.
<andkerosine>
Well, most.
<shevy>
I like paperbooks
<cout>
isn't that what torrents are for?
<andkerosine>
I read on my DS, and I could never go back.
<andkerosine>
Well, sure.
<cout>
I thought file sharing on irc went out when kazaa came along
<andkerosine>
But #bookz tends to have a wider selection.
<rue>
Nintendo DS?
<andkerosine>
Indeed.
<andkerosine>
It's so comfortable.
fatprog has joined #ruby-lang
<andkerosine>
Being able to read with one hand and smoke with another is just... so grand.
<cout>
(I would have said napster but it was sort of ircish)
<deryldoucette>
i have become is more accurate english
chendo has joined #ruby-lang
<seanstickle>
English has no formal spec.
<seanstickle>
So "more accurate" is a misnomer.
<seanstickle>
It's all about usage, beeyatches
<deryldoucette>
lol
<andkerosine>
Bitches love usages.
smartinsrenato has joined #ruby-lang
<andkerosine>
Are conditional default parameters evil?
hagabaka has joined #ruby-lang
hagabaka has joined #ruby-lang
<Asher>
why would they be?
<Asher>
they aren't really conditional so much as they are parameters iwth default values
<andkerosine>
But... conditional default values.
wmoxam has joined #ruby-lang
<Asher>
they're optional parameters b/c they already have values
<rue>
Are you two talking about the same thing?
<Asher>
duno
<deryldoucette>
i *think* he means if this param is set then set this other param to this value, if that one then set to this
<andkerosine>
Negative.
<deryldoucette>
ok
Jay_Levitt has joined #ruby-lang
<andkerosine>
I'm wrapping the reddit API. My comments method is a mixin attached to the Post, User, and Subreddit classes. The count parameter should default to 25 for users and subreddits, but 500 for posts, which imitates the site's default functionality.
<andrewvos>
FFS shevy you are such a hipster. What the hell is bluefish
<shevy>
an editor based on GTK mostly
<havenn>
I've been liking Sublime Text 2.
<shevy>
geany and bluefish are fairly similar, except that geany's default syntax highlighting for ruby sucks
<shevy>
but geany is much more active these days than bluefish :(
<BeTaMaX>
Thank you so much, I sometimes use Geany for java
<shevy>
cool
<shevy>
my dream would be an editor in ruby
<andrewvos>
shevy: redcar
<havenn>
shevy: ^ your dream is real
<shevy>
hmm
<shevy>
I think I tried out redcar years ago
<shevy>
dunno why I didnt use it
<BeTaMaX>
I've been giving my first steps in programming. I started a couple of weeks ago with Java and I find it quite dirty
<shevy>
oh wait
<shevy>
redcar requires java right?
<havenn>
Textmate 2 and Sublime Text 2 are pretty. Feed me colors. But, they aren't the One True Text Editor...
<BeTaMaX>
I mean I think Ruby is cleaner and simplier, less lines of code to write the same programm
<andrewvos>
shevy: I think so
<andrewvos>
shevy: Though, there was some talk of moving away from jruvy
<shevy>
BeTaMaX, well you have to compile java code into .jar files or?
<andrewvos>
jruby
<andrewvos>
*
<shevy>
so you can't compare ruby to java that easily. it's more fair to compare ruby with python php perl lua
<andrewvos>
java makes me sad
<shevy>
yeah andrewvos that was the strange thing
<shevy>
redcar looks really awesome from the screenshots
<theconartist>
BeTaMaX: ruby is a good language to learn in imo
<erikh>
there's also diakonos
<erikh>
which is a text-based ruby editor
<shevy>
but I dunno, java makes me really angry and violent. perhaps I am just cursed, with my brittle linux system, things tend to break, and java ends up not working. I can't even start libreoffice either, and it does not tell me what the specific error is
* shevy
is sad.
<shevy>
the less java for me, the better for my calmness
<havenn>
VIM or emacs?
<andrewvos>
vim
<andrewvos>
Noooo why did I respond
Jake232 has joined #ruby-lang
<burgestrand>
kill the troll
<shevy>
vim and emacs - two eternal enemies
<shevy>
but it's unfair to compare an editor to an operating system
<andrewvos>
Compiling node.js takes an eternity
smartinsrenato has quit [#ruby-lang]
<seanstickle>
ancient enemies, but they fight together against the horde of TextMate and Notepad+ users
<andrewvos>
We need a bot to kick anyone who mentions vim/emacs
* manveru
kicks andrewvos
<manveru>
only a little bit more... and i'll build my vimpire!
<andrewvos>
heh
<andrewvos>
Speaking of bots, I'm installing hubot
dfr|mac has joined #ruby-lang
wmoxam has joined #ruby-lang
wmoxam_ has joined #ruby-lang
Bwild has joined #ruby-lang
<andrewvos>
Wait hang on
<andrewvos>
Just realised I would need to write javascript
<andrewvos>
For hubot
<andrewvos>
Can anyone recommend a ruby bot?
<manveru>
andrewvos: you could continue with kamirc :)
<andrewvos>
manveru: A work in progress, I take it?
<andrewvos>
Basically, anything that shevy says is good is going to be whack
wmoxam has joined #ruby-lang
<csmrfx>
andrewvos: I bet you're right about that
<tbuehlmann>
for irc bot stuff I'm using ponder. most likely because I wrote it
<shevy>
andrewvos you are good
pxjorge has joined #ruby-lang
<shevy>
no, seriously, apeiros writes great ruby code
<shevy>
he wrote that audio thing ...
<dreinull>
I have a dictionary in a mongo db via mongomapper and it has a search interface that looks like this: Regexp.new(/^#{search_term}/i) It fails however when I search for this: ca (* -> complains about the parens
<Asher>
dreinull - sounds like the paren is not escaped for the regexp
<dreinull>
Asher: is there a way to just use the string or is there a simple escape method available?
<csmrfx>
I think it would be better to use something like 'pass' instead of '+' and 'fail' for same as '-'
darkf has joined #ruby-lang
<csmrfx>
or like testi pass { 0 == 0.to_i }
frequency9 has joined #ruby-lang
andrewhl has joined #ruby-lang
apeiros_ has joined #ruby-lang
lenilson_dias has joined #ruby-lang
daglees has joined #ruby-lang
tla has joined #ruby-lang
<deryldoucette>
got a question. when we build rubies under RVM, we're not adding a --static flag or anything like that, but ruby seems to be building both dynamic and static libs, though it only links against the dynamic.
<andrewvos>
injekt: You here?
<deryldoucette>
for some of our embedded userbase this adds an unneeded 12MB
<andrewvos>
injekt: Well, I can see you're here.. but are you active?
<Asher>
deryldoucette - i believe you can —disable-static
<deryldoucette>
is there a way to tell ruby *not* to create the static lib at *all*?
<deryldoucette>
ahh ok. thank you.
<Asher>
it's output into the Config.rb i believe
<Asher>
are you trying now? if that doesn't work lmk and i'll find the option for you
<deryldoucette>
no was going to pass it on too the client and let them :)
<deryldoucette>
(Working on less than 4h sleep so figured let them expend the energy and head off to bed after passing that along)
<Asher>
yeah seems to work here
<deryldoucette>
oh, thanks for verifyig that. appreciated
<Asher>
hmm it doesn't output a configure error at least gimme a sec i'll verify that it actually works
<deryldoucette>
thanks
looopy_ has joined #ruby-lang
gregf has joined #ruby-lang
<Asher>
looks like i was wrong
<Asher>
there is a way to enable/disable building shared but it looks like it builds static no matter what
<Asher>
what is the context where it's needed as a build option rather than simply removing it afterward?
<deryldoucette>
Asher: we have several embedded developers in our userbase, and they would like the static lib not to be built for space concerns
<Asher>
what do you mean by embedded developers
erpuds has joined #ruby-lang
<Asher>
developers working on code to be deployed on embedded platforms?
<deryldoucette>
yeah with the ruby installed on the platform. they just don't want the static since the shared works
<deryldoucette>
makes it easier for them to package everything up (not a gem, inhouse install packages) and not have to shed the 12MB static lib
<deryldoucette>
I personally am not an embed type, so not sure all of the specifics, just they specifically mentioned the space concern
<Asher>
the concern seems legit but presumably they are cross-compiling right?
<deryldoucette>
i believe so, yes (not my area of expertise, but conceptually it would make sense they are doing so)
<deryldoucette>
who would have thought working on full blown system would hamstring your knowledge for embedded environments hehe
<Asher>
it seems to me like the build system has little to do with the deployment packaging
<Asher>
i would think they could simply either 1. remove the static lib 2. not package the static lib
<Asher>
maybe ask if there is a reason they can't do that?
<Asher>
hard to figure out alternative options without knowing why that isn't sufficient
<deryldoucette>
right. they were asking if they could simply remove it, and then asked if they could stop building it completely. (they too couldn't see how not to build it)
<Asher>
yeah they can definitely remove it
<deryldoucette>
i wonder why it automatically builds both.
<Asher>
i believe standard deployment is static lib + shared lib if supported
<deryldoucette>
i know we're not passing nodynamic or --static-* flags or anything.
<deryldoucette>
ok
<Asher>
they seem to have changed the shared lib options
<Asher>
it used to be that you had to —enable-shared
<Asher>
but in 1.9.3 it's built by default, not sure when that changed
<deryldoucette>
which is? the static or the dyn?
<Asher>
the dynamic (shared) lib
<deryldoucette>
(I never checked that myself. interesting)
<Asher>
it's curious that they prefer the dylib
<Asher>
usually embedded systems will stick with static b/c it's smaller
<deryldoucette>
yeah, *that* thought passed through my head
<Asher>
actually it's not… so i take that back
<Asher>
3.5 vs. 5.5
<Asher>
here
<deryldoucette>
gives me something else to play with for a bit when i get some sleep.
<deryldoucette>
but that —disable-static definitely does *not* work?
wyhaines has joined #ruby-lang
<Asher>
correct
<Asher>
i'm surprised it didn't throw a configure error
<Asher>
for unrecognized option
<deryldoucette>
weird.
<Asher>
but i double checked the configure script and the rbconfig.rb params and there is nothing for it
<deryldoucette>
thanks for taking the time on this. appreciate it
<Asher>
np
<deryldoucette>
ok, will pass all this along. heading for the rack before i keel over.
<Asher>
later!
<deryldoucette>
o/ &
davidw has joined #ruby-lang
davidw has joined #ruby-lang
andrewhl has joined #ruby-lang
Bwild has joined #ruby-lang
Defusal has joined #ruby-lang
akahn has joined #ruby-lang
Caius has joined #ruby-lang
srbaker has joined #ruby-lang
turboladen has joined #ruby-lang
spuk has joined #ruby-lang
mika has joined #ruby-lang
pvh has joined #ruby-lang
hackeron has joined #ruby-lang
znouza has joined #ruby-lang
dominikh has joined #ruby-lang
werwolf has joined #ruby-lang
samuelkadolph has joined #ruby-lang
jarib has joined #ruby-lang
levifig has joined #ruby-lang
tomb has joined #ruby-lang
_br_ has joined #ruby-lang
cynosure has joined #ruby-lang
mitchty has joined #ruby-lang
apeiros has joined #ruby-lang
zmack has joined #ruby-lang
VladGh has joined #ruby-lang
jacobwg has joined #ruby-lang
wallerdev has joined #ruby-lang
m3nd3s has joined #ruby-lang
Bwild has joined #ruby-lang
nedbat has joined #ruby-lang
nedbat_ has joined #ruby-lang
S2kx has joined #ruby-lang
Spooner_ has joined #ruby-lang
looopy has joined #ruby-lang
Kyril has joined #ruby-lang
uniqanomaly has joined #ruby-lang
jeffmoss has joined #ruby-lang
havenn has joined #ruby-lang
m3nd3s has joined #ruby-lang
havenn_ has joined #ruby-lang
pemeon has joined #ruby-lang
ryanf has joined #ruby-lang
heftig has joined #ruby-lang
s0ber has joined #ruby-lang
y3llow has joined #ruby-lang
y3llow has joined #ruby-lang
y3llow has joined #ruby-lang
heftig has joined #ruby-lang
y3llow has joined #ruby-lang
<andrewvos>
I'm installing bundler under system ruby and keep getting this when I run bundle install ERROR: Gem bundler is not installed, run `gem install bundler` first.
<andrewvos>
I'm installing bundler under system ruby and keep getting this when I run bundle install: "ERROR: Gem bundler is not installed, run `gem install bundler` first."
<andrewvos>
What's that about?
smilingpeter has joined #ruby-lang
smilingpeter has quit [#ruby-lang]
leopard_me has joined #ruby-lang
leopard_me has joined #ruby-lang
<shevy>
good old bundler... never worked for me so far
<canton7>
shevy, do you know of anything which does work? I've also had lots of bundler headaches
<shevy>
canton7 with bundler? nope. But gems in general work very nicely, I like gems nowadays
<deryldoucette>
mpapis-bundler
<shevy>
I didnt test bundler thoroughly though
pemeon has quit [#ruby-lang]
<andrewvos>
To be fair, it's not really a bundler issue.
<andrewvos>
And bundler is normally my bro
voker57 has joined #ruby-lang
voker57 has joined #ruby-lang
<canton7>
the thing that bites me is if you're running cross-platform, with gems which have platform-specific versions
<apeiros>
canton7: stuff them into the package
<canton7>
don't really want to do that with eventmachine..
<apeiros>
worked fine for me for - I think it was libv8
<canton7>
compiled code, and all that
<erikh>
andrewvos: what's your $PATH
<erikh>
also do you have any RUBYLIB, etc set
<andrewvos>
erikh: Don't wory. I'm going to give up now
<andrewvos>
erikh: I just found out the gem I'm using doesn't run on 1.8.7
<andrewvos>
erikh: And I don't have the time to wait for ruby to compile on an ec2 micro instance
<erikh>
mmmm t1.micro
<andrewvos>
hehe
<erikh>
bringing new meaning, "linux is free if your time is worth nothing"
<andrewvos>
eyeah man
<erikh>
fwiw, virtualbox will be faster
<andrewvos>
Took an hour to compile ruby, then I realised I didn't install the dependencies
<erikh>
on almost any machine made in the last 5 years
<erikh>
... maybe not a netbook
<imperator>
real men use solaris
<erikh>
haha
<imperator>
because real men like pain
<erikh>
I prefer FreeBSD, but until they get their Xen stuff sorted out, it's useless to me.
mytrile has joined #ruby-lang
<andrewvos>
Ubuntu server ftw
<deryldoucette>
erikh: i'm with you on the fbsd
<erikh>
Ubuntu server's not bad but there are a lot of things I don't like about it.
<erikh>
FreeBSD offers a lot more flexibility and its defaults are definitely more aimed at sysadmins.
<erikh>
and the raw architecture is so much better