chessguy has quit [Remote host closed the connection]
spuk has quit [Ping timeout: 252 seconds]
spuk has joined #ruby-lang
<lianj>
jaimef: @host_translation_table[client.to_s] or client
<jaimef>
thanks. always know, if ugly, wrong.
keroppi has quit [Quit: keroppi]
<lianj>
wasn't really wrong. if c = @host_translation_table[client.to_s]; c; else; client; end works too.. yours was just doing the hash table lookup twice
happydude has quit [Read error: Connection reset by peer]
jxie has quit [Quit: leaving]
krohrbaugh has joined #ruby-lang
wallerdev has joined #ruby-lang
tylersmith has quit [Quit: tylersmith]
krohrbaugh has quit [Ping timeout: 264 seconds]
glebm has quit [Ping timeout: 252 seconds]
cored has quit [Ping timeout: 245 seconds]
cored has joined #ruby-lang
cored has joined #ruby-lang
cored has quit [Changing host]
glebm has joined #ruby-lang
divout has joined #ruby-lang
<Hanmac>
hm nokogiri for sample does include the lib source that it binds in its own code, is that the new style and should other gems do that too? or depends it on the size of the binded lib?
<whitequark>
Hanmac: on the plus side this doesn't require you to install the dev files in your system
aemeredith has quit [Remote host closed the connection]
<whitequark>
on the minus side this doesn't give you any security updates
<whitequark>
I'd say it is net negative
<whitequark>
except for maybe Windows, but you can distribute precompiled gems for mswin32
<Hanmac>
on the minus side: you install a gem and then you need to compile the entire wxWidgets lib ;P "building native extension, that could take a *while*" ;P
<whitequark>
Hanmac: or qt
<whitequark>
"linking webkit... please get a vacation for a week"
lguardiola has joined #ruby-lang
<whitequark>
"and don't you dare run this on a mere 32-bit machine or a 64-bit one with less than 8G RAM"
aemeredith has joined #ruby-lang
lguardiola has quit [Ping timeout: 245 seconds]
nickchow has joined #ruby-lang
kogent has quit [Quit: kogent]
bzalasky has quit [Remote host closed the connection]
<whitequark>
yorickpeterse: re symbols for identifiers. Basically yes. I thought about making them strings but there's not much sense in that.
jxie has quit [Client Quit]
jxie has joined #ruby-lang
<whitequark>
as per block parameters... it's not weird in the way you state it. do..end is not a block argument. it's a literal block passed to the method.
<whitequark>
there are several reasons I don't want it to go in the place of block-pass
thone has quit [Ping timeout: 264 seconds]
<whitequark>
1) we lose the ability to represent a not-well-formed AST for this: "foo(&bar) do end"
<whitequark>
2) it requires some extra hoops to convert one structure to the another in the grammar/builder
<ddfreyne>
The only thing I don't like about minitest is assert_equal and the order of arguments. I usually do assert_equal expected, actual... but there has to be a better way
<judofyr>
whitequark: then you do assert_nil(…)
<whitequark>
judofyr: ugly :/
ChuckJHardy has quit [Ping timeout: 268 seconds]
<judofyr>
whitequark: I guess I can add #expect too which always wraps
<whitequark>
judofyr: please do
<judofyr>
ddfreyne: the order is easier with chaining :)
<ddfreyne>
judofyr: Somewhat, yes
<ddfreyne>
judofyr: It's very smalltalky :)
<ddfreyne>
self assert: 2+2 equals: 4
<ddfreyne>
judofyr: Why #equal and not #equals?
<ddfreyne>
(s)
joshH3 has joined #ruby-lang
<judofyr>
ddfreyne: I had that initially, but it was difficult keeping everything consistent
joshH2 has quit [Ping timeout: 252 seconds]
<ddfreyne>
judofyr: Because of difficulties with #not_equal etc?
<ddfreyne>
Hmm yeah, the other stuff doesn't work well either
thone has joined #ruby-lang
dustint has quit [Ping timeout: 250 seconds]
thone has quit [Ping timeout: 264 seconds]
ChuckJHardy has joined #ruby-lang
FlyingLeap has quit [Quit: No Ping reply in 180 seconds.]
rwk1 has quit [Remote host closed the connection]
<judofyr>
whitequark: there. added #expect.
gnufied has joined #ruby-lang
<whitequark>
judofyr: cool. I might use that in my future projects, thanks
marr has joined #ruby-lang
jerrytgarcia has joined #ruby-lang
<judofyr>
whitequark: should be easy to add custom assertions too: def assert(*); super.tap { |c| c.extend(MyStuff) } end
lacrosse has joined #ruby-lang
<whitequark>
judofyr: isn't that going to be, like, slow?
<judofyr>
whitequark: probably. not sure if it would be noticeable
alessio_rocco has joined #ruby-lang
<whitequark>
judofyr: do a benchmark?
<whitequark>
here in parser, I have: 595 tests, 11993 assertions
<whitequark>
I can very well see how that can be slow
<ddfreyne>
That's a lot of test and assertions
jerrytgarcia has quit [Client Quit]
<whitequark>
ddfreyne: tests are all handwirtten. assertions are expanded from a DSL but are also handwritten.
<whitequark>
so, yes.
<whitequark>
oh also, most of the testsuite [assertions] are ran against 4 parser versions
<judofyr>
whitequark: is this expected on Parser? 595 tests, 10905 assertions, 31 failures, 2 errors, 0 skips
<whitequark>
judofyr: yup I'm fixing source maps
aemeredith has joined #ruby-lang
<whitequark>
judofyr: git pull, I've fixed some of these.
<Hanmac>
what do you guys think? should it be possible in a method to get the parameters the super method wants?
<judofyr>
whitequark: isn't it better do use the singleton's class ancestors?
<whitequark>
judofyr: hmm maybe
<judofyr>
whitequark: it went from user 0m11.815s -> user 0m12.188s
<judofyr>
whitequark: with .extend on every assert
<Hanmac>
darix because of the modules i include into my Classes, they overlay initialize , the class needs 1 parameter, but Object#initialize wants exactly 0
aemeredith has quit [Ping timeout: 245 seconds]
<whitequark>
super()
<Hanmac>
it might still not be 100% perfect
mbj has quit [Ping timeout: 264 seconds]
<whitequark>
what?
<whitequark>
snooping parameters is worse anyway
<ddfreyne>
Hanmac: Sounds like you're doing something wrong, because it shouldn't ever be necessary
<whitequark>
in racc, the AST for itself is constructed by hand
rwk1 has joined #ruby-lang
<whitequark>
not sure about bison but I think largely same
<whitequark>
treetop uses a bootstrapped parser.
mbj has joined #ruby-lang
thone has joined #ruby-lang
<ddfreyne>
whitequark: Grammars for grammars are usually fairly simple
<whitequark>
ddfreyne: yeah
GarethAdams has joined #ruby-lang
<whitequark>
I'd personally write an LL RDP
<ddfreyne>
Same :)
<whitequark>
(your grammar is most certainly not LL(k) for k > 1 :
<whitequark>
*:)
<ddfreyne>
I wrote a LL(1) RDP for an XML-like syntax a few weeks back
<ddfreyne>
Customer delivers something that looks like XML but actually isn't. AAARRRGGGHHHH
<whitequark>
LOL
<ddfreyne>
There was an existing parser (which did the "parsing" by splitting a string on < etc... yikeswtf)
<ddfreyne>
I rewrote it to use an XML parser... then it turns out it's not actually XML
<ddfreyne>
Sometimes I would like to be able to see my own face when I find out retarded shit like this
ohsix has quit [Ping timeout: 240 seconds]
<whitequark>
ddfreyne: you do not want that. trust me
<whitequark>
you'll be scared by your own murderous rage
<ddfreyne>
haha
ohsix has joined #ruby-lang
mbj has quit [Read error: Connection reset by peer]
<ddfreyne>
whitequark: Plenty of retarded stuff out there. Today I discovered that unloading a package in VisualWorks Smalltalk takes significantly longer than loading it...
<ddfreyne>
(Which explains why it's discouraged to do so)
<ddfreyne>
That's like "rm somefile" being 5x as long as "cp blah somefile"
ChuckJHardy has quit [Ping timeout: 252 seconds]
workmad3 has joined #ruby-lang
<whitequark>
how does it unload packages?
<whitequark>
something about tracing references? :)
thone has quit [Read error: Connection reset by peer]
<ddfreyne>
whitequark: Maybe... not sure. I have the feeling there's a combinatorial explosion when unloading, too
<whitequark>
...
<ddfreyne>
whitequark: When loading packages, it doesn't track dependencies either, so you have to put the packages in the right order
<ddfreyne>
whitequark: So package A contains class Foo and package B contains a method on Foo, then you need to put the packages in the order A,B because B,A will cause it to crash
<ddfreyne>
whitequark: If you have a few hundred packages then that relaly sucks
<whitequark>
ruby does neither, however...
Taranis has quit [Read error: Connection reset by peer]
Taranis has joined #ruby-lang
Axsuul has quit [Ping timeout: 252 seconds]
ChuckJHardy has joined #ruby-lang
<ddfreyne>
whitequark: It has a dialog in which you can modify the order. But you can't drag packages... it has a UP and a DOWN button and it moves the package up/down one place. If you have to move a package all the way to the top... then you're clicking like a spaz :P
<whitequark>
:D
<whitequark>
proprietary sw at its best
<ddfreyne>
whitequark: Often, it doesn't even tell you what the problem is. It just crashes with a really weird stack trace...
<ddfreyne>
Actually, most crashes happen because of a misconfiguration in some way. Very bad error reporting...
<ddfreyne>
whitequark: Packages are basically serialised object memory (compiled objects with code) and memory isn't always reliable and sometimes gets corrupted... which means your commits can get corrupted in the VCS too
<ddfreyne>
It's really icky :D
Averna has quit [Quit: Leaving.]
thone has joined #ruby-lang
adambeynon has quit [Quit: Computer has gone to sleep.]
<antbody>
Worst software I've ever used was used for FPGA design ;<
<antbody>
If you selected something, then you needed to click in empty space before you could delete it.
<antbody>
Every single thing you tried to do had something similar.
tsion has joined #ruby-lang
tsion has quit [Changing host]
tsion has joined #ruby-lang
<ddfreyne>
BUT... my professional Smalltalk career ends in 9 days and 3 hours :D
<antbody>
When you compiled the design, it took like 6 fucking hours
<antbody>
And after the six hours it told you, nope, can't compile.
<antbody>
Oh and you couldn't select anything in groups, so if you had a design with like 100 elements and needed to remove then all, you'd delete them one by one.
<antbody>
God the memories...
kgrz has quit [Ping timeout: 252 seconds]
<ddfreyne>
antbody: haha
<whitequark>
ando: which software?
adambeynon has joined #ruby-lang
Henry_ has joined #ruby-lang
<antbody>
I'll try to look it up, it was made by some Finnish university
<whitequark>
academic code
<whitequark>
you can finish at that :D
Henry_ has left #ruby-lang [#ruby-lang]
<antbody>
yeah...
<whitequark>
I wonder why LLVM is so damn good
<judofyr>
I like BankID's Java Applet in Firefox on OS X: the "Ok" button is disabled until you press the Cmd-button
<antbody>
Still studying. I still can't believe that people can become doctors and can't code for shit.
<whitequark>
antbody: their job does not include coding
<whitequark>
and/or their mind is simply trained at another (tbh more important) class of problems
<whitequark>
but universities should just hire/enroll good (open-minded!) professional programmers for these tasks.
<antbody>
Still, I don't find that having being a researches should exclude being a good coder.
<antbody>
Though maybe they just don't have the enthusiasm towards coding.
<whitequark>
it's not mutually exclusive. but researchers see coding as a means to achieve their final goal, not even a task of its own
<whitequark>
most simply don't care
ikrima has quit [Quit: Computer has gone to sleep.]
<tommylommykins>
at least I had the sense not to actually inflict that on my workmates
krohrbaugh has quit [Ping timeout: 256 seconds]
ndrst has quit [Quit: WeeChat 0.4.0]
dhruvasagar has joined #ruby-lang
<yorickpeterse>
...wow
ndrst has joined #ruby-lang
<whitequark>
is that autogenerated?
thone has joined #ruby-lang
<tommylommykins>
whitequark: I wrote it myself
<tommylommykins>
it's not that hard, just repeated condensing of code
<yorickpeterse>
you're an evil person
<tommylommykins>
I realised the more-sensible-looking-version was made up of all sensibly-lengthed one-line functions, no I just tapped my way to victory
r0bgleeson has quit [Ping timeout: 252 seconds]
thone has quit [Read error: Operation timed out]
thebastl has joined #ruby-lang
dhruvasagar has quit [Ping timeout: 268 seconds]
Mon_Ouie has joined #ruby-lang
dhruvasagar has joined #ruby-lang
mikestok has joined #ruby-lang
<antbody>
judofyr: haha :D
srbaker has joined #ruby-lang
sush24 has quit [Quit: This computer has gone to sleep]
srbaker has quit [Client Quit]
Domon has quit [Remote host closed the connection]
ChuckJHardy has quit [Ping timeout: 264 seconds]
mikestok has quit [Quit: Computer has gone to sleep.]
dr_bob has quit [Ping timeout: 245 seconds]
thone has joined #ruby-lang
ruskie has quit [Excess Flood]
mytrile_ has joined #ruby-lang
mytrile has quit [Ping timeout: 256 seconds]
dhruvasagar has quit [Ping timeout: 252 seconds]
llakey has quit [Ping timeout: 252 seconds]
llakey has joined #ruby-lang
thone has quit [Ping timeout: 256 seconds]
jaymes_ has quit [Ping timeout: 252 seconds]
tallship has quit [Ping timeout: 252 seconds]
dhruvasagar has joined #ruby-lang
jaymes has joined #ruby-lang
tallship has joined #ruby-lang
thone has joined #ruby-lang
benlovell has quit [Quit: Computer has gone to sleep.]
thone has quit [Read error: Operation timed out]
JohnBat26 has joined #ruby-lang
JohnBat26 has joined #ruby-lang
thebastl has quit [Remote host closed the connection]
ruskie has joined #ruby-lang
thebastl has joined #ruby-lang
charliesome has joined #ruby-lang
madb055 has joined #ruby-lang
nertzy2 has left #ruby-lang [#ruby-lang]
nertzy has joined #ruby-lang
ChuckJHardy has joined #ruby-lang
kstuart has quit [Read error: Operation timed out]
thebastl has quit [Ping timeout: 245 seconds]
krohrbaugh has joined #ruby-lang
baba has quit [Ping timeout: 268 seconds]
kstuart has joined #ruby-lang
ChuckJHardy has quit [Ping timeout: 255 seconds]
krohrbaugh has quit [Ping timeout: 252 seconds]
kstuart has quit [Ping timeout: 264 seconds]
kstuart has joined #ruby-lang
<whitequark>
antbody: btw, that thing you linked me to is totally awesome
<charliesome>
whitequark: you can do anything with eval and binding so i'd be surprised if special variables are the worst you can do
<khaase>
so, is there any code quality tool that works with Ruby 2.0?
<whitequark>
khaase: no.
<khaase>
hmm
<whitequark>
tldr: ruby_parser is shit and patches not welcome.
<ddfreyne>
whitequark: does parser parse Ruby 2.0?
<khaase>
whitequark: but you did a rewrite, no?
<whitequark>
ddfreyne: yes
<ddfreyne>
I need to play around with that :)
<whitequark>
khaase: yes, but it's not (yet) compatible with ruby_parser
<whitequark>
by AST format
<whitequark>
patches are welcome :)
scottschecter has joined #ruby-lang
<ddfreyne>
whitequark: So what's up with ruby_paser?
<whitequark>
ddfreyne: it doesn't support 2.0
<whitequark>
and issues and pull requests are just sitting there for months without any attention, etc
mistym has joined #ruby-lang
mistym has quit [Changing host]
mistym has joined #ruby-lang
<whitequark>
it is not likely that it will gain 2.0 support any time soon.
rwk1 has joined #ruby-lang
<yorickpeterse>
and didn't zen said no to keyword arg support?
<charliesome>
wait seriously?
<whitequark>
yorickpeterse: he already reopened
<charliesome>
you can't call yourself ruby_parser and not, you know, parse ruby
<yorickpeterse>
whitequark: ah
<whitequark>
charliesome: it doesn't currently do that anyway, see /issues ;)
<charliesome>
whitequark: i love how nearly all the issues are yours
outoftime has joined #ruby-lang
<ddfreyne>
whitequark: Does parse have modes for 1.8.x, 1.9.x and 2.0.x?
<yorickpeterse>
Completed 200 OK in 49561ms (Views: 3322.4ms | ActiveRecord: 91.6ms | Mongo: 42115.3ms)
<yorickpeterse>
WEBSCALE
<yorickpeterse>
Fucking internet connection is shit today
<injekt>
Flexa: lolmongo
<whitequark>
ddfreyne: 1.8.7, 1.9.3 (distinct from earlier 1.9.x), 2.0.0 and trunk
<whitequark>
parser.
<khaase>
what, did that nearly take a minute?
<whitequark>
it's also in readme
<injekt>
er
<injekt>
yorickpeterse: ^
<ddfreyne>
yeah, parser*
<yorickpeterse>
injekt: actually it's our internet
<ddfreyne>
whitequark: ok cool. I was too lazy to checkout the README :>
<injekt>
yorickpeterse: :(
<yorickpeterse>
Our Mongo dev DB is hosted by Amazon and our interbutts is shit
<yorickpeterse>
especially since there's about 10 of us atm streaming music
<injekt>
interbutts!
<whitequark>
a clbuttic mistake
<charliesome>
yorickpeterse: why on amazon?
<charliesome>
why not locally
<yorickpeterse>
charliesome: said DB is 120GB
<charliesome>
ouch
<yorickpeterse>
most of us have Macbook Airs
<charliesome>
can't you emsmallenate it
<yorickpeterse>
I'm the only one who has a copy of it
<yorickpeterse>
(locally)
<yorickpeterse>
but I don't actually use i
<TTilus>
uhh, i need to move .rvm directory and looks like it is far from being a simple task
<yorickpeterse>
* it
rwk1 has quit [Ping timeout: 248 seconds]
<TTilus>
anybody done it?
<TTilus>
how?
<charliesome>
TTilus: maybe ask #rvm
<yorickpeterse>
charliesome: also, emsmall what?
<injekt>
TTilus: why do you need to move it?
<charliesome>
yorickpeterse: can't you make a smaller version of it?
<charliesome>
that doesn't have all the data
<injekt>
TTilus: also moving it will screw up the rbconfig paths
kstuart has quit [Quit: kstuart]
<yorickpeterse>
charliesome: probably could but nobody can be bothered
<yorickpeterse>
we're moving offices anyway
kstuart has joined #ruby-lang
<TTilus>
injekt: reorganizing things
wesside has joined #ruby-lang
oneirophren has quit [Ping timeout: 256 seconds]
<TTilus>
injekt: need to find another way...
<whitequark>
TTilus: not a good reason to have such a PITA
<injekt>
TTilus: best bet: rm -rf .rvm
<injekt>
because rvm is shit
<charliesome>
use chruby
<TTilus>
whitequark: well, i foolishly assumed it would just be mv old new
<khaase>
TTilus: for moving stuff with rvm, you have to install it with --moveable
<injekt>
TTilus: you can't move stuff around like that because of paths
<TTilus>
charliesome: i am, this is somebody elses old deployment
<injekt>
lol @ --moveable
<khaase>
hehe, yeah
<khaase>
we use that for binary builds
<khaase>
I mean, I use rbenv locally
<injekt>
ah on travis?
<khaase>
yeah
<ddd>
rvm is shit only if you're thinking that because of all the features. for the core of its job, its fucking excellent at it
<khaase>
well, we don't have that in production yet
rippa has joined #ruby-lang
<khaase>
but soon
<injekt>
ddd: what job is that?
<injekt>
rvm has like 100 jobs no?
<whitequark>
ddfreyne: esipecially when it doesn't work without `rvm implode`
<charliesome>
ddd: i'd go the other way round
<khaase>
soon every passing ruby commit will update our preinstalled ruby-head
<charliesome>
rvm is great when you need all its features
<ddd>
however, having been on the dev team and leaving because of the feature bloat, i do agree that its gotten bloated
<charliesome>
but crap if you only need to switch rubies
<ddd>
but its core job of handling and installing multiple rubies its great at
<injekt>
khaase: nice!
<whitequark>
charliesome: does anyone need that
<ddd>
charliesome: bullshit
<whitequark>
ALL of rvm features
<charliesome>
well maybe not all
<charliesome>
but you get the point i'm trying to make
<ddd>
whitequark is the only one making a valid complaint. too many features
<khaase>
rvm is esp. painfull if just are working on one ruby implementation, and just want to go into rvm and change something for that
<ericwood>
how does the fact that rvm has lots of features interfere with its usage?
<khaase>
I tried a few times to dive into that code
<ericwood>
idk, maybe I'm the exception
<khaase>
and got the horrors
<khaase>
rbenv is really nice in that regard
<whitequark>
shims :(
<ddfreyne>
gemsets were added to rvm as part of a feature request for me... but I never used them :(
<charliesome>
khaase: try chruby next time you're setting up a machine
<injekt>
khaase: yeah, I got fed up with shims then stopped using any ruby version manager, then noticed I had something similar to chruby so just started using that instead
<ddd>
ericwood: because its slowed down considerably because of all the convoluted function calls, most of which carry some little shim for some additional feature / capability, and its suffering from feature creep *badly*
<khaase>
maybe
<khaase>
didn't have issues with rbenv so far
<khaase>
and the code base is really simple to dive into
<ericwood>
I guess that's valid
<ddd>
but 2.0 is a new rewrite, so it should improve and since it will be modular you can jus tload a core set, and drop everything else
dr_bob has joined #ruby-lang
wyhaines has joined #ruby-lang
<ericwood>
I do like how rvm will download and install ruby versions, idk if rbenv does that
dr_bob has quit [Read error: Connection reset by peer]
<whitequark>
ruby-build
thebastl has joined #ruby-lang
<whitequark>
but you can also do wget; tar xvf; ../configure; make; make install
<ddd>
yeah, i personally do like that you can download and install, and i like that you can choose between either a binary or source install
<charliesome>
even still compiling manually is not hard
<whitequark>
for something you do at most once a few months it's not hard
<whitequark>
srsly
<ericwood>
lol
* ericwood
shrugs, types "rvm install" and grins
<injekt>
I compile manually then use chruby
<ddd>
i don't think its a matter of 'hard', more along the lines of a single command to get it done and over with and have it centrally controlled
<ddd>
which is great in my mind
<ericwood>
yeah
<whitequark>
ddd: ericwood: I just use system ruby.
<wmoxam>
ericwood: I just type 'rbenv install'
<wmoxam>
:p
<ddd>
whitequark: which would explain why you're not a candidate for its use then
<ericwood>
stop liking things I don't like!
<whitequark>
ddd: well when I needed 2.0 I switched to chruby
<whitequark>
and compiled it myself
<whitequark>
took me what, one minute?
<whitequark>
hah.
<ericwood>
you must have a super crazy machine, then, to build it in 1 min
<khaase>
hmmm... I switch between common and uncommon ruby implementations a ton
<khaase>
and I don't want to be bothered with how to install them
<khaase>
some of them
<ericwood>
I switch between JRuby and MRI a lot
<ddd>
my refusal to use rbenv is totally a personal dislike of sstephenson, especially after his bullshit FUD fest upon release (half of his complaints were invalid AND he failed to actually do any checking), not due to the final product itself.
<ericwood>
ddd: yes
<ddd>
but if i were to change off of rvm ever, it would be to chruby most likely
<ericwood>
I think that has a lot to do with me not feeling like jumping ship
<khaase>
ddd: main reason I didn't switch at the time
<ddd>
i like chruby's approach
<charliesome>
dunno why rbenv needs to be so complex with its shims and all that crap
<ericwood>
I felt bad for Wayne during all of those shenanigans
<ericwood>
:(
<ddd>
mpapis and I spent weeks fighting off that crap
<ddfreyne>
Yeah, Wayne's a nice guy
<ddd>
nevermind the trolls that sprouted
<ericwood>
I'd be really sad if I were him and my project got trashed by jerks :\
<ddd>
he was hopping mad. he was going to quit working on rvm all together if mpapis remear and I hadn't of talked him out of it
<ericwood>
:(
<ericwood>
shit like that had a lot to do with _why leaving...
<ddd>
that was utter bullshit sstephenson pulled. most of it lies and half truths with a vicious attitude in delivering 'differences'
<ddd>
all it was was a project launch hype stunt. one with no class
<ericwood>
yep :\
<injekt>
hpricot didn't really get trashed, nokogiri was a better alternative, it was basically that simple
<zzak>
good morning!
<ddd>
morning zzak
<charliesome>
zzak howdy
<injekt>
hey zzak
<injekt>
zzak: I was away yesterday sorry I didn't reply
<whitequark>
injekt: I think it was burnout
<whitequark>
_why really did a lot of things.
glebm has quit [Quit: Computer has gone to sleep.]
<zzak>
injekt: np!
<zzak>
charliesome: ddd hi!
<injekt>
whitequark: possibly yeah
<ddd>
i've yet to try out nokogiri in isolation. i probably should
<whitequark>
ddd: it's an okayish libxml wrapper
<ddd>
_why left prior to me entering into the ruby world, but from everything i've seen of his work, his book for new users, etc, really wish he'd find the heart to come back.
<injekt>
libxml2 is meh on its own
<whitequark>
injekt: all use expat!
boytoy has quit [Ping timeout: 264 seconds]
<whitequark>
btw, what's better?
<whitequark>
it's fast, secure, and somewhat convenient to use
<wmoxam>
ddd: what was said about rvm that was so damn upsetting?
<injekt>
whitequark: I never looked into expat
<whitequark>
injekt: it was a joke
<whitequark>
my statement about expat
noop has quit [Ping timeout: 252 seconds]
My_Hearing is now known as Mon_Ouie
<injekt>
whitequark: oh hah I didn't know about it till now
<ddd>
wmoxam: that rvm executed 'dangerous and arbitrary code', that it was somehow dangerous to use, that it used a cd hack that no one could get by (the cd 'hack' always was and always will be an opt-out/in feature. it was changed to opt-in after the debacle, but it was ALWAYS an opt-* feature.
glebm has joined #ruby-lang
odcinek has quit [Quit: odcinek]
<ddd>
wmoxam: also he TOTALLY and UTTERLY failed to mention the md5 checksum security feature on all .rvmrc files, he straight out lied in relation to the arbitrary code execution, etc
<wmoxam>
ddd: or he was misinformed
<ddd>
wmoxam: he straight out said HE had checked all this out
<charliesome>
wmoxam: when you make bold statements, you don't rely on being "informed". you check em
<ddd>
so for you to directly and viciously target the way he did you have an OBLIGATION to be informed
<wmoxam>
IMO that's overly dramatic
<wmoxam>
as far as I can remember it was the HN crowd that ripped on RVM
<wmoxam>
HN hates everything
ra0zsojsg8 has quit [Quit: Leaving.]
glebm has quit [Client Quit]
<charliesome>
wmoxam: except golang
<ddd>
wmoxam: when you increase technical support levels *drmatically* for a project through lies, innuendos, and falsities, such that they have to spend several weeks dealing with the aftermath simply so you can launch your own project and get as much exposure as possible.. you're in the wrong
<ddd>
not to mention that we spent weeks dealing with companies that suddenly were thinking they had utterly dangerous software they were having their businesses rely on.. wow
melter has joined #ruby-lang
<ddd>
the smarter ones knew better and even stood up for rvm but its the ones that didn't and the new users that were just going around on forums, MLs, etc and just regurgitating sstephenson's FUD to the point that users were getting spooked away from RVM.. yeah thats a problem.
<ddd>
in the commercial world you could sue for defamation. what does an OSS do? spend developer time on triage rather than actually developing. Our support levels quadrupled *easily* after that fiasco.
<ddd>
and to top it off, not once would he publicly respond to our *technical* discourse on the falseness of his initial claims such that we could hav eput it to rest. To me, that points out that he *enjoyed* the trouble he caused Reveled in it if you ask me. </anger>
<ddd>
-
<wmoxam>
um
<yorickpeterse>
dat wall
<wmoxam>
they still list the same points against RVM now as then
alessio_rocco has quit [Remote host closed the connection]
jxie has joined #ruby-lang
<ddd>
wmoxam: its all been debunked. he never changed his wiki
<ddfreyne>
whitequark: ah btw, another VisualWorks Smalltalk WTF: moving a file copies it and deletes the original
<ddfreyne>
whitequark: Which is not only incredibly slow, moving a file to the same filename deletes it
<ddd>
but the community as a whole has been updated, the technical points covered, and as a result the support levels are now back to what they should be. He doesn't want to update, just makes him look bad, not us.
melter_ has joined #ruby-lang
<ddd>
sorry, wife took the computer to look up some numbers
<ddd>
ddfreyne: isn't that the way that Windows implements its mv as well? one reason moving files is so slow under it.
ChuckJHardy has quit [Quit: leaving]
swav has joined #ruby-lang
<ddfreyne>
ddd: no
thebastl has quit [Remote host closed the connection]
|Vargas| has quit [Quit: ...]
thebastl has joined #ruby-lang
<ddd>
ahh, updates file reference if on same logical disk, copies then deletes upon success when moving to separate physicals. (Was googling it)
sush24 has joined #ruby-lang
vlad_starkov has joined #ruby-lang
gjaldon has joined #ruby-lang
wudofyr has quit [Remote host closed the connection]
thebastl has quit [Ping timeout: 272 seconds]
wudofyr has joined #ruby-lang
wudofyr has quit [Remote host closed the connection]
<mbj>
Is there a ruby library (other than perftools) that allows to measure cpu time?
<mbj>
s/perftools/rubyprof/
workmad3 has quit [Ping timeout: 248 seconds]
<yorickpeterse>
zzak: can't remember exactly, not sure if it even was Athens
<yorickpeterse>
home tiem, bbiab
<zzak>
lates
rwk1 has quit [Ping timeout: 272 seconds]
vlad_starkov has quit []
sush24_ has quit [Ping timeout: 256 seconds]
tylersmith has joined #ruby-lang
vlad_starkov has joined #ruby-lang
voker57 has quit [Read error: Connection reset by peer]
sush24_ has joined #ruby-lang
voker57 has joined #ruby-lang
voker57 has quit [Changing host]
voker57 has joined #ruby-lang
__butch__ has joined #ruby-lang
francisfish has quit [Remote host closed the connection]
DomKM has joined #ruby-lang
<darix>
mbj: absolute_time - maybe this helps?
DomKM has quit [Client Quit]
DomKM has joined #ruby-lang
dc5ala has joined #ruby-lang
<mbj>
darix: Thx this is better, but I need to have access to the number of cpu time spend on a specific task.
srbaker has quit [Quit: Computer has gone to sleep.]
<mbj>
darix: The absolute_time gem does only provide the number of ticks that have passed between to measurements. If the task was delayed by IO it will still be counted.
<darix>
mbj: naw i mean ... does it count the time + waiting or without
<mbj>
spike|spiegel: I'm looking for a pure ruby solution that does not require c extensions. Maybe there is something in the os specific libraries that I can hook via ffi?
<spike|spiegel>
cpu time is architecture specific
<mbj>
darix: The operation system does some accounting for you, on linux for example you can use http://linux.die.net/man/3/clock
<whitequark>
spike|spiegel: that does not account for time spent in other tasks
<whitequark>
or rather, it does and it's the problem
kennyvb has joined #ruby-lang
<spike|spiegel>
the question was related to cpu time, hence the answer.. if it was about "I want the 'real' time the cpu(s) on my machine spent running this chunck of code", that's going to be different
mistym is now known as mistym_lunch
<darix>
mbj: how about things like valgrind? do they track cpu cycles as you want?
<darix>
code mapping from valgrind back to ruby might be a problem but hm
<mbj>
darix: interesting idea
<darix>
callgrind in your case of course
<darix>
later
io_syl has joined #ruby-lang
<mbj>
darix: hehe, just found it by googling also ;)
srbaker has joined #ruby-lang
io_syl has quit [Client Quit]
francisfish has joined #ruby-lang
io_syl has joined #ruby-lang
fragamus has quit [Quit: Computer has gone to sleep.]
<spike|spiegel>
mbj: any particular reason to need cpu cycles? ruby is so inefficent (and thank GC too) it's going to blame all the wrong things
<whitequark>
spike|spiegel: well it will blame the right thing, ruby :p
<whitequark>
but yeah
<spike|spiegel>
I had a patched ruby-prof that tracked gc time inside wall time, so that I can see and correctly atrribute time to GC... and even that had given me headaches
io_syl has quit [Quit: Computer has gone to sleep.]
<mbj>
I like the idea, but an empty fs cache on a CI box will blow up where my devbox will not.
<mbj>
Realtime is not the correct metric here.
flori has joined #ruby-lang
<mbj>
But better than no metric ;)
<spike|spiegel>
good lord, that thing is jusk yuck
<spike|spiegel>
mbj: you do know that Timeout lib spawns one thread per timeout requested right?
<mbj>
spike|spiegel: yeah
<mbj>
spike|spiegel: I still like the intention behind this ;)
tbuehlmann has joined #ruby-lang
banister`sleep has quit [Ping timeout: 268 seconds]
lsegal has quit [Quit: Quit: Quit: Quit: Stack Overflow.]
krohrbaugh has quit [Quit: Leaving.]
solars has joined #ruby-lang
adambeynon has joined #ruby-lang
g0bl1n has joined #ruby-lang
rwk1 has joined #ruby-lang
<g0bl1n>
My main ruby app forks a new process, which I get the PID. That new process launches a system app (sinatra rackup). When I kill the PID, the sinatra process keeps alive. If I kill the Process.getpgid(PID) only the parent process is killed. Any hint on how to kill the PID and the rackup sinatra child ?
Silex^^ is now known as Silex
vlad_sta_ has joined #ruby-lang
<whitequark>
trap(:TERM) { kill the child }
retro|cz has quit [Ping timeout: 256 seconds]
cruffinski has joined #ruby-lang
rwk1 has quit [Ping timeout: 252 seconds]
vlad_starkov has quit [Ping timeout: 248 seconds]
cruffinski has quit [Remote host closed the connection]
vlad_sta_ has quit [Client Quit]
vlad_starkov has joined #ruby-lang
imajes has quit [Excess Flood]
imajes has joined #ruby-lang
pkrnj has joined #ruby-lang
r0bgleeson has quit [Ping timeout: 252 seconds]
Silex has quit [Quit: WeeChat 0.3.8]
gnufied has quit [Quit: Leaving.]
erpuds has quit [Quit: erpuds]
Silex has joined #ruby-lang
jovy88 has quit [Read error: Connection reset by peer]
odcinek has quit [Quit: odcinek]
jovy88 has joined #ruby-lang
celinedior has quit [Quit: celinedior]
anachronistic has quit [Quit: anachronistic]
odcinek has joined #ruby-lang
sepp2k has joined #ruby-lang
divout has quit [Remote host closed the connection]
<g0bl1n>
whitequark: thanks, now using Process.spawn instead of Process.fork + system
krohrbaugh has joined #ruby-lang
srbaker has quit [Quit: Computer has gone to sleep.]
brianpWins has joined #ruby-lang
r0bgleeson has joined #ruby-lang
mrsolo has joined #ruby-lang
GarethAdams has quit [Quit: Leaving...]
odcinek has quit [Quit: odcinek]
vlad_sta_ has joined #ruby-lang
symm- has joined #ruby-lang
vlad_starkov has quit [Ping timeout: 268 seconds]
melter_ has quit [Ping timeout: 264 seconds]
melter has quit [Ping timeout: 264 seconds]
vlad_sta_ has quit [Remote host closed the connection]
bnagy has quit [Ping timeout: 252 seconds]
bnagy has joined #ruby-lang
Uranio has joined #ruby-lang
thone has quit [Remote host closed the connection]
<yorickpeterse>
whitequark: hm, it seems Builders::Default#n() isn't actually used
<yorickpeterse>
at least, when I subclass it and override it nothing changes
<IamBPAfree>
how do I search an array to see if it CONTIANS something. I know that if i use array.include? 'mytext' it will search the array for the element 'mytext'. But what I need is not an exact match. Like the array element my be 'thisismytextstring' but I need to search all the elements in the array for 'mytext' and need it to be TRUE when compared to the element 'thisismytextstring'
<yorickpeterse>
IamBPAfree: myarray.select { |element| element =~ /some_regex/ }
<mistym>
Or ary.any? {}, which returns a boolean instead of an array.
imajes has quit [Excess Flood]
krz has quit [Quit: krz]
rwk1 has quit [Ping timeout: 258 seconds]
<TTilus>
IamBPAfree: array.grep /mytext/
mbj has quit [Ping timeout: 256 seconds]
<IamBPAfree>
thanks for the suggestions....I think I can get .any? {}. boolean should be fine for my purposes.
mbj has joined #ruby-lang
<TTilus>
whitequark: i wast aware that many regex implementations are that bbad
imajes has joined #ruby-lang
<whitequark>
TTilus: MOST of them
vlad_starkov has joined #ruby-lang
<whitequark>
yorickpeterse: it is indeed used
<TTilus>
whitequark: sadly
<yorickpeterse>
whitequark: hmmmm
<TTilus>
IamBPAfree: !array.grep(/mytext/).empty?
swygue has joined #ruby-lang
<TTilus>
IamBPAfree: but .any? is a better idea
<yorickpeterse>
whitequark: hm, it seems that in the Git HEAD it does work
pkrnj has quit [Ping timeout: 268 seconds]
<yorickpeterse>
but not with the currently released Gem
mbj has quit [Ping timeout: 246 seconds]
<mistym>
.any? also calls fewer methods and is ever so slightly faster than !ary.grep(//).empty?, if that matters to you
<whitequark>
yorickpeterse: you probably want to use HEAD anyway
<yorickpeterse>
probably
mbj has joined #ruby-lang
<yorickpeterse>
heh, invalid Gemspec
<whitequark>
yorickpeterse: how so
<yorickpeterse>
ERROR: While executing gem ... (Gem::InvalidSpecificationException) ["bin/bin/ruby-parse", "bin/bin/ruby-rewrite"] are not files
<whitequark>
whoops
<yorickpeterse>
(don't want to use a bundle local config since I'll then probably forget about it)
<whitequark>
however... that's weird
<whitequark>
gem build works for me
<yorickpeterse>
you're using some regex to get two files, which is a bit silly