<havenwood>
lystra_: Does it work without the --local flag?
ur5us has joined #ruby
<havenwood>
lystra_: You want to install it from the remote source, yeah?
<havenwood>
gem install rails --local #!> ERROR: Could not find a valid gem 'rails' (>= 0) in any repository
eckhardt has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
dviola has quit [Changing host]
dviola has joined #ruby
AJA4350 has quit [Quit: AJA4350]
dbugger_ has joined #ruby
<havenwood>
lystra_: The default is --both, which allows local or remote sourced gems to be installed. It seems you're meaning to install from a remote source, so the default or --remote will work, but you don't have loquacious.gem locally, so it can't be found with --local.
bayprogrammer has quit [Quit: WeeChat 2.0.1]
<Radar>
seems a bit like "Installing Ruby: The Hard Way"
Dbugger has quit [Ping timeout: 245 seconds]
tvw has quit [Remote host closed the connection]
bayprogrammer has joined #ruby
<Radar>
But yeah, I think havenwood is right. --local is probably the cause of the problem here.
dbugger__ has joined #ruby
jenrzzz has quit [Ping timeout: 256 seconds]
dbugger_ has quit [Ping timeout: 252 seconds]
Dbugger has joined #ruby
dbugger__ has quit [Ping timeout: 260 seconds]
eckhardt has joined #ruby
eckhardt has quit [Client Quit]
nertzy has quit [Read error: Connection reset by peer]
nertzy has joined #ruby
Puffball has joined #ruby
white_lilies has joined #ruby
Deesl has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Guest85807 is now known as ged
ged is now known as Guest36153
hlmjr has quit [Read error: Connection reset by peer]
<lystra_>
havenwood: Let me try.
GodFather has joined #ruby
<lystra_>
havenwood: That worked. Thanks! I was using the same installation procedure as with ruby-2.0 so I expected things to work for future Ruby versions as well. Thanks a lot.
rivalomega has joined #ruby
<lystra_>
havenwood: I don't really get the difference between local/remote but things work with --local.
<havenwood>
lystra_: With --local you're saying you have the loquacious.gem file locally, and you'd like to install it from your local version.
venmx has quit [Ping timeout: 268 seconds]
<lystra_>
Radar: Yeah, it might not seem intuitive but we want to keep the base Ruby installation and any 3rd-party gems separate. We do something similar with Python, Perl, Tcl/Tk, etc.
kmurphy4 has joined #ruby
<havenwood>
lystra_: With --both (default) or --remote, it'll look for the gem in your remote sources as well (default https://rubygems.org/).
kapil___ has quit [Quit: Connection closed for inactivity]
<havenwood>
lystra_: See what you have set for remote gem sources: gem env remotesources
<havenwood>
lystra_: Yup, just the --local flag was overriding the --both default, telling RubyGems not to look at https://rubygems.org/.
<havenwood>
lystra_: Without already having a local copy of those gems, it couldn't find them.
<lystra_>
havenwood: Well, we have /tmp/ruby24p/vendor/lib/2.4.0/cache/loquacious-1.9.1.gem which I would expect it to find with --local.
alfiemax has joined #ruby
<lystra_>
havenwood: Actually, our preference is for gem install not to fetch anything from the Internet.
alfiemax has quit [Ping timeout: 256 seconds]
<lystra_>
havenwood: BTW, if I install to the default directory (i.e. omit --bindir and --install-dir), then even with --local, I can install little-plugger, loquacious, and bones successfully. Only when installing outside of the default installation path is there a problem. So, I think the problem remains that Ruby is not search out custom installation directory for gems to match the dependency.
<havenwood>
lystra_: I'm curious if we'd be able to give better suggestions if we knew more about what you're doing.
<lystra_>
havenwood: I just want to install all gems (via gem install) to a separate directory from where ruby was installed.
<lystra_>
havenwood: So if I install ruby 2.4.4 to /tmp/ruby24, I want to install all gems for it to /tmp/ruby24p.
<havenwood>
lystra_: Normally in prod we use a Gemfile.lock to specify the exact gem versions then install them to vendor/bundle relative to the root of the project.
<havenwood>
lystra_: If you don't want to download remote gems, how do they get on the server?
<havenwood>
lystra_: You can pass the location of the .gem file to --local.
Puffball has quit [Quit: Puffball]
<lystra_>
havenwood: I download them manually. I just don't want them downloaded automatically. Sorry for the poor wording.
<havenwood>
lystra_: You can also run your own local gemserver. It's hard to give recommendations without knowing more about why you're doing what you're doing.
<havenwood>
lystra_: no prob
<lystra_>
havenwood: We're packaging Ruby for some customers. When we provide an RPM to them, Ruby will be installed to /usr/local/ruby24. The customers also want some gems and we don't want to install the gems to the same directory as Ruby so we're installing the gems to /usr/local/ruby24p. We also want Ruby, when someone requires the gem, to search for the gems in /usr/local/ruby24 *and* /usr/local/ruby24p.
<havenwood>
lystra_: I'd say just pass the location of the *.gem to --local.
<lystra_>
havenwood: I am installing with /tmp/ruby24/bin/gem install -V --rdoc --ri --bindir=/tmp/ruby24p/bin --install-dir=/tmp/ruby24p/vendor/lib/2.4.0 --local /opt/src/2018.01.01/ruby/bones-3.8.4/src/bones-3.8.4.gem. Isn't that passing the location of the gem file?
pilne has quit [Quit: Leaving]
bga57 has quit [Ping timeout: 245 seconds]
alfiemax has joined #ruby
bga57 has joined #ruby
<havenwood>
lystra_: yes, looks right to me
<lystra_>
havenwood: That's the command that's failing. It succeeds without --local.
<havenwood>
lystra_: --rdoc and --ri are deprecated, you can use --document instead but you're not changing the default so I'd omit them
alfiemax has quit [Ping timeout: 245 seconds]
Azure|dc is now known as Azure
<lystra_>
havenwood: Ok, thanks. I see that with 'gem install --help. Didn't notice it. Holdovers from how we installed things with Ruby 2.0.
<havenwood>
lystra_: Sanity check that .gem file really is there?: file /opt/src/2018.01.01/ruby/bones-3.8.4/src/bones-3.8.4.gem
<lystra_>
havenwood: Yep. I use the same command line for Ruby 2.0 but just change 24 -> 20 and it works fine.
<lystra_>
havenwood: If the .gem file wasn't there, it wouldn't know loquacious was a dependency and generate the error message about not finding it.
<lystra_>
havenwood: If I strace the command, /tmp/ruby24p/vendor/lib/2.4.0/specifications is never searched unlike when I use Ruby 2.0.
<havenwood>
lystra_: oh! you're not trying to install loquacious, it's a dep of a gem?
<havenwood>
I mean not trying to install it directly?
<lystra_>
havenwood: Correct. bones needs little-plugger and loquacious. We manually installed both using the same method. Then, when we try to install bones, it fails because it cannot find its dependencies. So our custom gem installation path isn't getting searched for dependencies.
<havenwood>
lystra_: So you're installing the gems while online with --install-dir BLARG then moving all those gems and their deps over to your server from BLARG/cache/*.gem?
RyanMcCoskrie has joined #ruby
<havenwood>
lystra_: Interesting. I don't know why it's not finding the deps.
<lystra_>
havenwood: No, we're installing directly to our custom gem path, /tmp/ruby24p for now to test things. When we have this issue figured out, we'll rebuild Ruby and the gems and install to /usr/local/ruby24 and /usr/local/ruby24p and then generate an RPM directly from that directory. We're doing all this while online but using local paths to find all the gems.
<lystra_>
havenwood: Ok, thanks. I'll post to ruby-talk and hopefully someone will have an answer.
j416 has quit [Ping timeout: 256 seconds]
RougeR has quit [Ping timeout: 240 seconds]
RyanMcCoskrie has quit [Remote host closed the connection]
gnufied has quit [Remote host closed the connection]
Guest36153 is now known as ged
coderphive has joined #ruby
ged is now known as Guest22663
j416 has joined #ruby
sleetdrop has joined #ruby
cyberg has quit [Remote host closed the connection]
enigma has joined #ruby
enigma has quit [Client Quit]
sleetdrop has quit [Quit: Computer has gone to sleep.]
coderphive has quit [Quit: coderphive]
sleetdrop has joined #ruby
millerti has quit [Ping timeout: 240 seconds]
alfiemax has joined #ruby
hlmjr has joined #ruby
krawchyk has joined #ruby
alfiemax has quit [Ping timeout: 245 seconds]
tdy has quit [Ping timeout: 252 seconds]
cadillac_ has joined #ruby
thapakazi has joined #ruby
braincrash has quit [Quit: bye bye]
elphe has quit [Ping timeout: 252 seconds]
braincrash has joined #ruby
conta has joined #ruby
white_lilies has quit [Ping timeout: 264 seconds]
conta has quit [Ping timeout: 245 seconds]
howdoi has joined #ruby
alfiemax has joined #ruby
sleetdrop has quit [Quit: Computer has gone to sleep.]
alfiemax_ has joined #ruby
alfiemax has quit [Ping timeout: 256 seconds]
white_lilies has joined #ruby
ramfjord has quit [Ping timeout: 252 seconds]
elphe has joined #ruby
sleetdrop has joined #ruby
cagomez has joined #ruby
orbyt_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Radar>
Is roo still the latest and greatest gem to use for reading Excel files?
coderphive has joined #ruby
alfiemax_ has quit [Remote host closed the connection]
Deesl has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
lxsameer has joined #ruby
sysvalve has joined #ruby
mlkkk has joined #ruby
Guest23707 has quit [Ping timeout: 240 seconds]
elphe has quit [Ping timeout: 240 seconds]
yohji has joined #ruby
NL3limin4t0r has joined #ruby
marr has joined #ruby
dionysus69 has joined #ruby
Guest36902 is now known as SuperTux88
Burgestrand has joined #ruby
Mia has joined #ruby
Mia has joined #ruby
Mia has quit [Changing host]
duderonomy has joined #ruby
mrush has joined #ruby
mrush is now known as Guest92733
redlegion has quit [Remote host closed the connection]
redlegion has joined #ruby
Guest41156 is now known as Sp4rKy
Sp4rKy is now known as Guest17152
alfiemax has quit [Remote host closed the connection]
Guest17152 is now known as henry
Guest92733 has quit [Ping timeout: 252 seconds]
henry is now known as Sp4rKy
alfiemax has joined #ruby
Sp4rKy is now known as Guest34758
Guest2576 is now known as ged
alfiemax has quit [Remote host closed the connection]
ged is now known as Guest26197
alfiemax has joined #ruby
Guest34758 is now known as Sp4rKy
guille-moe has joined #ruby
KeyJoo has quit [Remote host closed the connection]
elphe has joined #ruby
duderono_ has joined #ruby
amar has joined #ruby
alfiemax has quit [Ping timeout: 268 seconds]
duderonomy has quit [Ping timeout: 256 seconds]
mrush_ has joined #ruby
marr has quit [Ping timeout: 264 seconds]
<ellcs>
guys, can someone explain me why this behavior exists? xD
<ellcs>
`nil.to_i # => 0`
<apeiros>
that's a very good question and I have no good answer for it.
<dminuoso>
ellcs: nil.to_a works too
<dminuoso>
nil.to_s works too
<dminuoso>
lots of surprising things.
<apeiros>
I guess you'd have to ask matz
<apeiros>
nil.to_s at least makes some sense. though you could argue about "" vs "niL"
lypsis|awy is now known as lypsis
<apeiros>
but yeah, nil.to_i, .to_f, .to_<numeric>, .to_a, .to_h all make no sense IMO
<dminuoso>
nil.to_s makes sense?
<dminuoso>
heh
<dminuoso>
oh goodie!
<dminuoso>
apeiros is on the righteous path of type theory
<apeiros>
sure. "printable representation of the object" makes sense for almost all objects.
mtkd has quit []
<dminuoso>
>> nil.to_str
Asher has joined #ruby
<dminuoso>
>> nil.to_ary
<phaul>
however is nil an object that represents the absence of any object? :)
<apeiros>
yes, that's its purpose. implementing the nullobject pattern :)
<ellcs>
i dont want to start a war, but aren't there two concepts mixed up? neutral elements and non-existence?
<ellcs>
but i see the purpose now, thanks!
mtkd has joined #ruby
<apeiros>
ellcs: that's probably where it comes from. but neutral element in numbers isn't always 0 either. it depends on the operation (e.g. multiplication, 0 is not neutral, 1 is). which makes 0 even less of a meaningful choice.
<phaul>
I mean what I said there feels like a contradiction in itself. :) but it's rather philosophical and a bit pointless argument..
<apeiros>
I understood. but having a value representing the absence of a value is a widely used concept, known as the nullobject pattern. it's not just limited to nil.
<apeiros>
in an authentication/authorization system you could e.g. have a NullUser
<apeiros>
so while it may be contradictory, it's useful
<phaul>
true
<apeiros>
I'd love it if ruby allowed subclassing NilClass for this purpose
<apeiros>
so that: `class NullUser < NilClass; end; user =NullUser.new; user ? true : false` => false
kapil___ has quit [Quit: Connection closed for inactivity]
<ellcs>
but why subclassing it, i mean you can just use nil itself
<apeiros>
can you do nil.privileges ?
<apeiros>
or nil.has_privilege?("edit_stuff")?
<ellcs>
no, but your programm crashes at this point
<apeiros>
which is not what I want
<phaul>
but doesn't that mean u need to sublcass User?
<phaul>
at least for the interface
<apeiros>
all I need is a duck
<ellcs>
i see
<apeiros>
if it quacks like a user and walks like a user, it is a user. it just is not allowed to do anything :)
<ellcs>
well, you might have a user "nil" entry xD
<apeiros>
it simplifies the design a lot.
<apeiros>
without the nulluser, I either have crashing apps, or I have to take care of nil *everywhere*
<phaul>
yeah, I'm just not seeing why subclassing NilClass is needed
<ellcs>
i mean, having a 'nil'-user in your database has the same benefits.
<apeiros>
it isn't. it'd just make a couple of things nicer.
<apeiros>
without that, I just use predicates
<apeiros>
user.logged_in? f.ex.
<apeiros>
hardcoded to true for User, hardcoded to false for NullUser
<apeiros>
but it means you have to invent a predicate for every kind of class
t0xik has joined #ruby
dhollinger has quit [Read error: Connection reset by peer]
<apeiros>
I'd rather treat NullObjects as falsy.
XB23 has joined #ruby
yohji has quit [Remote host closed the connection]
dhollinger has joined #ruby
venmx has joined #ruby
Charliebr73 has joined #ruby
Charliebr73 has left #ruby [#ruby]
Charliebr73 has joined #ruby
<phaul>
where Im unsure about your design is that would you def has_priviledge? in your NullUser < NilClass? Or would have checks outside
<apeiros>
NullUser has the same interface as User
<apeiros>
since NullObjects usually deviate wildly from their non-null counterpart, a shared codebase was rarely a consideration
<apeiros>
but yeah, maybe we could use a different approach than subclassing to turn an entire class falsy
<phaul>
at least to_s is there. I was thinking based on the same argument as the NullUser should have the interface of the User, NilClass should have the same interface as Object. But object doesn't respond to those things..
<ruby[bot]>
apeiros: # => [:instance_of?, :public_send, :instance_variable_get, :instance_variable_set, :instance_variable_def ...check link for more (https://eval.in/1009794)
<dminuoso>
Usual disclaimer: I will not be held responsible if you run this in production code. ;o
<catphish>
basically, i can't extend primative classes, so doing something like this to carry them around in a wrapper https://paste.ubuntu.com/p/FFyNh9BDHS/
<catphish>
not the actual code, i paraphrase, but that's the concept, not sure why it's necessary to define coerce explicitly here
<catphish>
not sure how ugly this is, but it works
codymj has joined #ruby
chouhoul_ has quit [Remote host closed the connection]
chouhoulis has joined #ruby
gnufied has joined #ruby
<apeiros>
catphish: your object probably recevies a respond_to?(:coerce)
<apeiros>
so if you omitted a proper respond_to_missing? impl, then it won't even try to invoke .coerce
chouhoulis has quit [Ping timeout: 240 seconds]
<apeiros>
and seeing your paste, you did in fact not add a proper respond_to_missing. you should amend that. there's a couple of things which will not work properly without it.
Burgestrand has quit [Quit: Good bye and have a nice day!]
<apeiros>
baweaver: ruby tip for tomorrow: "if you implement method_missing, also implement a corresponding respond_to_missing?" ;-)
roshanavand has quit [Ping timeout: 245 seconds]
alfiemax has quit [Ping timeout: 252 seconds]
<catphish>
apeiros: thanks, i'll implement respond_to_missing, it seems that at least #corerce and #to_s are implicitly definied, and so don't go to method_missing
<catphish>
or it may be the result of my respond_to_missing, or a combination :)
<dminuoso>
It directly pokes in the ruby land vm and changes the class pointer!
<dminuoso>
Haha
sysfault has quit [Ping timeout: 252 seconds]
marr has joined #ruby
Guest76059 is now known as ged
sysvalve has quit [Read error: Connection reset by peer]
ged is now known as Guest57223
sysvalve has joined #ruby
jarnalyrkar has joined #ruby
tag has quit [Quit: Connection closed for inactivity]
mjolnird has quit [Remote host closed the connection]
kmurphy4 has joined #ruby
gigetoo has quit [Read error: No route to host]
gigetoo has joined #ruby
Asher has quit [Ping timeout: 245 seconds]
jarnalyrkar has quit [Quit: Leaving]
jarnalyrkar has joined #ruby
chmurifree has quit [Ping timeout: 256 seconds]
shinnya has joined #ruby
XB23 has quit [Quit: Leaving]
mostlybadfly has joined #ruby
aupadhye has quit [Remote host closed the connection]
dionysus69 has quit [Quit: dionysus69]
chouhoulis has joined #ruby
duderono_ has quit [Read error: Connection reset by peer]
duderonomy has joined #ruby
bmurt has joined #ruby
rippa has joined #ruby
ta_ has joined #ruby
norfumpit has joined #ruby
mjolnird has joined #ruby
<norfumpit>
is it safe to change value of Net::HTTP::HTTPVersion inside a passenger environment? i think I might be having some concurrency issues due to my changing that constant around a block of code
ferr_ has quit [Quit: WeeChat 2.1]
quobo has quit [Quit: Connection closed for inactivity]
viq has left #ruby [#ruby]
houhoulis has joined #ruby
jyaworski has joined #ruby
rivalomega has joined #ruby
mtkd has quit [Ping timeout: 268 seconds]
mtkd has joined #ruby
rivalomega has quit [Remote host closed the connection]
grilix_ has joined #ruby
Rapture has joined #ruby
Asher has joined #ruby
paul_ has quit [Ping timeout: 260 seconds]
Asher has quit [Ping timeout: 240 seconds]
krawchyk[away] has joined #ruby
biberu has quit [Ping timeout: 256 seconds]
chouhoulis has quit [Remote host closed the connection]
bmurt has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
krawchyk[away] has quit [Client Quit]
conta has quit [Ping timeout: 268 seconds]
bmurt has joined #ruby
<apeiros>
catphish: sorry, got distracted by RL. you should google for examples of how to implement numeric-like classes in ruby. iirc there are good examples around.
houhoulis has quit [Remote host closed the connection]
kmurphy4 has quit [Quit: kmurphy4]
ta_ has quit [Ping timeout: 245 seconds]
Guest57223 is now known as ged
venmx has quit [Ping timeout: 268 seconds]
ged is now known as Guest34091
kmurphy4 has joined #ruby
venmx has joined #ruby
jyaworski has quit [Ping timeout: 245 seconds]
sysfault has joined #ruby
<catphish>
apeiros: RL is boring
\13k has quit [Ping timeout: 256 seconds]
syndikate has quit [Ping timeout: 256 seconds]
bmurt has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kmurphy4 has quit [Ping timeout: 240 seconds]
apparition has quit [Quit: Bye]
syndikate has joined #ruby
eckhardt_ has joined #ruby
\13k has joined #ruby
alfiemax has joined #ruby
bmurt has joined #ruby
phaul has quit [Ping timeout: 260 seconds]
DTZUZO has joined #ruby
DTZUZO has quit [Ping timeout: 260 seconds]
mlkkk has quit [Remote host closed the connection]
mlkkk has joined #ruby
jamiejackson has joined #ruby
Mike11 has quit [Quit: Leaving.]
bmurt has quit [Ping timeout: 248 seconds]
hlmjr has quit [Ping timeout: 245 seconds]
mlkkk has quit [Ping timeout: 256 seconds]
chouhoulis has joined #ruby
hlmjr has joined #ruby
Charliebr73 has quit []
DTZUZO has joined #ruby
mostly-harmless has joined #ruby
eckhardt_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
apeiros has quit [Remote host closed the connection]
conta has joined #ruby
lxsameer has joined #ruby
beefjoe70 has joined #ruby
coderphive has joined #ruby
ShekharReddy has quit [Quit: Connection closed for inactivity]
<havenwood>
arne: super without parens passes on args automagically
<havenwood>
arne: super() with parens doesn't
<arne>
havenwood: well, exactly that's what i expected
shinnya has quit [Ping timeout: 240 seconds]
thinkpad has quit [Ping timeout: 248 seconds]
<arne>
but super() will still send the block
<arne>
and so will super(some_fancy_params*)
<havenwood>
arne: yeah, that's true
<arne>
wow, that really caught me offguard
thinkpad has joined #ruby
emilford has quit [Ping timeout: 248 seconds]
<yehdawg>
@havenwood thank you!
amar has joined #ruby
jenrzzz has joined #ruby
emilford has joined #ruby
segy has quit [Ping timeout: 240 seconds]
<NL3limin4t0r>
yehdawg: The Ruby docs have also some documentation that pretty usefull when new to Ruby. See the first alinea of http://ruby-doc.org/core-2.5.1/ This however is not in tutorial form and just shows you things like controll expressions, assignment, etc.
venmx has quit [Ping timeout: 260 seconds]
yehdawg has quit [Ping timeout: 252 seconds]
segy has joined #ruby
segy has quit [Excess Flood]
emilford has quit [Ping timeout: 240 seconds]
Immune has joined #ruby
yehdawg has joined #ruby
roshanavand has joined #ruby
<apeiros>
havenwood: I really wish they'd have put this into JSON.<something>_load instead of plain .load
mostlybadfly has quit [Quit: Connection closed for inactivity]
<apeiros>
that kind of thing is a primary source for security issues
emilford has joined #ruby
Immune has left #ruby [#ruby]
sytherax has quit [Remote host closed the connection]
nachoman has joined #ruby
BTRE has joined #ruby
arashb has joined #ruby
alfiemax_ has quit [Ping timeout: 268 seconds]
NL3limin4t0r has quit [Quit: WeeChat 1.9.1]
beefjoe has joined #ruby
sytherax has joined #ruby
clemens3 has joined #ruby
sytherax has quit [Read error: Connection reset by peer]
sytherax has joined #ruby
nachoman has quit [Ping timeout: 245 seconds]
Puffball has joined #ruby
thapakazi has quit [Quit: thapakazi]
sanscoeur has joined #ruby
segy has joined #ruby
karapetyan has joined #ruby
mostlybadfly has joined #ruby
nitric_ has joined #ruby
roshanavand has quit [Ping timeout: 256 seconds]
physicist__ has joined #ruby
nitric has quit [Ping timeout: 245 seconds]
karapetyan has quit [Ping timeout: 260 seconds]
jyaworski has joined #ruby
jrich523 has joined #ruby
<jrich523>
what are the things like %Q{} called?
alfiemax has joined #ruby
jenrzzz has quit [Read error: Connection reset by peer]
jenrzzz_ has joined #ruby
<elomatreb>
Percent-literals is the usual term I think
Deesl has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<baweaver>
That should get you on a stable version.
<baweaver>
Latest is 2.5.1 on Ruby
<baweaver>
but the instructions are pretty well the same.
<Glenjamin>
ah ok, ta - i'll try an up-to-date ruby
banisterfiend has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
venmx has joined #ruby
<baweaver>
Yeah, anything that sudo installs a gem is a mess waiting to happen, because then you have an arms race of always needing it to do anything
<baweaver>
especially because sudo will root own everything it touches on that
<baweaver>
which is likely the cause of your problems
cthu| has quit [Ping timeout: 260 seconds]
<Glenjamin>
but the symptoms don't match that
xuyuheng has quit [Ping timeout: 260 seconds]
<Glenjamin>
bundler correctly prompts for sudo to try and install into the system gem folder, and the permissions error is on a tempdir
<Glenjamin>
i'm thinking this might be because of the version of rubygems that comes with 2.3.0 though
emilford has quit [Ping timeout: 245 seconds]
Guest98248 is now known as ged
emilford has joined #ruby
ged is now known as Guest62306
<Glenjamin>
hrm, I've added some debug logging into bundler to help understand the issue more - the generated tempdir itself is user-owned, but the `bin` dir inside it is root owned
<zenspider>
can anyone confirm that bbatsov/rubocop doesn't have a clean test run out of the box (I *just* checked out)
jenrzzz has quit [Read error: Connection reset by peer]
jenrzzz has joined #ruby
<zenspider>
ah. and it appears that it is literally running the tests twice and doco and and and when you run the default task.
<Glenjamin>
aha, found it - it's a bug in bundler that only kicks in under weird circumstances
<havenwood>
zenspider: cloned and run locally with 2.5.1, seeing two test runs and both passing with pendings
<zenspider>
havenwood: thanks. I'm seeing that too. Is there a cheap/easy way to filter out the pendings so I don't even see them?
<Glenjamin>
if `requires_sudo?` is true, and rubygems is < 2.7.5, then it runs `mkdir_p bin_path`, but internally that also checks `requires_sudo?` and makes the dir as root
<zenspider>
WHAT I REALLY WANT: a good ruby project that uses rspec idiomaticall and dosn't have nasty external dependencies
<zenspider>
preferably faster than these... 45 seconds for the one run is a bit much... totally doable... I could also just run a subset
kirun has quit [Quit: Konversation terminated!]
emilford has joined #ruby
xuyuheng has joined #ruby
venmx has quit [Ping timeout: 240 seconds]
<zenspider>
time rspec spec/rubocop/formatter/ # => 2.88 seconds. that's doable
<zenspider>
but man... 2.88 seconds for 1.12 seconds worth of tests. WTF?+
jmux has joined #ruby
ur5us has quit [Remote host closed the connection]
<zenspider>
"files took 1.04 seconds to load"... that means 25% of the time is unaccounted for