tomjoro has quit [Remote host closed the connection]
<GitHub67>
[jruby] pitr-ch commented on commit e797826: It's of minor importance. No trailing spaces is a simple rule (which can be usually enforced by an editor) that should prevent pollution of commits with a white-space changes assuming everybody has the same setting enabled. I do not remember that we ever discussed it though, I am just used to do it this way. Therefore I've kept the white-space removal there. I should have done an extra commit though. https://git
tomjoro has joined #jruby
<GitHub86>
[jruby] pitr-ch commented on commit 673a4c3: Yeah I agree, the exception was added based on someones suggestion so I'll defer to @eregon / @nirvdrum. https://git.io/vg9g3
jensnockert has joined #jruby
pawnbox has quit [Remote host closed the connection]
<GitHub34>
[jruby] nirvdrum commented on commit 673a4c3: I don't mind the use of `UnsupportedOperationException`, but I'd like there to be a message for the exception. It makes debugging a lot easier. https://git.io/vgHWs
jensnockert has joined #jruby
<GitHub94>
[jruby] chrisseaton commented on commit 673a4c3: Adding a custom message at each place where the exception is thrown should lead to better messages than a default message in a subclass. For example 'This branch of code should be never reached, because we already know that this value should be in this range'. https://git.io/vgHWg
phlebas has joined #jruby
thedarkone2 has joined #jruby
jensnockert has quit [Ping timeout: 240 seconds]
<GitHub174>
[jruby] nirvdrum commented on commit 673a4c3: Agreed. I'm just noting that they're currently all blank. https://git.io/vgHWj
<robacarp>
norc_: what sort of parser guru are you looking for?
nirvdrum has quit [Ping timeout: 265 seconds]
<norc_>
robacarp, take this code: p a if a = 1
<norc_>
I know why MRI will choke on this (the reason is that the parser builds up a lookup table for dvars at parse time)
<norc_>
In order of discovery.
<norc_>
*lvars
<norc_>
Not knowing much about jruby I expect the behavior to be mimicked, since jruby emits a NameError too.
<norc_>
In MRI I would have expected the compiler to build that cache when walking the AST.
<norc_>
And considering how the compiler implements unless_modifier and if_modifier, this behavior seems inconsistent, so I wanted to know the stance from jruby on this.
<donV>
To all: anybody up for confirming this issue? Would be great to know it is not just me, or if I am having the wrong expectation for the program.
<koochdog>
Hi all, I'm getting a weird bug hoping someone can help me figure out where to go. I using warble to create a WAR of a ruby on rails project. We have a need to run this application on a sparc server and I am getting an error on startup of "getppid unsupported or native support failed to load"
<norc_>
Granted, I am a bit scared about the tons of stuff that ruby-install is installing at the moment, but Im sure all of this is needed.
<norc_>
donV, there is no 9.0.2.0 release.
<donV>
Hmm
<donV>
norc_: Well, there is the tag for it, but I think there was a bad bug in it, so 9.0.3.0 was released really quickly.
<donV>
Testing 9.0.3.0 instead is great as well.
<donV>
…so you can skip 9.0.2.0 if you can’t find it. I see no binary release for 9.0.2.0 anywhere, either.
<norc_>
donV, wow the jvm bootup time is horrifying.
<donV>
yes.
<donV>
it is.
<norc_>
Is there some way of preloading it?
<donV>
I try not to lie awake at night thinking about it .
<norc_>
donV, 9.0.1.0 does not reproduce the issue
<donV>
There are several preload projects.
<norc_>
9.0.3.0 does
<donV>
Woohoo! Thank you!
<norc_>
master I would have to build first
<donV>
Now, it the example valid? I think so, but a second opinion would be great.
<donV>
I think master is not that important.
<donV>
I am already satisfied.
<norc_>
Testing jruby-1.7.24 still thoguh
<norc_>
Just for giggles.
<norc_>
Success!
<donV>
Yes!
<norc_>
Let me just respond to the issue
<donV>
Thanks!
<norc_>
donV, if you want to create a real test case
<norc_>
you should use plain old ruby autoload thoguh
<koochdog>
Hi all, I'm getting a weird bug hoping someone can help me figure out where to go. I using warble to create a WAR of a ruby on rails project. We have a need to run this application on a sparc server and I am getting an error on startup of "getppid unsupported or native support failed to load"
<chrisseaton>
Try -Xnative.enabled=false
<chrisseaton>
And, separately, .verbose=true
<koochdog>
Chrisseaton try that on a non-spark machine to replicate the issue?
<chrisseaton>
Try that on your actual machine
<koochdog>
Ok will do, thanks unforunately I don't have access to the sparc machine so I have to wait until that office is online
<koochdog>
and emulating sparc seems to be pretty much impossible
<chrisseaton>
Ha it's going to be slow debugging it then :)
<koochdog>
Very
<chrisseaton>
-Xnative.enabled=false should stop it trying to use this method, so that likely will fix your problem, but then the problem maybe that someone really needs that method, and the fallback doesn't work
<koochdog>
Right, unless it is implemented in Java?
<chrisseaton>
Yeah
<chrisseaton>
And it isn't
<koochdog>
When running the WAR where would I want to pass that in?
<chrisseaton>
But something may be designed to detect that and do something else
<chrisseaton>
Sorry I have no ideas about WARs
<chrisseaton>
Maybe JRUBY_OPTS? I'm just guessing
<koochdog>
no problem, it's on an application container I'm not familiar with so I'll let that team handle it
<d-snp>
hey chrisseaton did you see I fixed my problem yesterday by making a native Java JRuby launcher? I think the problem was with the JRE that ships with the JRuby binary distro
<d-snp>
anyway, I can control starcraft from jruby now :)
<norc_>
say what
<koochdog>
chrisseaton: Do you happen to know if this is pretty important to jruby from org/jruby/RubyProcess.java:752:in `ppid'
<chrisseaton>
The problem is I don't know if it's lazy or not - is the method being resolved, and failing, as someone is actually trying to use it, or is it just loading by default and so doesn't matter
<chrisseaton>
Can you see locally if your app calls ppid (maybe indirectly or course)?
donV has quit [Quit: donV]
<koochdog>
There is a pretty ugly stacktrace of library requring library
<koochdog>
That appears to be the line in question
<chrisseaton>
Right I get that, but where in your application does this stack trace lead you to?
<chrisseaton>
Is it before your app starts, or as your app requires something, or runs some code?
<koochdog>
Just rails startup
<koochdog>
Doesn't seem to be anything outside of rails calling an environment file to populate configs then it works its way through various libraries
<chrisseaton>
Can you be more specific? Is there RubyKernel#require on the stack trace?