<headius[m]1>
gah, I marked it for 9.2.14 but did it against master...
<enebo[m]>
headius: yeah I don't know
<enebo[m]>
It is a nice to address the reported issue but I think we need to fully map out how behavior changes since this also hits JRubyFile
<headius[m]1>
the JRubyFile part is small and just eliminates the possibility of a double slash at the beginning of the URI path
<headius[m]1>
I would not expect that to break anything since double slash can always be simplified to single
<headius[m]1>
my main concern was that it's now using getAbsolutePath from the JRubyFile-produced resource, which could have unintended consequences
<headius[m]1>
we'll see how CI shakes out I suppose
subbu is now known as subbu|lunch
<enebo[m]>
those last two sentences at this point would be a hard no but yeah I will also look at it more and see if anything jumps out
<enebo[m]>
mostly just that .14 really should be a true solid end point and if we land something where we even think there is any danger we should just use this as a carrot for 9.3
<headius[m]1>
the report in question already patched around it by using the preferred URI format, so perhaps that's an argument for 9.3
<enebo[m]>
ah ok well a work around definitely reduces the priority
<headius[m]1>
one failure on windows so far, for jarred gems
<headius[m]1>
probably normalizing the path to C:\uri:classloader or something
<headius[m]1>
I'll let the rest of CI run and then look into it
<headius[m]1>
looks like that might be a failure on linux too
<headius[m]1>
back to the drawing board
subbu|lunch is now known as subbu
ur5us has joined #jruby
nirvdrum has quit [Ping timeout: 265 seconds]
subbu is now known as subbu|afk
Liothen has joined #jruby
subbu|afk is now known as subbu
Liothen has quit [Read error: Connection reset by peer]