#techempower-fwbm

/

      • codygman- joined the channel
      • jberger
        TE-msmith_ / hamiltont: I got my mojolicious/meganuke port working!
      • hamiltont
        Nice!
      • I'm looking forward to the branch being merged
      • jberger
        found an actual bug in our preforking server doing so
      • hamiltont
        ha ha no way
      • that's insanity
      • jberger
        yeah, kinda crazy one, when using a perl that isn't understood by the environment (not /usr/bin/env perl) then the preforking server would start the main process in the specified interpreter, but the fork/exec-ed workers would start using the env perl :/
      • not usually a problem of course, but the way FWBM is laid out and the way I use Carton etc etc, that is why I couldn't get the JSON parser to work, it wasn't installed for the workers
      • hamiltont
        huh, makes sense
      • oddball issues like that are good to have developers find
      • jberger
        drove me crazy for a couple days
      • FWBM-bot
        [FrameworkBenchmarks] jberger opened pull request #1244: ported Mojolicious setup to bash, added app.conf, added stop.sh for symmetry (meganuke...meganuke_mojo) http://git.io/DtLoUQ
      • jberger
      • oops bot beat me to it
      • hamiltont
        Nice. FWIW I'm avoiding merging stuff while TE-msmith_ is still working on hisbranch
      • jberger
        I did add a stop.sh
      • hamiltont
        don't want to create a moving target for him
      • jberger
        hamiltont: mine is against his branch
      • hamiltont
        Ah ok. Is stop.sh ever actually used?
      • jberger
        not that I know of
      • hamiltont
        Ok
      • jberger
        basically I wanted to show how one WOULD stop it if it wasn't being meganuked (which for mojolicious/hypnotoad isn't necessary)
      • hamiltont
        Yup, I think that's a pretty good form
      • hamiltont is now known as hamiltont_away
      • jberger
        anyway, I intend to port the other perl frameworks to bash setup soon
      • less for msmith to have to do
      • they shouldn't be as hard
      • I was doing that json config trick which didn't port directly very well
      • codygman joined the channel