Reloader Defaults All Routes to 404

[Wardrop]

When I make a change that requires a full app reload, all of my routes stop working and return 404's.

Here's a log of the reload operation: https://gist.github.com/1797102

Here's what my route looks like (app/controllers/reservation.rb): https://gist.github.com/1797100

Does anyone know what's happening. The log looks like it's done everything correctly?

This issue comes after I just worked around another reload issue. It seems padrino doesn't like it when you use #require_relative in files that are auto-reloaded, especially when the #require_relative references a file that's also automatically reloaded. Anyway, this issue is less blocking than the one I face now.

[DAddYE]

Thanks man, I will try to reproduce the problem with ruby 1.9

[Wardrop]

Yep. I should have mentioned I'm running Padrino 0.10.5.

[robertjpayne]

Any update on this? It's making developing with Padrino a bit harder than I'd normally like.

[nesquena]

@robertjpayne You may want to consider disable :reload and checking out https://github.com/joshbuddy/ak47 until we resolve this reload issue. ak47 bundle exec padrino start

[Wardrop]

I'm using re-run for the time being. The pattern I use for tracking file changes is "*/.rb". That way it only reload if I change an .rb file anywhere in my application directory structure. ERB files are exempt as they don't need reloading. It would be nice if Padrino could launch faster, that way I could continue using shotgun as I do with Sinatra. My application in Sinatra took only about a 0.25 seconds to reload using shotgun. Padrino takes more like 2+ seconds.

[robertjpayne]

I did what @nesquena recommended and turned off reloading. And am using Shotgun to completely respawn the entire process on every request.

The app I have is extremely tiny and running ruby 1.9.3 on my local machine which is very fast so I can't even notice the delay.

[DAddYE]

Guys, in next version I will add a re-spawn builtin which is (almost on unix and mri) machines rock solid.

Btw, I worked a bit on this issue and I made few changes, I still need to check why require_relative make crazy our reloader, btw If any had a chance to try their code with latest padrino edge, be free to update this topic with yours logs. Remember to enable devel mode.

Thanks!

[nesquena]

Thanks @DAddYE the spawn built in reloader sounds really useful, I would probably use that in most cases over the current one and we can leave in this one for windows.

[DAddYE]

Other than windows other rubies implementations can have some problems

[nesquena]

@DAddYE Any update on this? or a reloader fix? I have experienced this as well.

[ujifgc]

Failed to reproduce this. Used a bunch of recursive require_relative calls of helpers, apps and system features. Please reopen and provide a minimal test case if the issue persists on padrino master.