Re: [cgiapp] Possible patches for ::Server (was: Re: ::Server vs ::Dispatch::Server: Worth having both?)
* Mark Stosberg <[EMAIL PROTECTED]> [2008-09-17T20:57:37] > > ::Dispatch::Server prefers static files first, ::Server prefers dispatching > > first > > I think it is a bug in ::Server that it prefers dispatching server. The > recommended use of ::Dispatch, and also for Drupal, is to use a rewrite > rule that will serve static files directly if they match, otherwise the > request is passed on for possible dispatching. As Michael Peters pointed > out, this design should have better performance as well. > > So Ricardo, would you accept a patch to change this ordering? Almost certainly. I'll have to check with the other big user/maintainer. > That will print out "/first/second" and then "/first", but never "/". I > think this is a bug in ::Server. Richardo, would you accept a patch to > make "/" dispatchable? ( "" Would be treated the same ). Only if you agree to never put an 'h' in my name again. -- rjbs # CGI::Application community mailing list #### ## To unsubscribe, or change your message delivery options, ## ## visit: http://www.erlbaum.net/mailman/listinfo/cgiapp## #### ## Web archive: http://www.erlbaum.net/pipermail/cgiapp/ ## ## Wiki: http://cgiapp.erlbaum.net/ ## ####
Re: [cgiapp] Possible patches for ::Server (was: Re: ::Server vs ::Dispatch::Server: Worth having both?)
Mark Stosberg writes:
>
> > Patches welcome.
>
> I'd like to provide feedback on various proposed changes to
> CGI::Application::Server
>
> > Need a way to specify parameters to dispatch()
>
> I disagree. The recommended way to do this to define a 'dispatch_args'
> routine in your CGI::Application::Dispatch subclass. I think George
> discovered seperately that the "dispatch_args()" method in
> ::Dispatch::Server was too magical for it's own good.
> [...]
Well, in the defense of ::Dispatch::Server's dispatch_args() method,
it was tripping over a sharp edge in ::Dispatch::dispatch, where the
specialized handling of args_to_new->{TMPL_PATH} blindly concatenates
the existing list of directories with the existing list without
checking for duplicates.
I'm not sure what the Right Thing To Do there might actually be (I'm
not a TMPL_PATH user), but given that the rest of the merging replaces
existing values with newer values the behaviour is Surprising.
Passing arguments to ::Dispatch::dispatch gets pretty prominent
billing in the ::Dispatch documentation, so it's not surprising that
folks use it.
g.
# CGI::Application community mailing list
####
## To unsubscribe, or change your message delivery options, ##
## visit: http://www.erlbaum.net/mailman/listinfo/cgiapp##
####
## Web archive: http://www.erlbaum.net/pipermail/cgiapp/ ##
## Wiki: http://cgiapp.erlbaum.net/ ##
####
