Re: Considering package proposal: Cgiapp

On Apr 11, 2005 8:09 AM, Tomas V.V.Cox <cox@idecnet.com> wrote: > Lukas Smith wrote: > > Matthew Weier O'Phinney wrote: > > > > Seriously though if we did allow frameworks then only those that we feel > > comfortable endorsing. If we have multiple frameworks then they simply > > would need to fit the "non-redundant" definition and so be it. > > A framework it self it's a reduced PEAR with form classes, templates, > i18n, .... Let's say there is nothing more redundant for us than > accepting a framework :) This is actually why I though Cgiapp might fit in -- it doesn't do forms or i18n -- and templating systems are pluggable (I use Smarty by default, but you can override the class to use whichever template solution you want). It's not a traditional framework. It handles determining what run mode of the application has been requested, and then performs the method for that run mode. Additionally, there are several points in program execution that may be customized -- pre-setup, pre-run mode execution, post run mode execution, and teardown; Cgiapp simply defines the steps, and lets the developer write the code for them. In a way, Cgiapp is simply some glue code to simplify developing applications. However, it's a framework in that it gives a specific modus operandi for developing applications: class == application, run mode == method. It's along the lines of an MVC controller, but can blur that definition significantly, based on how the developer uses it. -- Matthew Weier O'Phinney mweierophinney@gmail.com http://weierophinney.net/matthew/