From: Erich Schubert <newsgroups@...>
Date: 2001-04-18T09:10:08+09:00
Subject: [ruby-talk:13784] Web Applications

Has someone made a framework for web applications?
I'm looking for something like HTML::Mason, which is a just great Perl
Module; i believe this should be easier to implement with Ruby due to the
OOP.

What i need consists of two parts; one party is mostly done by eRuby, the
one of embedding ruby code into regular files.

The second part is what makes the big difference from HTML::Mason:
you do not work on single pages. it's more like "deriving" pages from
others.

That's what i actually want to do, and i'd love it to have some
eruby-like syntax for this as well as some caching structure.
These are things HTML::Mason does provide me with. (at some very-low-oop
level)

Basically my idea is as following:

You have some generic Request Object which outputs HTML (or XML, if you
prefer XML... or even Text) Header, Footer and has as "body" the message
"file not found".

From this object i derive my main pages by modifing some variables (like
title, description, meta-tags etc.) and replacing the body.

But i might also derive a subdir-"template" with slightly modified header
and footer; from this derive sites in this directory.

What do you think of this Website-Concept?

Is there some Ruby Project like this?
HTML::Mason is like it, but it is very Perl-dependent as it tries to
maximize mod_perl caching etc.
for example HTML::Mason translate all your "pages" into Perl source,
compiles this perl source into perl binaries and saves them on the disk.
so your mod_perl doesn't even have to recompile it, it can just take this
perl binary and translate it to machine code. HTML::Mason is very fast...

I'd love to do something like this in ruby...

Greetings,
Erich