On Sunday 11 June 2006 01:00, Sylvain Beucler wrote:
> PHP5 compatibility would be good, as long as PHP4 is also supported,
> and especially is register_globals could be set to off :)

Well I was thinking of temporarily forking the project so the 
fork did not depend on any legacy requirements of the current
Savane. If it gets up and running to a usable state then various
parts, or all of it, could be adopted by Savane proper.

I'd like the opportunity to really pull apart the code and put
it back together with coding practices developed in the last
year or two, like MVC (not strict but somewhat), AJAX, XHTML
and even more CSS. I'm running a test version on PHP5 + MySQL5
(Kubuntu dapper) right now so with E_ALL & ~E_NOTICE & ~E_STRICT
it's running okay. E_ALL produces a LOT of warnings so perhaps
eliminating them may be useful to Savane proper ?

> Mathieu is usually hostile to JavaScript ;) but as long as you do not
> _require_ JavaScript, there is no problem.

It's not really possible to have an AJAX interface that does
not require javascript. Savane, as is, already exists for
anyone who cannot tolerate javascript. I'm the opposite as
I've been so spoilt by AJAX that I can't tolerate web interfaces
that are not highly AJAXed.

 http://markc.renta.net/jquery/jquery_form_test.html

> As far as I'm concerned I'm planning new backend features (hooks
> management)

Such as ?

> but I may help with register_globals clean-up.

There are 731 instances of GLOBALS[] though :-)

I think the first major thing I'll try is to put all the global
config vars into an array inside a simple singleton class.

I'd be interested in porting all the backend scripts to PHP
too, but I have no idea what that involves yet.

I really only want a PHP5 based trac clone so I may be biting
off far more than I am bargaining for. I actually tried to do
something similar for SF 2.5 when it was released, but then
Savannah came out so I thought I'd wait and see what happened
with it... so here I am again, way too many years later.

PHP5 picked up one bug so far... line 33 in include/graphs.php
was "break;" instead of "return;".

--markc

_______________________________________________
Savane-help mailing list
[email protected]
https://mail.gna.org/listinfo/savane-help

Reply via email to