We really need to start to define what is and what is not Version 5 – plus, I think, we should put a few simple rules in place. In particular – separated by a massive ocean and three-quarters of the North American continent as we are – we need to write proposals up in more detail. I am not suggesting full-blown specifications but enough detail so that we do not trip over each other when developing independantly and in different but obviously connected areas. We should also document all functions as they are created.
I also strongly recommend that in the first instance we throw out ALL 'new' features. We should concentrate on the structure only. I would feel far happier releasing V5 with all the structural changes and no new features than a half baked version because we got sidetracked into adding new stuff.
Additionally – we need to keep focused on performance. If we can't improve on response times at every stage we should question the approach and not be frightened to change it. We know it's going to be a long haul.
I have listed below what I believe to be the structural changes we should focus on. All up for debate of course and – we are getting close to having to auction who does what.
TEMPLATING
Probably the biggest discussion area and the one we really do need to plan out in some detail. There are a multitude of approaches we could take for this and I reckon we need to know exactly what path we are going to follow before writing any code.
ICONS AS SPRITES
This should result in a major performance boost which is why I rate it highly.
PLUGINS
I believe the suggestion here is that we extend the WP add_filter/add_action into our code as mch as possible which should make allowing for third party plugins a breeze.
DATABASE
The whole approach to our queries, error trapping etc., needs a major overhaul and we still need to decide on the minimum MySQL version we intend to support. This decision may have a major impact on performance.
CACHEING
The areas where we have dabbled with cacheing – avatars, signatures, quick links, queries – have all proven to have been useful and successful. There is much more we could do including using the database to hive off repeatable chunks of code. Again – these items could prove tp be a major performance boost.
HIGHSLIDE & CHECKBOX/RADIO CONTROLS
if we are going to change these then this os the time.
EDITORS
We will probably need an early decision on whether we are going to change editing tools – the dream tool, of course, being one that handles rich text, html and bbCode. the filtering could be the problem as like it or loathe it, TinyMCE does a fairly good job at filtering and formatting.
INSTALL & UPGRADE
I think the model we have now is actually pretty sound. problems come through script conflicts and, in some cases, careless upgrade code. We need to review at the very least.
Well – this is my list – typos and all. have I left anything off?
