Slice the pie...

[Ingo Malchow]

2011/4/23 Lukas <1lukas1 at gmail.com>:
> Hi,
>
> You’re probably guessing right, this letter has nothing to do with food :)
>
> As far as I heard, the overhaul of kde www ecosystem is being planned. One
> thing I know for sure - its impossible to create a single CMS/tool fitting
> needs of
> any/every kde domain. And creating tools for each individual case is boring,
> as lots of code is almost the same.
>
> Instead, it would make sense to split those repeating elements into separate
> mini projects and “assemble” them back to web page at the browser level with
> Java Script.  To make it clear, think of comment forms, recent posts, top
> application lists, recommend buttons, related community blogs, whatever. The
> similar idea is already implemented by e.g. facebook as the social plugins.

Basically you are already nailing what we plan. Having a modular
webapp where we can plug new additions in. Not sure about the JS part
though, but as nothing really is started yet, this needs to be
evaluated.

>
> The key benefits:
>
> * Consistency
> Having same look and feel everywhere

Personally i prefer not the same look, but at least some basics in
common (corporate identity, not design).

> * Getting social
> Having to register to comment on each site is a known problem. Not for long
> :)

Identity is not a single sign in system, but a single sign on. So you
still need to log in, even though there is only one place to register.
Important point!

> * Easy to integrate
> As long as we have write access to sites theme, any element is inegrateble
> * Easier cross project maintenance.
> Since adding features done by adding a few html lines, even radical element
> changes wont break things up.
> * Easy to replace with better alternatives.
> In case if projects like Diaspora become reliable alternative, KDE-www wide
> integration could be made smoothly.
> * Faster page loads
> As long as JS is loaded at the bottom of the page, the key content will be
> loaded faster, while supporting elements loads in the background.
> * Less dependency on single programing language/platform
> most elements could be developed as separate projects, best one with most
> contributors could be chosen on individual case (no more PHP ws Ruby ws
> Python... flames).
>
> Since http://lists.kde.org/?l=kde-promo&m=130332826712417&w=2 comments
> element is primary target to test the idea. Add identity integration,
> RSS/email subscription, BB code, syntax highlighting and even bugs.kde.org
> could use make use of it :)
>
> Would this approach be feasible?
>
> Cheers,
>
> Lukas
>

Overall, yes, sounds good. Details would need to be evaluated. Would
you like to attend our webworld sprint? Some seats got empty. Might be
a good opportunity to meet and make something happen ;)

Cheerio,
Ingo