Long Term / KDE.org Design Update

Wed Jul 8 08:33:57 UTC 2015

> Hello Everyone;
>
> My name is Ken Vermette, I'm a member of the VDG and by trade a web
> designer/developer. We're interested in modernising the kde.org websites
> and getting the ball rolling on better unifying the sites as a whole. I'm
> personally willing (and expecting) to invest the time and programming
> effort required to do this. I know it's obvious, but we'd also like to work
> closely with the web team.
>
> It's a huge undertaking and we know it will easily take years to finish all
> the sites; we would like to know we aren't stepping on toes, duplicating
> existing efforts, etc. Before any of that we need the green-light from
> here, and would also like to invite web-team participation in building a
> maintainable and more universal kde.org network.
>
> Thank you, and great work;
>   - Ken Vermette
Hi all,

I think a redesign of the kde.org sites is a great idea. Since responsive design
is now a "must have", I suggest building the site upon one of the responsive
frameworks (for example Bootstrap) rather than building a own responsive framework.
Also the new design should be more consistent. There is no need for inline CSS
as it makes overriding hard, the same thing also holds for internal <style> tags.

Although design is important, there is a component that desperately needs a rewrite:
the backend.

I have been digging through code in an attempt to customize one of the kde pages and
it was not a pleasant sight. The main problem of the current design is that a lot of
things are hardcoded. Basically, if you include the header.inc file, you are stuck with
all the things that other sites have. This might be good for the normal websites, but as
  soon as you have to customize a thing (for example add/remove a css file or a div),
you are out of luck. There is no easy way to override things, but with minor tweaks
  this could be achievable, without breaking things.

My proposition is to split the hardcoded content into modules loadable by a loading function:

// pseudo code
function load_module($type, $file, $allow_override = false) {
if $allow_override == true
   // look into the current script directory for $file and include it if found, return
else
   // look into $type for the $file, include it if found, return
}

example header.inc:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
     "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"  lang="<?php print $site_locale; ?>" xml:lang="<?php print $site_locale; ?>">
<head>
   <title><?php print $title; ?></title>

   load_module($document_root, 'meta.inc', false); // meta is loaded but cannot be overridden
   load_module($template_path, 'css.inc', true); css can be overridden by a local css.inc file
   load_module($document_root, 'js.inc', true); js can be overridden by a local js.inc file

  ...
</head>
<body>
   load_module($document_root, 'breadcrumbs.inc', true); breadcrumbs can be overridden by a local breadcrumbs.inc file
   load_module($document_root, 'plasmaMenu.inc', true); menu can be overridden by a local plasmaMenu.inc file

   etc.

This is (somewhat) the same mechanism that Wordpress child themes use and it's really easy to use.

The second problem with the backend is that it's outdated. If you turn on error reporting and
warnings on a server with recent PHP version, the number of messages will blow your mind :). If
there is redesigning being done, why not update the backend to current standards? The current
backend does a good job, but sooner or later it will have to be modernized.

Anyway, these are some things I have noticed by browsing the code. I would be happy to help
improving the frontend/backend, so I'm wondering what is the current status of the redesign.

Cheers,
Andrej