[Foundation-l] Re: Launch of the Wikimedia Commons

[Ashar Voultoiz]

Angela wrote:
> I am pleased to announce that a database has now been set up for the
> Wikimedia Commons.
> 
> The site is temporarily located at http://commons.wikimedia.org/ until
> the project gets its own domain.
> 
> This provides a space for those interested to begin forming the
> guidelines for the site, and to discuss how the software will need to
> adapt to make the Commons as useful as possible.
> 
> The previous discussions are now linked to from
> http://commons.wikimedia.org/wiki/Commons:Project_plan.
> 
> Batches of files may be uploaded as compressed zip files. Some time in
> the next few days, a script will be written to decompress these zip
> files on the server side. Please make sure you use appropriate file
> names inside the archives.
> 
> Thanks to Tim Starling for creating the database, and to all those who
> have contributed to the discussion so far.
> 
> Angela.

Hello,

In my mind, MediaWiki is not a software to manage a gallery of pictures 
or media.

As I understand the announcment, the community already choose MediaWiki 
and will just ask developpers to code new features. Basicly developpers 
will have to write from scratch a gallery system whereas there is 
already such software under GPL (example: http://pnavy.com/dcgallery/ 
with dublin data core support).

There is so much features that are needed for a media gallery, that soon 
the dev team will be overhelmed. What will happen when the community 
will want dublin core metadata, EXIF automatic parsing, e-commerce, 
XML-RPC, multiple fields search, ogg tags parsing, videos preview ... We 
don't even have a license system : we rely on using templates instead 
wich is clearly not a solution in term of search capabilities.

A wiki is not a database, it's a tool to create collaborative text. A 
media gallery is a database. Guess what ? A wiki system is not meant to 
act as a media gallery.

As for the batch zip uncompression, that just lead to easy vandalism 
(uploading a 3MB zip can easily generate 1GB of data).


In conclusion, the "commons" idea have been around for sometime now, we 
should not start things in a hurry and hope to fix it later. It will 
just give us a LOT more works. Instead I think we should carefully plan 
the project needs, do and don't, find an already existent solution and 
then either use it or enhance it (that's what GPL softwares are all about).

cheers,

-- 
Ashar Voultoiz