[Foundation-l] Instant Commons

Sun Apr 2 03:04:41 UTC 2006

Robert Scott Horning wrote:

>Anthere wrote:
>
>  
>
>>ChitChat with David over whether there is a meta community ... or not... 
>>is nice, but let's get back to real business for a while :-)
>>
>>There is a proposal on meta : InstantCommons
>>http://meta.wikimedia.org/wiki/InstantCommons
>>
>>I would be happy if some of you could have a look at it (if not at the 
>>technical details, at least to the general concept - it will not take 
>>you much time) and give a feedback about it (here or on meta or on irc).
>>
>>I am not looking for a comment by those who proposed the project... but 
>>by the others ;-)
>>
>>Thanks for your feedback.
>>
>>Anthere
>>Special Project Committee
>>http://meta.wikimedia.org/wiki/Special_projects_committee
>> 
>>
>>    
>>
>I think Instant Commons is an incredible concept.  I don't see how it 
>will directly help individual Wikimedia projects, but it is something 
>that is going to be beneficial in terms of having goodwill to other 
>MediaWiki website administrators and perhaps other individuals who are 
>sucking bandwidth out of the Wikimedia Commons image archive.
>
>Are there any statistics available to show how much bandwidth is going 
>to serving images to non-Wikimedia projects who are using HTML markup 
>links to commons images?  I'm especially talking about site mirrors (who 
>may not mirror the images), but other projects as well.
>
>  
>

I am interested in this reasoning:
"

    * It does not respect the license terms of the image, and does not
      allow for other metadata to be reliably transported
    * It does not give credit to Wikimedia
    * It consumes Wikimedia bandwidth on every pageview (unless the
      image has been cached on the client side or through a proxy)"

Has any consideration been given to allowing a pass through type of 
feature where the Wiki Media Commons becomes more of a registry and 
published online backup rather than a centralized master archive?

For example:  If NASA could publish some of the famous Apollo 
photographs or Mars data documenting water erosion in partnership with 
another entity under a GPL (this is already done by for profit entities 
who attempt to enforce copyright so there should be no problems using 
GPL on public domain NASA data) by submitting the photographs, detailed 
information and the master URLs where they keep the masters.

Under this scenario, perhaps the NASA site could be the primary URL with 
Wikimedia Commons used as a secondary accessed when the original is not 
available.    Rather than becoming the primary source for the data, 
Wikimedia Commons serves more as a catalogue of GPL'ed data available 
elsewhere on the web.

Rather than acting as a centralized single massive server Wiki Commons 
might save substantial bandwidth by only providing data when the primary 
source is not available when an update to local caches is initiated.

This might also encourage other smaller sites to collaborate with Wiki 
Commons as they would receive exposure and traffic in exchange for 
publishing their content under the GPL via the Wiki Commons registry.   
Wikimedia Foundation thus becomes a welcome large partner or umbrella 
resource improving wiki technology and available FDL'ed content rather 
than the 800 pound competitor avoiding exposure of Achilles tendons.

The technical developers responsible for the Wikimedia Commons might 
wish to also look this over:
http://dijjer.org/

It is a caching technology being designed for web browsers to cooperate 
in reducing the load on central resources using a peer to peer transport 
similar in some ways to bittorrent.  I am trying it out at the moment 
but I am uncertain whether it will catch on widely.  Bit Torrent 
provides the luxury of supporting transmission of specific files the 
individual users chooses.  This dijjer.org project appears to place all 
control in the hands of web publishers and thus I think it becomes 
vulnerable to unaccountable freeloaders.  If this is true, it will 
probably not prosper or be an influential technology in the long run.

regards,
lazyquasar