[Dextrose] [Sugar-devel] [DESIGN] Extending activity microformat spec to include optional olpc_activity_size

[Bernie Innocenti]

On Mon, 2010-11-22 at 10:51 -0500, C. Scott Ananian wrote:
> I avoided doing this in the original specification because I meant the
> microformat to be human-writable, and I didn't expect humans to be
> able to reliably update these fields correctly.

The field is optional, though.


> Besides, the size can be ascertained with an HTTP HEAD request, which
> takes very little bandwidth.

Yes, but it introduces latency proportional to the number of activities.
Lots of latency from schools in deployments.

>   Checksum should also be HTTP ETag. Again, less than 100 bytes.
> I believe that my original implementation
> did both of these optimizations, although I wouldn't be surprised if
> some servers didn't properly support HTTP HEAD and/or ETag.  But I
> believe it is the server which should be fixed in that case.

Hmm... I'm not sure this approach could be made to work beyond simple
caching proxies. For example, ASLO delegates downloads to a CDN based on
HTTP redirects to mirrors (some of which are nearby deployments).

With a well-defined digest on content we can reliably identify
alterations along the distribution chain. If the update page were on
https with a certificate signed by the deployment, we could ensure
secure distribution of activities.

-- 
   // Bernie Innocenti - http://codewiz.org/
 \X/  Sugar Labs       - http://sugarlabs.org/