> From Steven Hazel <sah at thalassocracy.org> >It seems to me that date-based redirects are becoming a very >fundamental part of how many freenet client apps work. Given this, >can anyone think of a good reason not to replace date-based redirects >with URIs of the following forms: > >for a KSK: >freenet:DBK at baseline,increment/foo > >for an SSK: >freenet:DBS at baseline,increment,public key/foo > >assuming that "baseline" and "increment" are base16 or base64 (I don't >care) seconds since the epoch, and the final keys look like: > >freenet:KSK at mostrecent-foo > >and: > >freenet:SSK at public key/mostrecent-foo > >respectively, where "mostrecent" is also base16 or base64 (respective >to your previous choice) seconds since the epoch? > >The advantage to these URIs would be that retrieving the data they >point to would require only one request, whereas retrieving the data a >date-based redirect points to requires two. This would make it >possible to write a date-based redirect to a CHK which requires only >two requests instead of three. And the keys aren't any uglier than >ordinary SSKs, so as far as I can tell, this is pure winnitude.
I still think we should use Gregorian/GMT/Base10, since these are numbers that are meaningful to even non-technical humans. The other problem I see with this (besides the "if it ain't broke, don't fix it" argument) is that it puts the responsiblity of configuring the DBR correctly on the person linking to the document, instead of the person uploading it, which seems wrong to me. -- Benjamin Coates _______________________________________________ Devl mailing list Devl at freenetproject.org http://lists.freenetproject.org/mailman/listinfo/devl
