On Thu, 28 Dec 2000, Benjamin Coates wrote:
> >From Travis Bemann <bemann@bemann>
> >There has been lots of discussion on using stuff like mapfiles/MSKs to
> >handle pseudoupdating. However, pseudoupdating is just a temporary
> >measure that will be obsoleted by real updating. So why don't we not
> >waste our time working on a temporary measure and instead actually try
> >testing pass-through and selective-broadcast updates with Serapis and
> >then actually implement the better of the two updating schemes?! If
> >we did this right now we might be able to get real updating in Freenet
> >0.4, but if we work on the temporary measure of mapfiles, real
> >updating will probably not be in Freenet 0.4 and will wait to Freenet
> >0.5.
> >
> >--
> >Travis Bemann
>
> Unless there are serious performance issues with the date-based redirect and
> MSK combination, I do not see a reason to implement real* updating at all.
>
> * I'm not entirely clear what the difference between "real" and "fake"
> updating is, however.
Real updating allows you to replace the contents of an existing SVK
derived key with new content, provided that your signature checks out. It
will work. The problem is how to get the update 'messages' to the
locations in Freenet where the content exists.
Fake updating is creating a new key entirely but redirecting the reader to
the new key each time he/she wishes to view it.
Scott
_______________________________________________
Freenet-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/mailman/listinfo/freenet-dev
