If you have any questions on our proposition do not hesitate to ask us :) Tim
2013/3/28 Jaussoin Timothée <[email protected]>: > Le 11/02/2013 12:25, Ashley Ward a écrit : > >> On 11/02/2013 10:24, "Sergey Dobrov" <[email protected]> wrote: >> >>> Don't you guys think that the fact that PEP has subscribers from a >>> user's roster does such feature dangerous? >> >> Yes, I agree. You wouldn't want to make this the general case as there are >> serious privacy concerns with automatically sharing this information >> (roster items or general pubsub subscriptions). >> >> With pubsub subscriptions there is also the case that your home server >> doesn't actually know your subscriptions on a remote server. This is >> another one of the reasons that I think storing and sharing this >> information should be application specific. >> >> PEP provides a good place to store this information, but I think the data >> should be stored in the pep nodes in an application specific manner, >> rather than trying to generalise it. >> >> The application itself is then responsible for obtaining user's consent to >> share this information, and deciding what information is stored and >> disseminated. >> >> I don't think anything about this should be mandated in the xmpp spec at >> the moment. The tools are already there to handle this sort of thing. >> >> There may be a case for adding discovery of other user's subscriptions to >> the xep-0060 spec, but you would have to also add methods of controlling >> access to that information too, and it's already complex enough! >> >> -- >> Ash >> >> > Hi ! > > We've made some modifications of the XEP here > http://lite.framapad.org/p/zIMeGxQZfS . > We also start an experimental implementation in Movim :) > > The most important new element of the XEP is the ID generation (line 97). > > edhelas
