2017-05-10 22:08 GMT+02:00 Sam Whited <[email protected]>: > All this being said, I could sway either way at this point, and as an > actual client developer you probably have a better perspective on what > makes the most sense for the client ecosystem, so I defer to your > knowledge on what's best here. >
I actually was trying to make this easier for the server developers. I was assuming that managing multiple tokens in a database is more complicated than doing some quick calculations on one. One more argument for being able to generate new tokens based on a secret in a PEP node is, that once I have that secret cached (because it was sent to me at the beginning of the session) I can create tokens at any time without needing a server connection right then. I can very easily come up with a lot of scenarios in which I want to show a QR code to someone when I don't have a connection to the server (session is hibernating) or only a very lagging one. Or in other words I would like to avoid a spinning wheel 'generating QR code / requesting token' before being able to invite someone. Anyway I think my arguments are made and now I guess it's up to Georg to decide which way he likes to go with his XEP. My main concern with current PARS is only that I strongly believe that this should be a server side XEP and I think this thread has shown that at least on this point there is a strong consensus. How exactly this becomes a server-side XEP doesn't actually matter to me personally. I'd implement both. cheers Daniel
_______________________________________________ Standards mailing list Info: https://mail.jabber.org/mailman/listinfo/standards Unsubscribe: [email protected] _______________________________________________
