The problem with all of this is that it is not (from my perspective as an end user) expected behavior.
Arguments about Camino doing the "standard" thing or not aside, losing my Camino passwords when I reset Safari but not losing my Safari passwords when I reset Camino is a problem. What are we to do? Wait for Apple to fix their bug? Is there a way we can have an option for Camino to do the non-standard thing and store its passwords differently? Being stuck on the right side of the argument but still being hit by defective behavior is something that I think should be dealt with in some way. ~Aric On 10-Mar-08, at 11:43 AM, Stuart Morgan wrote: > On Mar 10, 2008, at 7:31 AM, Noemi wrote: > >> It's clearly possible to create keychain entries that Safari can't >> read, as evidenced by all the other apps that don't lose their >> passwords when Safari is reset. > > Sure; Safari uses web form and HTTP auth passwords, so those are the > types it reads. Most applications aren't browsers, and thus don't save > passwords of those types. > >> Why does Camino create entries that Safari can read? > > Because we are storing the same information, for exactly the same > purpose, that Safari would. It's not like we reverse-engineered > Safari's keychain storage format and bent over backward to rig up > entries that Safari can read; Safari can read them because we updated > Camino's keychain code to correctly use the standard keychain types > (which we didn't use before only because Camino's original keychain > implementation predates the creation of standard keychain type for > these sorts of passwords). > > Using the keychain APIs correctly is a feature, and the > interoperability with other browsers that use the keychain correctly > which comes along with that is also a feature. The only arguable point > is whether or not Safari should view the reset feature as applying > only to things it created, or anything applicable to Safari, and that > is something that we have no power to decide or change. If you > disagree with Safari's current behavior, you would need to file a bug > with Apple--although they may view this behavior as correct, just as > Safari uses the system cookie storage system, and reseting cookies > from within Safari's interface will affect just about any network- > based applications that haven't explicitly opted out of using the > system cookie store. > > -Stuart > _______________________________________________ > Camino mailing list > Camino@mozdev.org > https://www.mozdev.org/mailman/listinfo/camino _______________________________________________ Camino mailing list Camino@mozdev.org https://www.mozdev.org/mailman/listinfo/camino