Re: New Package Proposal: Enterprise A&A

> Hi, > > ok the way I see it both A&A and LiveUser aim to provide the same goals > and are somewhere in alpha and beta stage. > > For the most part the approach is the same (container/provider approach > to allow different backends), but there are some structural differences. > The backends that are supported atm are not the same however. I don't > know if there is a language barrier or a communication barrier, because > Tony seems to dispute this fact a lot. As far as I am concerned there does not seem to be any language/comunication barrier, we've been having a very interesting discussion. > When I said that I think it is wacked to have two single sign on classes > in PEAR I say this because I think it would make sense to through as > much weight as possible behind one single sign on system. Even though it > is possible to make the to work with each other it is less than ideal > and does not make the choice easier for the enterprise folks :-) It makes sense. > To conclude: > I would suggest deciding on one of them and merging whatever additional > features or useful concepts the other has into it. Again, it makes a whole lot of sense. Combining efforts seems the more fruitful approach. > So my question to Tony is: Is this a sensible course of action for you? > Will you leave the code out in the open even if A&A is not chosen? > > For LiveUser: The code is already in the open. The question is if the > current developers are willing to switch if A&A is chosen. I for one > haven't started using LiveUser in production. If I understand Markuses > situation correctly he is the original author of LiveUser but uses an > older version in production that is incompatible with the PEAR version > anyways. Arnaud seems to be the person closest to using PEAR::LiveUser > in production. Ok, not to put pressure, I resumed work on LiveUser. By that I mean I actually began using it in my current development, so there should be bugfixes (or not if everything works ;) in the next couple of weeks. The fact is that I did not start using it in a production environment ... yet. If everybody agrees on keeping one package it does not really matter, it is more an API question for people already using it. Are there any liveuser users out there ? Anybody who has ideas on authentication/permission should speak its mind. I mean, I think everybody benefits from everybody ideas, right ;) > I hope to find some time to look at the A&A source tonight and will then > voice my vote. I just had a look at the code and I must say it looks very clean. My thoughts on the subject are as follows: i cannot spend a month without a user management system. So I will use liveuser. So I think A&A can be put into pear because merging both would require a lot of time which I am not sure I have (as lukas and markus i'd dare say) at the moment. We could then start from ground-up, build a common API and take the best of the two packages. Wrapper could be provided as well to make things easy. Both packages are somewhat in the same stage, the design is different but the aim is the same. Too bad we could not do that two months ago, efforts could have been merged more easily then, oh well, too bad ;) So I'd say, go for it +1 from me. Arnaud.