Hey, guys, please, please, cool down !!!!
There are things which are important, and things which are not. I have nothing against PKINIT inclusion into the code, but at some point, we also have to consider the community around the project itself. Pushing some new code is *good*. But we must also support the users, and we also have to be around when a new committer/user tries to use it. We have had some concerns about the kerberos code at the beginning of the year, because there was no documentation at all (and the situation is slightly better), and also absolutly no tests. Lacks of Documentaion seems to be a common disease, so I won't blame people for not pushing perfect docos, as we are all faulty. Lacks of tests is damn bad, because not only we have no clue about what is doing the code, but also we are stuck with the existing code base without any way to guarantee that a modification will not break the whole stuff. Lacks of test and lacks of javadoc is even more evil. David, i'm sorry, but I have to back Alex on that, for the reasons I just pushed. Before adding some new code to a stack of 15Klocs nobody but Enrique has a clue about is really not an urgent thing. I would prefer that Enrique spend some time he has on improving the tests on the existing code base. <chair hat on> I would like to say that it's utterly important for the project to have simple rules at this point : - adding new code is _good_ - code without tests, javadoc, documentation is _bad_ - adding some new code to an existing stack of not documented, javadoced and with no tests is not something that will help the project. It's not about doco, it's about tests. With tests, everyone of us can play with the code. Without tests, the code deserves to be sandboxed. Sandbox can be used to push this kind of new code, I see no problem. When it reaches a level of tests which allow any other committer to anter into it, and do some modification without breaking the tests, then it can be added to the project. <chair hat off> In my mind, and for having worked on Kerberos code base in the last 3 months, PKINIT implementation and maybe the whole kerberos project should be moved to the sandbox, except that we have a need for kerberos into trunk. And it's also the reason why In have spent days on kerberos this summer : to avoid a one man show situation. -- Regards, Cordialement, Emmanuel Lécharny www.iktek.com
