Hi list,
Could everybody please have a look at Bug #564793, and give me your input on such an authentication system as I have proposed.
For working on PMA,
Here is my plan of attack on things: early 2.3.1 release candiates will have: Parser analyzer system functions in place and basic use DB-config helper functions in place and used Initial code for bug #564793 (either the session cookie OR the symmetric encryption, not both)
late 2.3.1 release candiates / early 2.3.2 release candiates: full DB-Config system (actual database storage) Parser analyzer UI (better responses than MySQL server responses!) Final code for bug #564793
I have a strong reason for this spread of feature addition. The key part of it being seperate of what is happening to each sub-area of the code, is that is becomes a lot easier to trace which of the changes caused problems, if any. Especially as some of these features are really two parts intertwined.
With the second set, I'm more inclined to have them ready for 2.3.2 release candiates, rather than 2.3.1, to increase the testing time, and allow me some more time, as I have university courses starting up next week as well as a part-time job (working on open source! http://mvs.sf.net/)
Additionally, this gives the translators more time to work on things, as there will be a LOT of strings coming in with the Parser Analyzer UI.
Has anybody given thought to the release schedule? Exactly one year ago today, we released PMA 2.2.0. Our releases in the 2.2 series were about 35 days apart. How busy is everybody else expecting to be over the next few months? I think Alex is starting university, and I'm definetly back at university, so I don't think we can quite keep up with some of the frequent release schedule. I'd tentativly like to suggest that we aim for 45 days between releases. That gives us a month for features, and two weeks for a pair of release candiates.