On Jan 30, 2008 2:35 PM, Vadim Lebedev <[EMAIL PROTECTED]> wrote: > Why is it better than subversion? > > In my experience ditributed VCS are better suited for decentralsed dev than > centralised VCS
Answer is: - faster - more reliable in case of server fail - easy forking/merging (everybody knows that) but what I would have liked to hear from you is that *in the case of OpenWengo* it will ease the work for vendors (because of easy forking/merging). Here comes the "know you clients, what they need, what they do with the product". Maybe time to make a survey? How do you plan to change from svn to hg? What technical problems do you think you will encounter and solutions? How long do you think it will take? What are the steps? Will it be possible to switch back easily if something goes wrong? What about integration with Trac? (full vs partial vs none) ect... In a more general way I would like to know: How many people will be coding on OpenWengo at MBDSys (full time vs part time)? What do you want to achieve with OW? What are your goal/goals? How do you plan to achieve these goals? What are the steps? time schedule? How will you get money? What are your other projects at MBDSys beside OW? What is your business plan? your strategy ect... *Your global vision of the project will determine my enthusiasm* or lack of (and surely of other contributors). I had one a long time ago... (don't worry if you are blond with nice boobs can help too) > Well Tanguy as you know i was raised in Soviet Union, > so, naturally i'm inclined toward dictature :) LOL My preferred character is Napoleon I, what a great dictator :p Let's develop a game featuring Napoleon I vs Staline huhu > Tanguy what would be your sugestion about subproject/submodule structure > Would you volunteer to be maintainer of any subsytem? Thing is that I don't have that much time (that is why also I want to be sure that you think 3 times before changing things like subversion). I have no idea if I will have time in the future but sure I will follow the project (like a lot of people here I guess). Inactive does not mean that you won't become active in the future. Libraries, libraries, libraries !!! divide to conquer, divide to simplify, divide to ease access to other developers ect... (and very important, work on good APIs between these libraries, APIs are the most important thing with design - UML) 1 library = 1 directory = 1 subproject = 1 submodule If you check trunk/libs directory you have already 28 projects/libraries/modules/whatever you call it i.e libsound, webcam, coipmanager... Work should be done on trying to merge trunk/wifo (osip, eXosip...) and trunk/libs. How is it possible to standardized libraries handling? why not consider wifo and dependencies as libraries like others? On trunk I don't really see any other library that can be created. Some can be eliminated with the come of Qt 4.4 (phonon vs libsound). I started to remove boost from trunk (Aurelien convinced me), I didn't finished nor commited something. I think it can simplify access to new developers since only one big library will be needed: Qt (QtCore and boost are similar). Good luck, may the force be with you -- Tanguy Krotoff <[EMAIL PROTECTED]> +33 6 68 42 70 24
_______________________________________________ Wengophone-devel mailing list [email protected] http://dev.openwengo.com/mailman/listinfo/wengophone-devel
