From: P Z. <zol...@gm...> - 2009-02-09 14:31:07
|
On Sun, 08 Feb 2009 22:11:41 +0100, Julian Bäume <ju...@sv...> wrote: >> The idea of integrating it with kdevelop to a greater extent is >> interesting and has a great deal of potential... (wonder if you could >> flowpart a C program or something. =P) > It won't integrate directly into kdevelop, just use some of their > infrastructure. But basically your right, of course. Does this mean that besides kdelibs, kdevelop will also be a dependency of ktechlab? That won't make non-kde users happy. >> But it is good dicipline to keep to the smallest number of incremental >> changes at a time, even if the intermediate results aren't optimal. > True, but hard to manage if each change you need to do is a quite large > step > in the project. Base on what I've seen in the code, I prefer a rewrite. (an the KDE guy also did the same :)) ). But first, we need a _design_, instead of starting coding directly. >> Right now the tree is frozen for release. I'm not sure what the best >> practices are with SVN, but I hope the SVN administrator will create a >> 0.4 fork as soon as possible so that the multitude of efforts at the 4.x >> port can be integrated and testing can begin. > There already is a branch in SVN containing the next stable release. I > manually backported all bug-fixes from trunk. If this branch is > considered > stable, we can tag it as released and create packages from that copy. > Trunk > could be opened for development again. I don't see anything against > doing so. > If there is the need to create bug-fixes for the new stable release, we > could > easily provide fixes and release another bug-fix release in the 0.3.x > series. > Since the release-candidate version is branched separately, we could mess up the trunk, right? :D My opinion is that we should start a new "trunk" for the kde4 port from zero, and when it reaches the level of functionalty of the 0.3 versions, that should become the new trunk version. |