Bug#95870: ardour & quasimodo debian packages
>I think getting ardour into Sid could benefit the project in the sense that it
>opens the code to a larger audience and speed up tracking bugs (after all,
>that's all about what Open Source is).
at this time, i sincerely doubt that. the CVS log for the last 2 days
records 2000+ lines of changed code. debian packaging is cool, but its
not targetted at the actual development of a rapidly changing
system. thats what CVS is for.
in addition, its not easy to define what "the package" is here. the
program "ardour" is really just a symlink to another program called
"aes" that causes aes to dynamically load a shared object. in
addition, there are 6 of my own libraries that have to be installed,
but are not specific to ardour or aes (i have several other
applications that use them). these libraries are also undergoing
development, some very rapid, some are now fairly stable.
>Your program would stay in Sid until it
>gets stable and thereby benefit the contribution of a large community of
>developpers.
I think it would be appropriate to put it into Sid at the point where
I start to release tarballs. Doing so before then just opens me to up
to Debian users saying "I found this problem with the debian package
X.Y.Z", when I've already changed 5000 lines of code since then. Its
just not a situation I want to be in (or put them in).
I don't know if you've read the ardour-dev archives; there are people
(reasonably skilled programmers) who tried to get ardour to run on
their machine for 2 days and never succeeded. the precise version of
libxml, whether the package of Gtk-- they use is broken or not
(several of them are), the version of Guile their system has
installed, which compiler version, and so on and so forth. These are
all important things to fix, but its definitely not my focus at this
time. Getting the full functionality of the program is; since I know
that I can compile it, and I know of several others who can, thats
enough for now.
>Since Debian has really strict quality policy, your code wouldn't be
>allowed to enter 'unstable' version anyway until it's mostly
>bug-free.
heh. that will be interesting to establish for a program of this
complexity.
--p
Reply to: