VDR User wrote:
> What ever happened to the idea of setting up VDR deveopment on
> mercurial to allow the main contributors who want to work on it to do
> so without hassle/delay?
My 2c on this:
Lets think this through: An open VDR repository for everyone to get in
their personal 'I want this' patches. Surely this will lead to chaos,
since no one really oversees the needs of all VDR users. (Take
sourcecaps vs. lnbsharing as example.)
So we have to restrict things. Only a few trusted dev's getting write
access. In any case changes that go beyond simple bug fixing would
require discussions about whether things go into the right direction,
and I'm pretty sure that none of the trusted dev's would ever check in
any changes without a previous ack by Klaus anyway.
So if Klaus ack's anything anyway, why does anyone except Klaus actually
need write access?
There are however some pro's for a public repository:
- With a public repository, the need for frequent developer releases
wouldn't be that high, as people who really want the latest greatest
could pull it at any time. Dev releases could be done less frequent at
more stable points.
- Bug fixes can be pushed between point releases more easily. (There are
bug fix patches that are waiting for 1.6.0-2 or 1.7.1 for some time now.)
- With a distributed repository system like Mercurical, it would also
easily be possible to do external user branches to support more
cutting-edge features, as long as some other maintainer keeps them up to
date. Think of it as something like the -mm kernels, or as the bigpatch
/ extensions-patch. And if things get Klaus' approval, they can be
easily merged back into main VDR too.
To sum up things: A public repository may be useful, but probably in a
different way than some people think.
vdr mailing list