On Mon, Oct 13, 2008 at 11:26:22AM -0400, Arjan van de Ven wrote:
On Mon, 13 Oct 2008 17:05:08 +0200 Patrice Dumas pertusus@free.fr wrote:
In that case skipping some of the fedora release, like doing the LTS stuff only each year may help.
before solving the "how many" problem... lets be realistic
I agree, but the issue of space may be a show stopper for some, so better think about it.
- Making every release LTS is not feasible, there's just not going to
be enough manpower for that (history has shown that) for now at least.
Agreed.
- I don't think users would even really demand such a beast
I don't know in general, but I know that I stopped proposing people that were not absolute power users to use fedora, even if they were fluent enough because of this issue. So fedora lost at least 2 users. But they were not users likely to contribute, so they don't weight that much.
- The volunteer base is a bit thin; current fedora volunteers tend to
work on the latest version (or maybe two). In fact, so do many users.
There are only 2 versions... But it is true I don't think there are that many packagers interested in LTS, for the user, I am not sure. But I also think that the amount of work for the 2 first year is not high, no updates, for most packages is right. There are indeed some big and complex core packages that may be problematic, we'll see.
I would suggest focusing on first getting ONE such LTS release going. Figure out a plan on how to get out of it first (communication wise etc) in case interest isn't what you expect, and how to monitor the health. With that I suspect the fedora leadership will be happy to endorse it happily.
My suggestion is first to communicate only to the packagers that they can update the packages, look whether there are some that are updated, especially those in @code and @base, and especially when there is a knowledge that something is broken or a security issue, create a SIG if it looks that there is support, discuss the details, see how many participants, make the rules and only afterwards communicate something to the exterior.
Do one. Get that working. Don't even think about multiple until you have gotten one down. You'll learn so much that it's not worth planning beyond that.
Indeed, maybe the best would be to start with F9 such that the infrastructure doesn't have to be revived.
-- Pat