There is also another point that I think wasn't yet mentioned - I, as a non DD, create packages mainly because I myself want them, because it is very comfortable. However, with a new package, it takes usually a month or two, until it hits the unstable (many days to find a sponsor and fix all small problems, on month in the new queue), but I need the packages immediatelly.
So what I do is that I create a working deb in a few minutes (maybe not lintian clean and conforming to all debian policies, but nevertheless working), upload it to my repository with a version 0.6.0-0oc1 and then gradually improve it using versions (0.6.0-0oc2, 0.6.0-0oc3...) in the meantime search for the sponsor, he also tells me some tips to improve, then the package is ready, so the sponsor uploads it to the NEW queue with a version 0.6.0-1, but it waits there for a month, but I need updates of my package now, so I create versions 0.6.0-1oc1, 0.6.0-1oc2, ..., and when the package hits unstable, I just create a new changelog entry 0.6.0-2 and ask the sponsor for upload. So I am satisfied, because I always have a freshly updated package available immediatelly on all my computers and also it gets to Debian. Ondrej On 7/16/07, Neil Williams <[EMAIL PROTECTED]> wrote:
On Mon, 16 Jul 2007 19:28:09 +0200 martin f krafft <[EMAIL PROTECTED]> wrote: > Anyway, we're not here to agree on one procedure, are we — because > we likely never will. True. :-) > You gave your new policies and I added my 2¢ > in the form of showing you how I do things. Or would like to do > them. Actually, the discussion that ensued from my original email has helped me finalise how I want to handle the changelog in future. It has been very useful to see how different sponsors deal with the same problem. All I need to do now is ensure that I put it on my website somewhere so that maintainers know what each sponsor needs. -- Neil Williams ============= http://www.data-freedom.org/ http://www.nosoftwarepatents.com/ http://www.linux.codehelp.co.uk/
