Jeremy Katz wrote:
On Mon, 2007-08-06 at 23:24 +0200, Jeroen van Meeuwen wrote:
> Jeremy Katz wrote:
>> On Sun, 2007-08-05 at 20:29 +0200, Jeroen van Meeuwen wrote:
>>> This enables other applications using livecd-creator as a backend
>>> application, while already having initialized a yum object configured
>>> with a certain installroot (which once the yum object is initialized,
>>> cannot be changed). Matching the build environment to what
>>> livecd-creator expects results in trouble with other builds running
>>> (eg., symlinking the actual yum installroot to "%s/install_root" %
>>> self.build_dir).
>> In practice, though, this is a bit of a moot point, since the API for
>> using livecd-creator is the command-line and passing it a kickstart
>> config. At which point, the fact that you have another yum object
>> doesn't really make much difference...
> No matter what API you support, it doesn't prevent you from applying the
> patch.
But why add (and therefore have to maintain, follow indirections and
then deal with potential odd fall-out) options that aren't ever going to
be hit or used?
Because we move forward and that's what we do. We focus on new features,
software and technology.
The issue of maintaining things on the long-term applies to anaconda,
pykickstart, it's deps. Not livecd-tools. Not Revisor either. Not at
this point in time anyway.
If maintaining the code is your biggest concern though... Try and assign
all livecd-tool's bugs to me.
Kind regards,
Jeroen van Meeuwen
-kanarip