> Besides the fact that the document disregards most of the discussion
> here on the D&R list so far concerning an installer.
I didn't think it did that and if it did, it was because we have to find
out if all projects involved can come to a reasonable method of
integration before we do anything else. It's just a document to help
explain why we would bring all three together. A general understanding
of each project is assumed and for those that don't know them, links are
provided as well as a short description.
> It -in my opinion-,
> like most other installer projects (except pen2), describes an installer
> that is a full replacement for the current installation process, which i
> think shoots way beyond the target audience of an installer.
As for replacing the current process, that is not the idea. The idea is
to open up Gentoo to people who do not like, do not wish to repeat, or
do not have time to repeat the current process and there are a number of
those people. Many of them were at LWE. Of course, that's not really the
point, either. The console is always going to be there as will the
documentation; no one is looking to replace that and mandate some new
method of installing machines. This is another option for those that
want it. For those people with 300+ machines, the current process
doesn't exactly scale and this kind of project was desirable. That's all.
With regard to the target audience, there are two, yes. There are people
who want to install desktop systems using a graphic installer and those
that want an automated, unattended system. The processes vary to only a
small degree and being able to work from both a common code base and
community will not hurt any of the parties involved. Neither will take
precedence or overshadow the other. Neither is more important than the
other. I don't think it's unreasonable to work together on a project
that is beneficial to both types of users. Gentoo, itself, takes both
types of users into account already and extending the same thought
process to a project such as an installer works.
I understand that some people don't want an installer and other don't
want an automated installer and others don't want a GUI installer, ad
infinitum. No one will be forced to use either and with the amount of
people involved, working on a combined system should not detract from
either. If everyone involved in the three projects remains involved (and
I hope they do) we'll have 6+ people coding. That's a large enough team
to work on a project that is modular enough to be worked on by that many
people when separated properly.