[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Latest Mandrake



Mariusz Przygodzki wrote:
> 
> Unfortunately I have used Mandrake Cooker (beta 7) many times. I can agree
> draktools look good on a screen (and it's a good idea generally) but a
> quality of draktools is/was still very low (some important features were
> dissapearing between upgrades randomly so an usability is/was far from good).

The guys who used it could start "net sharing" without editing any
stupid config files. They just didn't know good enough about the fields
they had to set for their eth cards, and I helped them figure out.

But overall, I think it would take me a lot of time in Debian to
make that machine provide internet services it got from cable modem
to other people (using windows) with dhcp. And surely, those people
would probably not be able to do it by themselves. They would need
a horrible "linux system administrator" guy who would configure the
thing for them. With that config tool, they don't. I was awe-struck at
the speed of things. I didn't read _any_ documentation to do it.

If we take a tool like that and standardize/stabilize it for debian,
it would add many many points to debian's value. It seemed to me as a
config tool that was concerned about the user, which is something I
haven't seen for ages in config tools.

Also, Mandrake seems to have better default configurations for gnome/X11. 
Well, in a way, stuff like internationalization works... Whatever good
they've got in there, we lack it.

For instance, guys had xemacs on a version, but the default configuration
had syntax highlighting turned on and had fancy nextstep icons. It looked
and felt better than our default configuration of xemacs. (Of course
it sucked in some ways, but that isn't the point. What they do we can do
better if we want)

My suggestion is this:
  * A group works out certain "scenario"s. Certain common tasks that
    a debian user should be able to perform easily.
  * Works them out in such a way that all relevant subsystems are enhanced
    to make that scenario happen.

Example: _no_ kernel recompilation to have "net sharing". Whatever,
the specifics aren't important.

Just my 2 cents,

Thanks,

-- 
Eray (exa) Ozkural
Comp. Sci. Dept., Bilkent University, Ankara
e-mail: erayo@cs.bilkent.edu.tr
www: http://www.cs.bilkent.edu.tr/~erayo



Reply to: