Gnome System Tools toughts

[Mike A. Harris]

On Sun, 19 Oct 2003, Jaap A. Haitsma wrote:

>I came across the announcement of a new version of Gnome System Tools (a 
>project lead by Ximian) a few days ago. They resemble the redhat config 
>tools quite a bit, though they have only have 5 tools up to now (not 
>really production ready according to the website but still)
>
># Users and groups
># Date and time
># Network configuration
># Bootloaders
># Runlevels
>
>See: http://www.gnome.org/projects/gst/index.html
>The nice thing is that the front end (The GUI) is distribution 
>independent and the backend (configuration files etc.) are distribution 
>dependent.
>
>Just some thoughts/questions.
>
>Wouldn't it be better for RedHat, SuSE, Debian etc. to just join this 
>effort. It's a lot more effective then every distro provider making 
>their own tools. (I see that there's maybe an issue for KDE users)
>
>Are there any initiatives to have a standard of configuration files, 
>which are used by all providers. I heard of LSB, but it seems to me they 
>  do not standardise configuration files.

We've been down that road before.  "Linuxconf", as well as other 
config tools here and there throughout the years.  The biggest 
problem with externally maintained configuration tools, is that 
they tend to *NOT* be distribution specific, and so they don't 
track the development of the given distribution.  As such, they 
suffer greatly from NIH syndrome (Not Invented Here), and tend to 
not match what the distribution has to offer without a LOT of 
customization, tweaking and bug fixing, which must then be 
repeated by internal developers every single release.

It was decided to develop our own configuration tools internally, 
so that we know they are up to date with the software we ship, 
and we are the upstream maintainers, as well as not having to 
worry about distribution specifics being ignored because an 
external maintainer uses some other distribution, etc.

Before we'd be willing to throw away all of the tools we've 
already written, and future plans to improve upon that, there'd 
have to be a rather major selling point.  And "same tool works 
the same on all distros" is a nice thought, but that's not the 
*highest* priority.  What is more important than our tool being 
the same one SuSE and/or Mandrake, etc. uses, is that our tool 
actually works for our distribution without a lot of hacking, or 
developmental latency.  Our tools are developed on _our_ 
distribution release cycle.  An upstream tool is developed on 
volunteer time, and different components of such a suite would 
more likely than not have different release cycles.

So, while it is interesting to see a new tool being developed, 
the tool would have to prove itself out there before any 
consideration would be giving to throwing away a couple of years 
worth of invested development which does work for us now already.

<DISCLAIMER>
That is just my own personal opinion however.  I do not speak for
Red Hat specifically, nor would I be the one here making such a
decision as to wether or not Red Hat would include such software
in future distributions.  Such a decision would be based on 
meritocracy of the software in question, and wether it would meet 
the project's goals and/or Red Hat's goals.
</DISCLAIMER>

TTYL

-- 
Mike A. Harris     ftp://people.redhat.com/mharris
OS Systems Engineer - XFree86 maintainer - Red Hat