DConf configuration system
wheeler at kde.org
Sun Apr 10 07:41:18 EEST 2005
On Wednesday 06 April 2005 23:36, Chris Lee wrote:
> Ah, and that's the fun part. See, it turns out, this is a solution in
> search of a problem.
Actually -- while I'm not going to get into the specifics of GConf vs.
KConfig, it's most definitely a problem -- and in fact a problem that's more
important than whether GConf or KConfig is cooler.
Here's the basic problem:
Configuring a Linux box is hard.
It's hard because it's damn near impossible to write a tool that can configure
most of the stuff on it because everything is using its own configuration
formats. Let's just assume for the sake of argument that YaST is the best
one out there for doing this -- it's still a huge hack that's incredibly
difficult to maintain because of the huge domain of encapsulated information.
There are places that Linux on the desktop won't be able to go until we can
get the number of configuration formats down to a very small number. And
really I don't mean just on the desktop, but I don't think we're going to
have a sane case to pitch to other tool developers until we can get this
sorted out just between a couple of desktop projects.
This is one of the Big Problems (tm). We've got some smart people on this
list and it's certainly not beyond them to solve problems when them come up
if they start working on something like this. At some point we'll just have
to get everyone in a room with enough beer, pizza and stimulants and knock
this out once and for all. Until then iterative approaches, even if they're
not likely to a lot of buy in certainly don't hurt. I think your point about
the naming is rather appropriate, but with a slightly different
interpretation -- if some iteration of the above won't work for KDE it's not
we're going to be tricked into using it, so why fuss about the name?
Sorry, normally I ignore the N-th iteration of this thread on here (apply the
standard excuse of "this is important, but I'm way too overcommitted
already"), but I just felt like you were quite off here Chris. ;-)
The fact that an opinion has been widely held is no evidence whatever that it
is not utterly absurd.
More information about the xdg