HAL must die!
perrin at apotheon.com
Thu Mar 17 20:00:35 UTC 2011
On Thu, Mar 17, 2011 at 05:35:04PM +0100, Ivan Voras wrote:
> Well yes, that's one thing: if you use HAL, everything must use HAL
> and you can't pick and match incompatible applications and force half
> of the things in xorg.conf.
There should be a way to override it on a piecemeal basis. Seriously.
Everybody who thinks it's a good idea (by way of analogy) to write
command line utilities that default to not letting you specify any
options at all, and if you use one option to do something non-default you
have to specify *all* options even when the specification is exactly the
same as the default -- raise your hands.
I don't expect to see many hands.
> As for me, I'm running without any xorg.conf (really, I don't have the
> file at all), all my desktop applications are using HAL and everything
> is autodetected and "just works", which is as it should be.
In my experience, about one third of the time HAL makes X work great, and
the other two thirds of the time it fails in some way that requires me to
create a complete xorg.conf file just for one or two options.
Worse, from what I've seen, the xorg.conf auto-generation doesn't always
match the defaults HAL provides very well. As a result, auto-generating
an xorg.conf often results in a configuration that works even *worse*
than HAL's automatic configuration, and as a result a lot of work needs
to be done to get configuration to work.
For something like what HAL does, in addition to allowing piecemeal
custom configuration, it should also provide a way to get the *exact*
configuration the automatic configuration management software. I haven't
looked into it in *too* much depth, but so far I have not seen any sign
of that kind of helpful functionality either.
It's a common problem for popular software systems these days, I think.
People design software meant to eliminate the configuration and
management hassle from the end user, but it doesn't always work
perfectly. Unfortunately, it so zealously attempts to guess what the
user wants that it effectively *disallows* easy fixes when the user
discovers that something needs to be "fixed". This is, in short, bad
I blame Microsoft, GNU, and Canonical for this trend, mostly.
Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 196 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-questions/attachments/20110317/098cedbd/attachment.pgp
More information about the freebsd-questions