Left out of the Xfce/udev example is an explanation of why software can't stick with HAL on systems that don't support udev. Is it just a matter of code divergence going too high up the abstraction layers, and not enough HAL-using testers?
Posted Mar 2, 2011 18:40 UTC (Wed) by drag (subscriber, #31333)
[Link]
Sound like a combination of lack of manpower and lack of application developers involved that use Unix systems that require HAL.
Xfce, udev, and HAL
Posted Mar 2, 2011 19:37 UTC (Wed) by nix (subscriber, #2304)
[Link]
The whole reason for the existence of things like udisks is that they provide a consistent API that another package can implement to provide (e.g. disk) hotplugging info on systems that don't support udev. It's probably easier to implement that than to implement all of udev.
(If the problem is that FreeBSD doesn't have a useful hotplugging infrastructure yet, that's a different matter, but I thought it had one.)
Xfce, udev, and HAL
Posted Mar 2, 2011 19:43 UTC (Wed) by mezcalero (subscriber, #45103)
[Link]
I am pretty sure that's not what davidz had in mind when he wrote udisks...
Xfce, udev, and HAL
Posted Mar 2, 2011 22:26 UTC (Wed) by nix (subscriber, #2304)
[Link]
I was thinking of the original 'HAL is deprecated, use udisks/u*' email, but I didn't actually look it up so I could very well be wrong.
Xfce, udev, and HAL
Posted Mar 4, 2011 8:38 UTC (Fri) by michaeljt (subscriber, #39183)
[Link]
> I am pretty sure that's not what davidz had in mind when he wrote udisks...
From a quick look at the API though it does look reasonably suitable for implementations on other platforms, even if that wasn't the idea. With a big exception for the LVM stuff.