Goals for 0.3.0
mnews22 at wp.pl
Fri Jul 30 03:18:28 PDT 2004
Dnia 30-07-2004, pią o godzinie 04:09 +0200, Kay Sievers napisał(a):
> Now, I'm a litte lost :)
> I started to look into the hard disk serial number thing. For ide it
> seems as easy as the attached program. For scsi it would be about three
> pages of code.
> But while thinking about, I not sure anymore, that we want it. Do we
> really need stable udi's for the disk itself? Isn't it just fine to use
> the topological position for the drive-udi? Then the drive-udi changes,
> while moving the disk around, but all what we really need to care are
> about, are the udi's of the volumes. Here we have labels and uuid's to
> create a stable udi, to find the already stored HAL-properties.
> Hmm, any ideas?
Stable disk UDI also has imaginable usages, particularly all sorts of
tweaker daemons that could want to apply per-physical-device settings,
like spindown time, DMA on/off, etc. IMHO, if it is possible to provide
stable UDI for any given device, it should be done, unless there are
strong reasons not to. We can't think of all possible applications of
HAL, that's the point of it, so we shouldn't weaken guarantees without
"Tautologizm to coś tautologicznego"
Maciej Katafiasz <mnews2 at wp.pl>
hal mailing list
hal at freedesktop.org
More information about the Hal