Re: udev device naming policy concerns
Roger Leigh wrote:
> Tore Anderson <email@example.com> writes:
>> * Tollef Fog Heen
>> > I think you are overestimating the problems. I used to run without
>> > any compat symlinks, just devfs, and apart from the fact that you have
>> > to fix inittab, it mostly Just Worked.
>> When I experimented with devfs and attempted to run without the
>> my system broke down to the point of not booting fully, due to the fact
>> that the devices where my system partitions was supposed to be located
>> (/dev/hda1 and such) wasn't available any longer.
> I didn't have that problem, though I didn't remove the compatibility
> symlinks for just this reason.
>> After having fixed that I just found that many other things was
>> broken, such as for instance XMMS (due to the missing /dev/dsp and
>> /dev/mixer), my floppy drive was inaccessible (no /dev/fd0 either),
>> and so on for almost every piece of software using device nodes
> The vast majority of software in Debian should be devfs-aware already.
> When I used it for a few months a few years ago I provided patches for
> all the init scripts, config files, and so on that were broken on my
> system, and all of them were included.
>> Anyway, my complaint isn't that it will be technically impossible to
>> build a system exclusively using non-standard device names, be it
>> devfs names or something else entirely, but rather that this attempt to
>> do so right now is undesireable and serves no good purpose.
>> The current naming scheme is, after all, a universally accepted and
>> ubiquitous standard, one which has been formalized by the Free Standards
>> Group, and one which no major player in the GNU/Linux arena seem intent
>> on changing in the foreseeable future. I believe we should have a very
>> compelling rationale at the table before deciding to stray from it.
> The current single directory with a bazillion device nodes in it
> doesn't scale well. Not only is it difficult to find the nodes
> actually available on your system, it also has the potential to hurt
> performance. For example, an "ls /dev" on my system shows 1439 device
> nodes, of which perhaps 10% are actually available and 5% have
> actually ever been used. When any program scans /dev, that's 1400
> dentries created for no good reason, and I haven't got the full set of
> device nodes.
However, that isn't a problem with udev, provided udev is operating
properly, because udev is supposed to *only* create device nodes for
devices which *exist*. So this issue does not really give any advantage to
> I do vastly prefer the devfs-style names, vc/1, tts/3 etc., and I use
> these without devfs. The only aspects of devfs I disliked were:
> /dev/cdroms }
> /dev/discs } too unpredictable which device gets which number
> /dev/ide0/lun0/.... far too unweildy when /dev/hda/3 would do as well
It's all very well to provide the devfs names (it doesn't hurt), but we
shouldn't be "moving away from" the traditional names, because the
traditional names are going to be standard, and devfs is basically
Nathanael Nerode <neroden at gcc.gnu.org>
US citizens: if you're considering voting for Bush, look at these first: