supported hardware

[Tom Rhodes]

On Mon, 5 Jan 2004 11:07:13 -0800
Brooks Davis <brooks at one-eyed-alien.net> wrote:

> On Mon, Jan 05, 2004 at 01:45:34PM -0500, Tom Rhodes wrote:
> > On Mon, 5 Jan 2004 08:56:32 -0800
> > "Bruce A. Mah" <bmah at freebsd.org> wrote:
> > 
> > > If memory serves me right, Christoph Devenoges wrote:
> > > 
> > > > if i connect my nikon SQ digicam via USB, (after setting USB mode to Mass
> > > > Storage in the camera setup menu)
> > > > 
> > > > this appears:
> > > > 
> > > > umass0: Nikon Nikon COOLPIX SQ, rev 1.10/1.00, addr 3
> > > > da0 at umass-sim0 bus 0 target 0 lun 0
> > > > da0: <Nikon Digital Camera 1.00> Removable Direct Access SCSI-2 device 
> > > > da0: 1.000MB/s transfers
> > > > da0: 983MB (2013922 512 byte sectors: 64H 32S/T 983C)
> > > > 
> > > > you might what to add this to the 5.1|2 i386 supported hardware list
> > > > with the other umass driven hardware.
> > > 
> > > I (with the help of several other folks) have been trying to move
> > > instances of specific devices out of the hardware list (where it
> > > largely duplicates the content in manual pages) into the manual pages.
> > > 
> > > umass(4) has some of this content already, but before somebody brings
> > > over the devices from the hardware list, there's a question of how
> > > long we want that manual page to get.  There's a *lot* of umass(4)
> > > devices in the world and it's not clear to what extent we want to even
> > > *try* to exhaustively list everything there.
> > > 
> > > I see where trhodes@ has done some work on this manpage recently.
> > > Tom, I haven't had a chance to ask you about this yet, but do you have
> > > an opinion on how to handle this?
> > 
> > I've read over the comments made by Brooks Davis and must agree
> > that listing every device will be tedious if not annoying.  But
> > if you look over the uscanner manual page it has a ton of devices
> > listed also.  For the time you put into the release and hardware
> > notes, Bruce, you are to be soluted; I just cannot think of a
> > better alternative right now.  We have the hardware notes, we
> > condence them by adding to the manual pages, now we have the
> > possibility of having bloated manual pages.  It's a never-ending
> > battle.
> > 
> > Off the top of my head, while it isn't the best idea in the world,
> > a single manual page with all the USB devices might be an option.
> > Something like usbdevs.4, and perhaps the ability to link to a
> > specific section via the hardware notes.  Or just the manual page
> > with a simple comment:
> > 
> > "Due to the overwhelming amount of USB devices FreeBSD supports,
> > a single manual page was added to list them all.  Please scan through
> > this list for your specific device."
> 
> I intended my comment to be specific to umass devices that are probed
> based on their class rather then devices like uscanner that need to
> be recorded by vendor/device.  The path to listing all working umass
> devices is also the path to listing all ATA and SCSI disks, all ATAPI
> CD-ROM drives, external serial modems supporting the AT command set, all
> USB and PS/2 mice, etc...  In general, I think we need to avoid listing
> devices that are probed by class unless there's something weird about
> them.  With umass, even if you had a couple people with unlimited credit
> cards sitting around and ordering every USB keychain gizmo they could
> find, the list would always be out of date and incomplete even in CVS.
> 
> For devices like uscanner which have large lists, but where support is
> hard wired, I don't think it's unreasionable to keep the list in the
> manpage.  Sure it's long, but there's some actual basis for it in the
> source tree so maintanence is possiable.

Point well taken and I have to agree.  It would be difficult to
list every single device; it would be nice if someone who ran
one of those large information sites (like freshports) to set something
like this up.  That way users could just submit what works and
what doesn't.  But this is getting a bit beyond the original scope
of this discussion.

-- 
Tom Rhodes