Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 9 Apr 1996 12:58:24 -0700 (MST)
From:      Terry Lambert <terry@lambert.org>
To:        bde@zeta.org.au (Bruce Evans)
Cc:        freebsd-current@freebsd.org, uhclem@nemesis.lonestar.org
Subject:   Re: floppy format detection [was Re: devfs questions]
Message-ID:  <199604091958.MAA05719@phaeton.artisoft.com>
In-Reply-To: <199604090830.SAA14558@godzilla.zeta.org.au> from "Bruce Evans" at Apr 9, 96 06:30:11 pm

next in thread | previous in thread | raw e-mail | index | archive | help
> I was pleased to forget about these complications for PC floppies.
> Originally, PC floppies were all 512*8 interleave 1.
> 
> I don't like automatic media detection because it slows things down and
> reduces robustness

Without automatic media detection, automounting is impossible.
Automounting speeds things up and increases robustness, especially
for users who wouldn't know a /dev/fd0.1440 if it jumped up and
bit them on the butt.

A user can see a floppy, and a user can see a floppy drive, and
can pretty much figure out what goes where (here come the war stories
about the secretary who folded the 5.25" disk in 4 to get it into
the 3.5" drive, etc.).  But a typical user *can't* see the disk format
and doesn't *want* to see a mount command or the /dev directory.


> - if media detection isn't implemented or working,

That's easy.  Make it work.  You can't argue from an assumtion of
failure.

A failure to identify media format on disk insertion means that
the disk is unformatted and you need to send a message to the
media daemon telling it to put up a "Format not recognized -- format
this disk now?" requester.

> then you have to assume that an error means that the media [density]
> changed, and flail around attempting to detect the new media.

Yep.  Don't buy old hardware.  Fortunately, old hardware isn't sold
in stores near you.

If you absolutely must use old hardware to make the "pro old hardware"
argument work (not that it worked before, keeping the install under 2M
or on 1.2M floppies)... then make the flailing about come up with the
right answer.

It is never wrong for old hardware to be slower than new hardware at
a given task.  Format recognition is one such task.

Want a higher resolution display?  Buy newer hardware.  Want 8.5ms
seek times on your hard sidk instead of 65ms seek times?  Buy newer
hardware.  Want faster format recognition?  Buy newer hardware.


> >Don't forget that even with PC-AT compatible systems, there are still
> >drives out there that have broken media-change reporting (which MS-DOS
> >conceals by assuming media change is broken on ALL drives), so the media
> 
> Nope, versions 4.01 and 6.22 screw up my disks by assuming the the
> change line works on my broken 3.5in drives.

Yep.  You need to test the change lines by looking for a media change;
the first time you see one that also gets change notification, you
set a bit in a writable store somewhere.  Like userconfig does.


> >damage to the hardware.   I personally prefer having an semi-automatic
> >dev plus manual devs or an ioctl available for explicit settings.  In the
> >PC/AT world, the CMOS provides the "hints" needed to perform semi-automatic
> >operations and get the test time down to an acceptable level.
> 
> I prefer using only standard formats so that format detection is
> normally unnecessary.  Reading foreign formats is probably sufficiently
> rare that the driver shouldn't have any complications to support it - it
> should just support it.

Yes, yes, yes.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199604091958.MAA05719>