Re: Linux tape drivers

Re: Linux tape drivers

[Andrew Morton]

On Wed, 4 Apr 2007 12:22:35 -0700
"Willem Riede" <wrlk@riede.org> wrote:

> On 4/4/07, Andrew Morton <akpm@linux-foundation.org> wrote:
> > On Wed, 4 Apr 2007 16:26:14 +0200 Kern Sibbald <kern@sibbald.com> wrote:
> >
> > >  I'm all in favor of deprecating ide-tape if ide-scsi is a viable alternative,
> >
> > ow.  ide-scsi is in very bad shape and nobody is maintaining it or fixing
> > bugs in it or anything.  The only reason we retain ide-scsi at all is, err,
> > because some tape drives need it.
> >
> > If we have no alternative to using ide-scsi, and if the tape controllers in
> > question are not some terribly obscure things which nobody would seriously
> > use then we have a bit of a problem.  I don't know of anyone who we can
> > turn to to get anything done in ide-scsi.
> 
> Andrew,
> 
> Kai suggests that libata should be able to take over from ide-scsi.
> Do you concur?

That would be very nice if we can make it happen.  I don't know how close
we are to that, but the fine folks on linux-ide might be able tell us.

> That may be a better way forward than ide-scsi.
> I know, because I had a hard enough time with it when I needed
> to fix it so it worked for osst, at least.
> 
> Regards, Willem Riede.

Re: Linux tape drivers

[Bartlomiej Zolnierkiewicz]

On Wednesday 04 April 2007, Andrew Morton wrote:
> On Wed, 4 Apr 2007 12:22:35 -0700
> "Willem Riede" <wrlk@riede.org> wrote:
> 
> > On 4/4/07, Andrew Morton <akpm@linux-foundation.org> wrote:
> > > On Wed, 4 Apr 2007 16:26:14 +0200 Kern Sibbald <kern@sibbald.com> wrote:
> > >
> > > >  I'm all in favor of deprecating ide-tape if ide-scsi is a viable alternative,
> > >
> > > ow.  ide-scsi is in very bad shape and nobody is maintaining it or fixing
> > > bugs in it or anything.  The only reason we retain ide-scsi at all is, err,

I haven't heard about any major bugs besides well known module unload problem.

What have I missed?

> > > because some tape drives need it.
> > >
> > > If we have no alternative to using ide-scsi, and if the tape controllers in
> > > question are not some terribly obscure things which nobody would seriously
> > > use then we have a bit of a problem.  I don't know of anyone who we can
> > > turn to to get anything done in ide-scsi.
> > 
> > Andrew,
> > 
> > Kai suggests that libata should be able to take over from ide-scsi.
> > Do you concur?
> 
> That would be very nice if we can make it happen.  I don't know how close
> we are to that, but the fine folks on linux-ide might be able tell us.
> 
> > That may be a better way forward than ide-scsi.
> > I know, because I had a hard enough time with it when I needed
> > to fix it so it worked for osst, at least.
> > 
> > Regards, Willem Riede.

Re: Linux tape drivers

[Andrew Morton]

On Wed, 4 Apr 2007 23:31:06 +0200
Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> wrote:

> On Wednesday 04 April 2007, Andrew Morton wrote:
> > On Wed, 4 Apr 2007 12:22:35 -0700
> > "Willem Riede" <wrlk@riede.org> wrote:
> > 
> > > On 4/4/07, Andrew Morton <akpm@linux-foundation.org> wrote:
> > > > On Wed, 4 Apr 2007 16:26:14 +0200 Kern Sibbald <kern@sibbald.com> wrote:
> > > >
> > > > >  I'm all in favor of deprecating ide-tape if ide-scsi is a viable alternative,
> > > >
> > > > ow.  ide-scsi is in very bad shape and nobody is maintaining it or fixing
> > > > bugs in it or anything.  The only reason we retain ide-scsi at all is, err,
> 
> I haven't heard about any major bugs besides well known module unload problem.
> 
> What have I missed?

oop, caught making unsubstantiatable assertions.

I've seen an ongoing dribble of doesn't-work and it-crashes reports and I
simply have not made any record of them, because it's ide-scsi and I don't
expect there's anything we can do about them :(

Often these reports are from people who are writing CDROMs and the usual
response is "use cdrecord dev=/dev/hdc".

I think Alan and Jens might have some thoughts on the ide-scsi status?