On Mon, 23 Jul 2001 [EMAIL PROTECTED] wrote:
> >At the end of scd0 cdrecord reports an error (can't copy now, but the gist
> >of it was "input stream ended") and fails to fixate the disc. Basically,
> >all the data from the CD has been read, but since the reading child dies,
> >I think cdrecord notices it and dies on the spot. I would very much like
> >to hear a report that this works - what happens when the source CD ends?
> >No error message?
> Read README.copy & README.verify
I'm afraid I wasn't logging with the proper facility, so the error is
lost. It's basically an error reporting the input stream ended. From what
I can grasp from README.copy the problem is probably the "two unreadable
sectors" at the end of the disk, perhaps combined with read-ahead on the
device. Apart from using sdd or readcd, is the only way to avoid this
subtracting the two sectors from the sector count of the copy?
> >cat /dev/scd0 | cdrecord -v dev=0,1,0 -waiti -
>
> >And this worked, though it reported "fifo=0%" through 99% of the process.
> >Is there a way to do this in a more intelligent fashion?
>
> If this works, then there is a cat/kernel bug!
It worked fine (though the buffer reported 0% all the way). Why shouldn't
it? Would a 0% fifo break the recording?
I switched to using dd if=/dev/scd0 bs=64k as the source of the pipe and
the buffer reported being 100% full all the way; it worked fine,
additionally.
Perhaps a pointer to the README.copy/verify files could be included in
http://www.fokus.gmd.de/research/cc/glone/employees/joerg.schilling/private/problems.html
or, if this is a faq, add a little FAQ section so people aren't
overwhelmed by the 35 README files distributed with cdrecord. ;)
Take care,
--
/\/\ Christian Reis, Senior Engineer, Async Open Source, Brazil
~\/~ http://async.com.br/~kiko/ | [+55 16] 274 4311
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
