Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 2 Nov 2001 15:50:13 -0600
From:      David Kelly <dkelly@hiwaay.net>
To:        Jim Arnold <jim@ohio.com>
Cc:        Cliff Sarginson <cliff@raggedclown.net>, freebsd-questions@FreeBSD.ORG
Subject:   Re: Splitting a tar archive
Message-ID:  <20011102155013.B60019@grumpy.dyndns.org>
In-Reply-To: <a05100313b808ba9bc433@[206.128.102.10]>; from jim@ohio.com on Fri, Nov 02, 2001 at 04:16:31PM -0500
References:  <3.0.5.32.20011102113219.00f94888@mail.sage-american.com> <20011101160655.O92340-100000@bsd.smnolde.com> <20011101222033.C53366-100000@jodie.ncptiddische.net> <3.0.5.32.20011102113219.00f94888@mail.sage-american.com> <3.0.5.32.20011102124221.00f94888@mail.sage-american.com> <a05100310b808b104839b@[206.128.102.10]> <20011102204953.D6967@raggedclown.net> <a05100313b808ba9bc433@[206.128.102.10]>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Nov 02, 2001 at 04:16:31PM -0500, Jim Arnold wrote:
> >>
> >What do you mean it doesn't work !?
> >What happens then..?
> 
> 
> The file size that gets created when I cat the files
> is smaller than the original file. Here's what I did to test
> this out:

Well, lets first check the split files. As others have commented you
should be using the -b option to split so that it knows the files are
binary. Otherwise its trying to break on line count thinking the files
are text.

For homework dig into the source of split and see how it handles text.
Wonder if there is a goof where it might assume a null is end of line
when processing in text mode.

In the Bad Old Days I had one heck of a time on machines administered by
NASTRAN users. The NASTRAN (or was it PATRAN?) package had its own split
command which was for splitting models (created in PATRAN). The
administrator got bit trying to use the Unix split and "solved" the
problem by replacing it with the other split.

Back in those days SGI's IRIX split didn't know binary. So I'd use dd in
a shell script. And stashed it in ~/bin at the front of my $PATH.


Back to the current thread. If the file (movie?) you are splitting is
only to be used between Unix machines via CDROM and no need of Windows
of Mac access then you might skip the mkisofs step. And you might not.
Use burncd or cdrecord to write the split image directly to the CD-R.
Then you might resurect the file using "cat /dev/cdrom > movie.mov" for
the first segment, then "cat /dev/cdrom >> movie.mov" for each of the
remainder. The problem with this method is (I've noticed) some CD drives
report end of media with the block of data which has touched EOM and
possibly the read returned less than a full buffer, and others report
EOM only after one has attempted to read past EOM and no data was
returned. So its possible to lose the last block with some drives.
Noticed this when I wanted to verify a FreeBSD ISO by lifting it raw
off of the disk with dd. Philips ATAPI CD-RW read short. HP SCSI CD-RW
read correctly. Believe Sony ATAPI CD-RW read correctly.

Same sort of thing applies with tar. Skip the mkisofs step. Then to read
the CD-R simply, "tar -tvzf /dev/cdrom". Saves the fuss of mounting.
This is exactly the way we use tape.

-- 
David Kelly N4HHE, dkelly@hiwaay.net
=====================================================================
The human mind ordinarily operates at only ten percent of its
capacity -- the rest is overhead for the operating system.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011102155013.B60019>