head	1.4;
access;
symbols
	multivol-2-1:1.4;
locks; strict;
comment	@# @;


1.4
date	98.10.26.13.58.23;	author schaefer;	state Exp;
branches;
next	1.3;

1.3
date	98.10.23.16.34.37;	author schaefer;	state Exp;
branches;
next	1.2;

1.2
date	98.10.22.18.12.50;	author schaefer;	state Exp;
branches;
next	1.1;

1.1
date	98.10.22.18.03.27;	author schaefer;	state Exp;
branches;
next	;


desc
@@


1.4
log
@fix
@
text
@OK
- boundary case when WRITING to NTL tested (NTL-X136)
    tar --block-size=64 -cvf - /usr/bin /usr/users /usr/lib | /usr/local/bin/multivol --medium-changer-script=/usr/local/bin/medium_changer.sh --output-file=/dev/st0 --block-size=32768
NOT YET OK- boundary case when READING from NTL (NTL-X136) tested, including
  data integrity (--compare with tar), provided you eject the tape
  first (mt -f /dev/st0 offline) so the right one can be loaded.
    /usr/local/bin/multivol --medium-changer-script=/usr/local/bin/medium_changer.sh --input-file=/dev/st0 --block-size=32768 | tar --compare --block-size=64 -vf - 
- for OL8 (not NTL), you need to put the tape-init and tape-deinit scripts
  as shown in examples but not in ./test.sh

BUGS
   - code is very badly written
   - no error checking on scripts.

[ untested yet: no small tape ]


@


1.3
log
@fixes
@
text
@d14 3
a16 1
   - TODO: if infile specified, and short read, load next tape.
@


1.2
log
@more
@
text
@d1 14
a14 7
- should only write at the specified block size (ie buffer till size
needed)
- for the above does only work at 4096 and less block sizes.
- but bizarrely, this is not what diagnosticed. Also tar|dd does this
  funny 4096 ...
- boundary case when creating a multi-tape untested
- reading it was tested (with cpio), however we changed stuff.
@


1.1
log
@Initial revision
@
text
@d4 2
@
