Hi
I just found out that forcing applydeltaiso to use libbz2-1.0.2 instead
of 1.0.3 solves the problem.
bzip2 did change its compression slightly in 1.0.3. Gentoo had that
problem before with its deltas and the scripts handle it now.
applydeltaiso seems to do not.
If anybody can reproduce the problem (and the solution) then maybe it
would make sense to bundle the 'right' libbz2 with the deltarpm package
and link it in statically (although i'm as opposed to static linking as
anybody else...).
Heiko Helmle wrote:
I have the same problem, albeit on a non-SuSE distribution (gentoo), trying
to get RC3 ISOs.
[EMAIL PROTECTED] ~/tmp $ md5sum SUSE-Linux-10.1-RC1-Addon-BiArch.iso
SUSE-Linux-10.1-RC1_RC3-Addon-BiArch.delta.iso
05644d16b959a54d6c68a997ffa95fb5 SUSE-Linux-10.1-RC1-Addon-BiArch.iso
49c163be6fcfd11ab4e43d3b3acaabc3
SUSE-Linux-10.1-RC1_RC3-Addon-BiArch.delta.iso
but the resulting RC3-ISO has the md5:
4d3c42182655136a112f4d01a35cedd2 SUSE-Linux-10.1-RC3-Addon-BiArch.iso
I'm re-running applydeltaiso to check if the corrupt ISO is created the
same
every time (and to rule out memory issues) and under valgrind - as
suggested in an earlier post.
Regards
Heiko
Andras Mantia wrote:
>> Hi,
>>
>> Until now I could apply the delta ISOs without problems. Now I cannot
>> create the CD1. I verified, the md5sum for the RC1-CD1 is OK, the
>> md5sum for the delta-iso is OK, but after applying, the result CD is
>> NOT OK, both according to applydeltaiso and a manual checking of the
>> md5sum:
>>
>> md5sum SUSE-Linux-10.1-RC1-x86_64-CD1.iso
>> 6fcd57b438477ebb90162236678ebefc SUSE-Linux-10.1-RC1-x86_64-CD1.iso
>>
>> md5sum SUSE-Linux-10.1-RC1_RC2-x86_64-CD1.delta.iso
>> ad44f70c5eee8ee7001fce102b229899
>> SUSE-Linux-10.1-RC1_RC2-x86_64-CD1.delta.iso
>>
>> The result of applydeltaiso:
>> md5sum mismatch, iso is corrupt
>>
>> Manual checking:
>> md5sum SUSE-Linux-10.1-RC2-x86_64-CD1.iso
>> d55dadb1b636d3fe3433a7ecb5e1dc32 SUSE-Linux-10.1-RC2-x86_64-CD1.iso
>>
>> It should be 47f3cf67680c0cafdf64911c0eabfc53 .
>>
>> Any idea what can I do?
>>
>> Andras
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
