RE: rpm: bzip2 compressed payload

["Wichmann, Mats D" <mats d wichmann intel com>]

>> > Why not just say that the payload can be compressed, implementation
>> > specific, and that gzip support is required, but bzip2 support
>> > is optional, and other types of compression can/will be added
>> > if/when necessary. That's what I would do anyways, your LSB
>> > mileage may vary ...
>>
>> If I can build a package that can't be installed everywhere, then the
>> standard has failed ...
>
>By that criterion, the cause may be hopeless.  What about stuff like
>glibc versions?

That's abstracted away; depending on 'lsb' (or for 2.0, 'lsb-core')
gives you a promise about certain behaviors from glibc and other
libraries and there's no need (nor is it allowed) to depend on
specific glibc versions.