Hi,
Steve McIntyre wrote:
> Let's go with "-jigdo-checksum-algorithm" to keep the separation?
The user is king.
Other mail:
I proposed to keep versioning and SONAME at 1
> > LIBJTE1 {
Steve McIntyre wrote:
> Surely that works for upwards compatibility, but won't stop somebody
> from building and depending on the newer libjte2 then trying to run a
> binary with libjte1?
The only user of libjte already has precautions against this:
> > Fine check for compatibility happens at compile time by the required
> > minimum version (e.g. iso_libjte_req_* versus LIBJTE_VERSION_*) and
> > at run time by calling libjte__is_compatible().
This kind of test works well between e.g. xorriso's binary and libisoburn,
and then between libisoburn and libisofs.
The test between libisoburn/libisofs and libjte has never been challenged
but i am optimistic that it works.
The compile time test is at
https://sources.debian.org/src/libisofs/1.5.0-1/libisofs/messages.c/#L157
(This is how i let a passive piece of source kick an unknown compiler:
https://sources.debian.org/src/libisofs/1.5.0-1/libisofs/messages.c/#L174
)
Runtime test
https://sources.debian.org/src/libisofs/1.5.0-1/libisofs/messages.c/#L180
Here i let libjte judge whether it finds its own version not older than the
compile time version seen by libisoburn (LIBJTE_VERSION_* from libjte.h).
As long as you stay with a plausible sequence of version numbers,
libjte__is_compatible() should not need your attention.
(For the curious:
https://sources.debian.org/src/jigit/1.21-1/libjte/libjte.c/#L68
)
Have a nice day :)
Thomas
