Greetings,
I'm trying to bump the htslib suite to version 1.19 in Debian sid, and symbols tracking caught the drop of the bgzf_idx_amend_last function. As far as I understand, this function was not supposed to be part of the ABI in the first place, given its declaration in hts_internal.h. Besides, running the test suites of the many reverse dependencies, without rebuilding them first, does not raise any ill effects (except perhaps cyvcf2, which seems to have a test item failing to write VCF4.1 files properly, but as far as I can see, it is unrelated to the missing symbol). So the function does not look to have made it to the ABI by accident to consumers of the htslib.
Just checking, may I go ahead with the missing symbol without soversion bump, or was the drop actually unintended?
Have a nice day, :)
Étienne.
Greetings,
I'm trying to bump the htslib suite to version 1.19 in Debian sid, and symbols tracking caught the drop of the bgzf_idx_amend_last function. As far as I understand, this function was not supposed to be part of the ABI in the first place, given its declaration in hts_internal.h. Besides, running the test suites of the many reverse dependencies, without rebuilding them first, does not raise any ill effects (except perhaps cyvcf2, which seems to have a test item failing to write VCF4.1 files properly, but as far as I can see, it is unrelated to the missing symbol). So the function does not look to have made it to the ABI by accident to consumers of the htslib.
Just checking, may I go ahead with the missing symbol without soversion bump, or was the drop actually unintended?
Have a nice day, :)
Étienne.