libcdio-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Libcdio-devel] ABI problem with: [PATCH v2] add multi extent ISO966


From: Thomas Schmitt
Subject: Re: [Libcdio-devel] ABI problem with: [PATCH v2] add multi extent ISO9660 file support to libcdio
Date: Thu, 28 Jun 2018 09:21:19 +0200

Hi,

Rocky Bernstein wrote:
> The next release will include the CD-TEXT changes but not the multi-extent
> changes? And will be ABI and API compatible? Or not?

I deem branch ts-cdtext-fix ready for inclusion.

The ABI compatible variant of multi-extent still has to be developed.


> I am concerned that changing the API/ABI too quickly that it doesn't give
> distributions time to adjust.

By introducing a new "v2" API part, it should be possible to avoid breaking
the ABI of the existing iso0660_stat "v1" API and ABI.


Pete Batard wrote:
> Sorry for your loss.

Our government is supposed to step down and the autobahn will be closed
for an hour of national depression.

> You can still play an old recording of the last WC semifinal against Brazil,

Yeah. "Gol da alemanha !". What glorious times.


Rock Bernstein wrote:
> > And will be ABI and API compatible? Or not?

Pete Batard wrote:
> If Thomas achieves what he set to do, I think we should ultimately be okay
> in terms of ABI and API compatibility. But I'll leave it to him to confirm.

Yes. That's the plan.

In my own scheme of .so numbering, the current development would be
  libcdio.so.18.1.0
  libiso9660.so.11.1.0
with the warning that the newly introduced parts of this API are still
prone to changes.

The next release would then install
  libcdio.so.18.2.0
by
  libcdio_la_CURRENT = 20
  libcdio_la_REVISION = 2
  libcdio_la_AGE = 0
and
  libiso9660.so.11.2.0
by
  libiso9660_la_CURRENT = 13
  libiso9660_la_REVISION = 2
  libiso9660_la_AGE = 0
with the promise that the new API parts will be kept stable in future.


It will probably last a few days until my proposal of iso9660_statv2 is
ready. Then it will need heavy testing.
(I could reduce the risk of regressions by leaving the old iso9660_stat
 implementation as is and using a copy of it for deriving v2. But that
 would lead to ugly code duplication and future maintainance pitfalls.)


Have a nice day :)

Thomas




reply via email to

[Prev in Thread] Current Thread [Next in Thread]