This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: are there any exceptions(cases) that glibc is not strict backward-compatibility?
- From: Paul Eggert <eggert at cs dot ucla dot edu>
- To: Jason Yang <jasonyangshadow at gmail dot com>
- Cc: libc-alpha at sourceware dot org
- Date: Sat, 18 Jan 2020 13:52:04 -0800
- Subject: Re: are there any exceptions(cases) that glibc is not strict backward-compatibility?
- References: <0b567792-5065-296d-eac0-4b4a70bc52da@gmail.com>
On 1/18/20 7:09 AM, Jason Yang wrote:
are there any cases(exceptions) that we need to notice in
terms of glibc backward-compatibility?
Every time glibc's behavior changes, there is a potential backward-compatibility
problem. Even if the behavior is just to add a feature that wasn't there before,
it's theoretically possible that old programs assume that the feature is absent
and misbehave if it's present. Plus, old code may depend on buggy glibc behavior
and thus may break when a glibc bug is fixed.
That being said, the glibc developers are quite conscious of
backward-compatibility issues and go to extraordinary lengths to avoid them.
Where can I find these exceptions?
The source code's top-level NEWS file is a good place to start. Although it
doesn't list *every* backward-compatibility issue, it lists the ones deemed
significant for practical code.