C. Michael Pilato wrote:
> Justin Erenkrantz <justin@erenkrantz.com> writes:
>
>
>>>That said, the common case is to create svn_error_t's via our constructor
>>>functions, and I can't think of a good reason someone would be depending
>>>on the size, but it is technically a violation of our compat policy.
>>
>>I really don't think adding a field to the end of the structure
>>violates our compatibility policy. So, this doesn't warrant bumping
>>the structure number. No one is likely to be allocating it themselves
>>(as we provide it) and we only return a pointer to the svn_error_t not
>>the structure itself.
>
>
> Man, I hope so. I can't imagine how painful patch application and
> backporting will be if we convert all our svn_error_t's to
> svn_error_t2's.
I still think it's technically a violation of the policy, but
considering the difficulty of actually reving that structure and the
fact that anyone making use of svn_error_t is almost certainly not going
to notice any problem as a result of this change, I suspect we can
simply make the change, document it in the release notes, document the
fact that the size of svn_error_t cannot be relied on in the future, and
get on with our lives.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Apr 13 20:49:20 2005