On 03/03/2015 01:46 AM, Andrew Stitcher wrote:
I'm happy to let the new API work be more carefully reviewed. The only
reason to me to get it in 0.9 is that 0.9 was intended to be a point for
API stability from then on. And the transport API is a significant
change in the engine API. Pushing it off means allowing 0.10 to break
API compatibility.

To be confident enough to guarantee stability, I think an API needs to be well understood and have been used for a reasonably wide range of intended scenarios.

From the thread on the wiki, it sounds like the proposed API is still changing. There are also some aspects not yet implemented which may impact the API. I know I am still trying to get my head around all the implications and without wanting to impute my slowness to anyone else, there may be some who have not yet read the details.

If this was something entirely new - purely an additive change - then this might not be a big deal and getting it released would allow for even wider review. However, for changes, which may impact other code, it seems risky to squeeze it in while aiming for an RC next week.

One thing I think would help would be some examples, even just code snippets, showing how the API would be used in various different scenarios (including e.g. explicit EXTERNAL SASL over SSL, DIGEST-MD5 authentication with no encryption - max ssf 0 -etc).

Reply via email to