On 06/20/2013 05:18 PM, Andrew Stitcher wrote:
On Wed, 2013-06-19 at 15:30 -0400, Andrew Stitcher wrote:
On Tue, 2013-06-18 at 18:30 -0400, Ken Giusti wrote:
Folks,
There's some old QMF-related code in our repo that appears to be quite dead.
I've create a JIRA to track this work:
https://issues.apache.org/jira/browse/QPID-4940
I'm in the mood to remove lots of code from the tree ;-), so I'll
prepare a change for review and we can have a lighter tree soon.
I've prepared a change and review that encompasses this work and a
little more:
The 'little more' here refers to hiding the entire qpid::client API.
That is a significant enough change that it needs to be explicitly
brought up in its own right, not bundled in with the removal of the much
more obscure QMFv1 stuff.
I'm not clear on what the immediate benefits of this are (the code is
all still there, its just that the headers are not installed), but the
potential impact to some users is certainly significant. I've been
encouraging people to move towards qpid::messaging, but this seems
drastic and sudden for the 0.24 timeframe.
If we want to make a clear signal regarding deprecation, what about
having the qpid::client headers installed only through a separate make
target (e.g. make install-deprecated). That way users still relying on
the API get a clear warning. That option can then be removed in the 0.26
or whatever.
https://reviews.apache.org/r/12006/
Code changes:
https://github.com/astitcher/qpid/commits/remove-old-api
JIRAs:
https://issues.apache.org/jira/browse/QPID-4940
https://issues.apache.org/jira/browse/QPID-4941
https://issues.apache.org/jira/browse/QPID-4942
I really like some of the implications of this work - specifically that
it really tidies up the header files that we export to be installed for
our APIs. I think this is very important for some of the suggested Qpid
modularisation work going forward as it exposes a much cleaner
interface.
Any omments etc. please make soon on the review (or on this email
thread) as I intend to push it to trunk for the 0.24 time frame.
Thanks
Andrew
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
