On Wednesday, 20 March 2019 14:59:39 PDT René J.V. Bertin wrote:
> >Because it is a major rewrite of QtNetwork code interfacing with OpenSSL.
> >Such change cannot go to LTS branch [1]
>
> Now maybe (though I'd argue this is a bug fix; OSSL 1.0 will go EOL 5 months
> before Qt 5.9). But that was not the question.
>
> 5.9.0 was released on May 31st 2017
> (https://en.wikipedia.org/wiki/Qt_version_history#Qt_5), two full months
> after the fix in question.
You're looking at the wrong date.
$ git show --pretty=fuller cfbe03a6e035ab3cce5f04962cddd06bd414dcea | head -7
commit cfbe03a6e035ab3cce5f04962cddd06bd414dcea
Author: Richard J. Moore <r...@kde.org>
AuthorDate: Thu Mar 23 12:43:22 2017 +0100
Commit: André Klitzing <aklitz...@gmail.com>
CommitDate: Tue Jul 4 18:03:59 2017 +0000
QSslSocket: OpenSSL 1.1 backend
The commit was *begun* two months before the Qt 5.9.0 release, or one month
after the alpha1 release. It took more than three months from the first upload
for the change to be accepted. 5.9.1 had been released by that time.
> Qt 5.10 was released half a year later, so the
> fix could easily have gone into 5.9.0 or 5.9.1 .
It was decided not to. As the length of time shows, the development of this
change was not trivial. Applying it to the released branch was not only a
violation of the feature freeze, but also potentially irresponsible, since it
was new code that had bugs. If you search the commit log, you'll see a number
of OpenSSL 1.1 bugfixes.
PS: you should backport those fixes too.
--
Thiago Macieira - thiago.macieira (AT) intel.com
Software Architect - Intel System Software Products
_______________________________________________
Interest mailing list
Interest@qt-project.org
https://lists.qt-project.org/listinfo/interest
