[Development] Reverting the QRegExp change?

[lars.knoll at nokia.com]

On 5/2/12 10:48 AM, "ext Thiago Macieira" <thiago.macieira at intel.com>
wrote:

>Given the negative reaction here and on Gerrit, I'm wondering if we
>should 
>revert it.
>
>In fact, for the simple fact that the SIC change wasn't discussed here
>before 
>it went in (my bad, sorry for that), it deserves to be reverted.

My bad as well for approving it.

>What do you think?
>
>Another option is what Olivier has proposed:
>	https://codereview.qt-project.org/#change,24986

The change is clearly causing more issues than anticipated. Given the fact
that we'd like to deprecate QRegExp in 5.1 or 5.2 anyway and should not be
using it anymore for new code, I wonder whether changing this was the
right idea.

So I think we should at the minimum add Olivier's patch. Reverting it
completely would fit in with saying that QRegExp is done and close to
being deprecated, so there's no benefit in SC breakages in these classes.

In total I am more leaning towards reverting it, but would also be ok to
add Oliviers patch. Let's hear some other opinions.

Cheers,
Lars