[Webkit-unassigned] [Bug 29803] [Qt] QWebHistory and QWebPage autotests crash

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=29803





--- Comment #28 from jedrzej.nowacki at nokia.com  2009-10-27 01:33:04 PDT ---
(In reply to comment #27)
> (In reply to comment #25)
> > > I think the API contract should be changed to not emit frameCreated from the
> > > QWebPage c'tor and that it is guranteed that the mainFrame() exists...
> > This is an awesome catch! Thanks. Because of that the solution is completely
> > bad. The only workarounds I can find for the problem are based on
> > MagicSignalRouter(tm) class or on some operation on the EventLoop. This would
> > be evil and an overkill.
> > So I believe we should leave current solution. It is more readable and it will
> > be easier to maintenance.
>
> hehe, I like the killing of createMainFrame in your solution a lot. I will try
> to poke Simon once he is back from vacation.
Another solution: Add info to documentation that QWebPage constructor always
creates the main frame and it is not possible to catch the signal. It isn't big
deal because of the QWebPage::mainFrame method. The solution have one
disadvantage it can be implemented only in a major release of Qt, it is
behavior change. Maybe it is worth to wait for it. Current solution is not so
bad.

> > Really, the lines are meaningless and actually I think they are confusing, an
> > empty lines should divide some logical blocks, but this isn't the case. Since I
> > had to modify the part of source code I decided to reformat it.
>
> Right they are meaningless, why do you change them then? The main point is that
> you propose to fix a specific problem and when the patch contains unrelated
> things it gets harder to review. As a reviewer you will have to ask yourself if
> the contributor is using a Yedi mind trick to fool you or what he tries to
> hide.
OK, I understand now, I'll try to avoid it.

> The other part is that if one is not enforcing a conherent coding style one
> ends up having a mess like the symbian kernel where {} are freely placed with
> every kind of indention.
Agree, but I don't think it should be done by humans, we have computers and
cool scripts for it. OK, I know it isn't the place and the time for this
discussion.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.