Ghost effect while editing table cells

Dirk Hohndel dirk at hohndel.org
Tue Jan 7 14:04:39 UTC 2014


On Tue, 2014-01-07 at 17:46 +0100, Patrick Valsecchi wrote:
> On 07. 01. 14 17:15, Lubomir I. Ivanov wrote:
> > On 7 January 2014 17:12, Patrick Valsecchi <patrick at thus.ch> wrote:
> >> On 07. 01. 14 16:02, Lubomir I. Ivanov wrote:
> >>> On 7 January 2014 16:52, Lubomir I. Ivanov <neolit123 at gmail.com> wrote:
> >>>> On 7 January 2014 16:44, Patrick Valsecchi <patrick at thus.ch> wrote:
> >>>>> Hi,
> >>>>>
> >>>>> Am I the only one seeing a ghost effect when editing a text field within
> >>>>> a
> >>>>> table? For example, look at the attached screenshot. I was editing the
> >>>>> end
> >>>>> pressure of a tank and you can see both the previous value and the value
> >>>>> I'm
> >>>>> editing superposed.
> >>>>>
> >>>>> I'm using the latest GIT version, with Qt 4.8.4 on Linux Ubuntu 13.10.
> >>>>>
> >>>> i think it's not storing the "end pressure" values.
> >>>>
> >>>> but also asserts:
> >>>> ASSERT: "current != NULL" in file qt-ui/models.cpp, line 94
> >>>> with test30.xml and if editing "end pressure" of the first cyl.
> >>>>
> >>> this is the new getDiveById() that Dirk added recently.
> >>>
> >>> hmm, both the cyl. and weight models seem to call getDiveById() and it
> >>> returns NULL, yet previously setDive() is called and it assigns '
> >>> currentId'.
> >>> so i don't see what is wrong on a quick look. will test some more...
> >>>
> >> Lubomir,
> >>
> >> I was not talking about an assert problem. Just a visual glitch.
> > patrick, do you see the same glitch after 409c2e9?
> >
> The screenshot is taken with 409c2e9. I guess I'll have to figure out 
> how to turn on debugging for seeing the asserts...

No, you guys are talking past each other.

Lubomir saw another result of the bug that Patrick reported before and
that I fixed in 409c2e9bcfd8.

Patrick is seeing another instance of "Qt is broken on Ubuntu", IIRC.
I certainly see this on neither Fedora, Windows or Mac.

/D



More information about the subsurface mailing list