Page 1 of 1

More on tooltip vibration...

Posted: Mon Aug 02, 2010 15:33
by agamemnus
Ok, I noticed that when I simply update a window's tooltip text while I am hovering over the tooltip, it vibrates. This same thing basically happened with the slider and slider thumb tooltip vibration -- since I was changing sheets, cegui was updating the text every time.

I really don't want to manually reset the tooltip Y-size to what it was before I update the text... not to mention that this trick won't work when the amount of lines changes...

I don't really know what causes it, even now. Here are the results of a few seconds of polling the tooltip x/y size data: (the position stays the same)

Code: Select all

159           24.49524
159           24.78876
159           24.08228
159           24.37579
159           24.66931
159           23.96283
159           24.25635
159           24.54987
159           23.84338
159           24.1369
159           24.43042
159           24.72394
159           24.01746
159           24.31097
159           24.60449
159           23.89801
159           24.19153
159           24.48505
159           24.77856
159           24.07208
159           24.3656
159           24.65912
159           23.95264
159           24.24615

Re: More on tooltip vibration...

Posted: Sun Aug 08, 2010 19:34
by Ident
Could you submit a small test project which displays this behaviour?

Re: More on tooltip vibration...

Posted: Mon Aug 09, 2010 20:20
by agamemnus
That would take hours and hours... perhaps days... I am not set up to create and test C++ executables and don't know how. I would also need a workable CEGUI/Irrlicht draw loop in C++, and my loop is in Freebasic, using an Irrlicht wrapper. So, no.

I could provide my program EXE showing this behavior, but what would that show?

Re: More on tooltip vibration...

Posted: Tue Aug 10, 2010 14:15
by Ident
agamemnus wrote:That would take hours and hours... perhaps days... I am not set up to create and test C++ executables and don't know how. I would also need a workable CEGUI/Irrlicht draw loop in C++, and my loop is in Freebasic, using an Irrlicht wrapper. So, no.

I could provide my program EXE showing this behavior, but what would that show?


How is anyone supposed to reconstruct the faulty behaviour your described without a test case being given or a specific error reconstruction described making it possible to reconstruct this in an environment a CEGUI developer can easily set up himself?

The executable file would indeed not help us at all because we cannot debug it.

Can you at least give any hint as to why this behavior could come up?

Could you take a CEGUI demo and change it in a way it reflects the changes you made which lead to this erroneous behaviour? This would for example make it possible to debug the behaviour.

How do you think anyone is supposed to rebuild that behaviour to find the error? It works for everyone. We don't know what you did specifically. Provide a modified demo (I think this is the fastest way) and we can help you.

Re: More on tooltip vibration...

Posted: Tue Aug 10, 2010 14:27
by agamemnus
How is anyone supposed to reconstruct the faulty behaviour your described without a test case being given or a specific error reconstruction described making it possible to reconstruct this in an environment a CEGUI developer can easily set up himself?


I explained in my first post how to recreate the error.

Have you tried to create a test program where a window's tooltip text changes or updates constantly (say, every second), and have you tried hovering over the window while doing that?

Re: More on tooltip vibration...

Posted: Thu Aug 12, 2010 01:35
by Ident
agamemnus wrote:
How is anyone supposed to reconstruct the faulty behaviour your described without a test case being given or a specific error reconstruction described making it possible to reconstruct this in an environment a CEGUI developer can easily set up himself?


I explained in my first post how to recreate the error.

Have you tried to create a test program where a window's tooltip text changes or updates constantly (say, every second), and have you tried hovering over the window while doing that?


No, why would anyone ever want to update a tooltip _each frame_ anyways? However you are lucky, since CrazyEddie is a magician and can solve such problems even without a test case, I heard. In this case a solution might soon be provided to you. Be patient, young soldier.

Re: More on tooltip vibration...

Posted: Thu Aug 12, 2010 02:39
by agamemnus
Ident wrote:No, why would anyone ever want to update a tooltip _each frame_ anyways?


1) Just because you don't have a reason, or can't think of one, doesn't mean there isn't any.
2) I said "every second", not each frame.

However you are lucky, since CrazyEddie is a magician and can solve such problems even without a test case, I heard. In this case a solution might soon be provided to you. Be patient, young soldier.


Sarcasm, really? Again... I provided all the info necessary to repeat this bug. There is no test case, but not everyone needs a test case to reproduce behavior... sometimes only a description is needed. You won't even bother to reproduce it based on the description, so instead of prancing around here and getting off on "making fun of me" instead of trying, just don't try if you don't want to. :D

Re: More on tooltip vibration...

Posted: Thu Aug 12, 2010 11:57
by CrazyEddie
The issue of misaligned tooltip rects (which was the actual cause of this) should be fixed in branches/v0-7 @ r2588.

I've locked this topic in order to prevent things descending into outright abuse.

CE