Author Topic: Strange cursor position with wx3.0.2 (scintilla)  (Read 19163 times)

Offline headkase

  • Almost regular
  • **
  • Posts: 159
Re: Strange cursor position with wx3.0.2 (scintilla)
« Reply #30 on: December 28, 2015, 11:41:07 pm »
There are many more problems in wx3.0 builds that needs to be fixed.
At the moment it is best to stick to the wx2.8 builds.

wx3.0 builds are for people that are not afraid of crashes/asserts/bugs and they're willing to report them.
Ah, thanks oBFusCATed, I don't actually need a 64-bit build of C::B, it's more of a vanity thing.  Under Linux a wx 2.8.12 64-bit build is trivial, but under Windows where I am one is not.

Offline oBFusCATed

  • Developer
  • Lives here!
  • *****
  • Posts: 13413
    • Travis build status
Re: Strange cursor position with wx3.0.2 (scintilla)
« Reply #31 on: December 28, 2015, 11:41:41 pm »
@ollydbg:
I don't like the fix. We have to know the reason for this change in behaviour.
Also the check should be against 3.0 and not 2.9.5. wx2.9.x is not supported in any way, only wx3.x, thus the code should not be misleading.
(most of the time I ignore long posts)
[strangers don't send me private messages, I'll ignore them; post a topic in the forum, but first read the rules!]

Offline ollydbg

  • Developer
  • Lives here!
  • *****
  • Posts: 6035
  • OpenCV and Robotics
    • Chinese OpenCV forum moderator
Re: Strange cursor position with wx3.0.2 (scintilla)
« Reply #32 on: December 29, 2015, 12:06:17 am »
@ollydbg:
I don't like the fix. We have to know the reason for this change in behaviour.
I don't know the reason, I have debugged this bug for several days, until I asked in the wx-user maillist.
Eran just told me this is an workaround, this also get better performance as he said. see his wx-user maillist(in my previous posts)

Quote
Also the check should be against 3.0 and not 2.9.5. wx2.9.x is not supported in any way, only wx3.x, thus the code should not be misleading.
I see the CallAfter's document is said this function is from 2.9.5, so I changed my first idea which is 3.0.0 to 2.9.5.  :)

Right bleepin' on.  ;)  I await it being pushed to SVN and then I will build wxWidgets 3.0.2 and C::B wx30_64.  Finally, 64-bit.  That's all I want - is it too much to ask!? ;)
I need more comments about those patches, especially Morten, because he is the maintainer of that scintilla branch. You can apply those patches locally in your computer, and test those changes too.
If some piece of memory should be reused, turn them to variables (or const variables).
If some piece of operations should be reused, turn them to functions.
If they happened together, then turn them to classes.

Offline headkase

  • Almost regular
  • **
  • Posts: 159
Re: Strange cursor position with wx3.0.2 (scintilla)
« Reply #33 on: December 29, 2015, 01:01:56 am »
I need more comments about those patches, especially Morten, because he is the maintainer of that scintilla branch. You can apply those patches locally in your computer, and test those changes too.

I tested the patch.  Built wxWidgets 3.0.2 64-bit, applied your patch to SVN 10630, compiled C::B wx30_64.  Resulting build: opened main.cpp from the C::B project, scrolled around, removed and added text, over the course of five minutes testing everything worked.  Compiler used was TDM-GCC 5.1.0 64-bit.

Offline headkase

  • Almost regular
  • **
  • Posts: 159
Re: Strange cursor position with wx3.0.2 (scintilla)
« Reply #34 on: December 29, 2015, 03:57:33 am »
Ok, I did a sanity test.  I know that using MinGW-Builds GCC 5.2.0 / posix threads / seh exceptions / Revision 1 / 64-bit I had the cursor errors.  So, I recompiled wxWidgets 3.0.2 and Code::Blocks SVN 10630 with the patch with that toolchain just to make sure the issue was also fixed using that.  It is.  The patch corrects the errors.

If anyone would like to try the binaries I built in a different environment, their machine, you can find them: Here.