User forums > General (but related to Code::Blocks)
"file updated" dialogue bug
jarro_2783:
I agree with jens.
editormanager.cpp is definitely creating a ConfirmReplaceDlg for that dialog box. If that's using an XRC file then the patch looks alright.
thomas:
Ah, see... I told you that you were talking of a different dialog. I was talking about the one 15 lines earlier (in the same function), which is why I had no idea what the heck you were talking about :)
You're right, this one uses ConfirmReplaceDlg for some reason. Funny... I don't see why that one couldn't be a cbMessageDialog like in the other one.
MortenMacFly:
--- Quote from: thomas on November 07, 2007, 02:07:25 pm ---Funny... I don't see why that one couldn't be a cbMessageDialog like in the other one.
--- End quote ---
Agreed. I think we have discovered a bug or at least very bad design. Thanks for clarifying.
dmoore:
--- Quote from: MortenMacFly on November 07, 2007, 03:48:48 pm ---
--- Quote from: thomas on November 07, 2007, 02:07:25 pm ---Funny... I don't see why that one couldn't be a cbMessageDialog like in the other one.
--- End quote ---
Agreed. I think we have discovered a bug or at least very bad design. Thanks for clarifying.
--- End quote ---
well if you look at the code, the user is given the following options when a file has been modified:
* Yes (Reload)
* No (Don't Reload and make the internal date in the editor the same as the file)
* Cancel (Don't reload and don't change the internal date, break out of the loop)
* All (Reload all changed files)
cbMessageBox can't do that.
The earlier use of cbMessageBox concerns deleted files. Arguably, there are only two reasonable options: Yes - keep the contents of the open editor, No - close the editor.
As someone who has been annoyed by this bug on Linux in that past I'd like to see it fixed. I think Jens' patch does that without breaking anything, so why wouldn't it be accepted?
MortenMacFly:
--- Quote from: dmoore on November 07, 2007, 04:25:11 pm ---As someone who has been annoyed by this bug on Linux in that past I'd like to see it fixed. I think Jens' patch does that without breaking anything, so why wouldn't it be accepted?
--- End quote ---
Because it leaves a pointer to a "bug" that needs to be fixed... ;-) On the other hand... let's see...
Navigation
[0] Message Index
[*] Previous page
Go to full version