Closed GoogleCodeExporter closed 8 years ago
Checked Ayar2Uni (and Uni2Ayar, though that shouldn't be called) and neither
seems to have a potential for error.
Uni2Zawgyi shouldn't be glitching either, since the other testers reported that
WZ never crashes.
So it's probably something in Burglish. Hmm....
Original comment by seth.h...@gmail.com
on 1 Dec 2010 at 12:31
Added a bunch of index checks, and fixed one (non-crashable) bug.
So now we need to type a bit with Burglish and see... should also check
pat-sint words while we're at it.
Something makes me thing that the ID we use to type each word should be causing
an error too.... maybe we can bounds check that?
Original comment by seth.h...@gmail.com
on 1 Dec 2010 at 1:59
Side-note: If this really did only show up recently, it's probably got
something to do with kinzi.
Original comment by seth.h...@gmail.com
on 1 Dec 2010 at 1:59
Possible reason:
1) chit + 8 + NGAY + right + right + backspace
The problem is that the PS word list is generated for NGA and re-generated for
NGAY.
Then there is something with the index being out of bounds maybe? (Like, if NGA
has 2 words, moving to 3 on NGAY and hitting backspace is an error).
Easier way to trigger it:
1) NGAY right right backspace
It falters on this instruction:
int currReplacementID = (model->getPossibleWords().empty()) ? -1 :
model->getWordCombinations()[actID];
So, it's getting the wrong word entry. IDs are being saved wrong.
Still very frustrating, but not insurmountable.
Original comment by seth.h...@gmail.com
on 12 Dec 2010 at 8:44
Attachments:
The error was that we weren't resetting the selected index on backspace.
Amazingly, I'd just fixed this on_keypress but not on_backspace.
No more crashes~ though time will tell of course.
Original comment by seth.h...@gmail.com
on 12 Dec 2010 at 9:08
Original issue reported on code.google.com by
seth.h...@gmail.com
on 30 Nov 2010 at 9:40