The attached program runs fine on a Red Hat Linux 8 system, but displays a mixture of characters and square boxes on Shrike.
Created attachment 91178 [details] Sample program to trigger utf-8 issue
Reproduced. Wonder if this is related to the move to ucs4 for python.
Indeed it is. Removing tcltk-8.3.5-ucs4-for-py.patch, it displays just fine. Misa, do you have any utf-8 test-cases for tkinter?
In tcltk-8.3.5-90 I have disabled the ucs4 patch for now, since it seems to be breaking utf-8 support. But this is probably not the final solution. At least now "utf-8.wish" displays ok again, though the equivalent(?) Tkinter python script does not.
Created attachment 92758 [details] Tkinter equivalent? to utf-8.wish Actually this script doesn't seem to display correctly with or without the patch. What is the right way to do this with python and Tkinter?
try: ... text=u'...' instead of ... text='...' Works on my FC2 box