Description of problem:
calibri fonts are misconfigured
Version-Release number of selected component (if applicable):
How reproducible: every time
Steps to Reproduce:
1. open a libreOffice document (spreadsheet, text document)
2. change font to Calibri 11 pt
3. type 'abc'
Actual results: the display will read JMN
Expected results: the display will read abc
Additional info: only point size 11 is affected; if one selects the text and changes font size to 10.5 pt or 12pt or (I think) any other point size, the underlying text displays correctly as 'abc'
With the "real" Calibri or with the default replace of Carlito.
(i.e. what's the output of fc-match Calibri)
I have to say that I can't reproduce this on two F-20 boxes, one with the true Calibri installed and one with Carlito available as a replacement
Carlito-Regular.ttf: "Carlito" "Regular"
So, your problem is with Carlito. i.e. if you change the font to Carlito does the problem persist?
I don't see any problem like this, and this should be a super common experience, so there is something unusual at your end.
I can't reproduce this on F20 system. Please check again if this really is a problem.
I cannot reproduce this either. Are you absolutely sure that your fonts are intact? What happens if you do the same thing in Abiword?
I cannot reproduce this behavior now, either. I noticed it in a spreadsheet I received from Windows; it was unreadable, and I tracked the reason down to the simple case I reported. I could reproduce the buggy behavior repeatedly then, but it seems to work OK now, including the spreadsheet that showed the bug originally.
I had LibreOffice loaded into memory--it is quite possible that I did not reload it while trying the bug, so it could have been a corruption of the font in the memory of a running application.
I can confirm that this looks like in-memory font corruption.
I usually keep several docs open for several days, and it just happened to me again that incorrect glyphs are being displayed. Closing and opening the file doesn't help---I have to close all instances of oocalc. The newly run spreadsheet displays the strings correctly.
Should I reopen this bug or just let it remain closed because this is not a repeatable problem in the new formulation--I don't know what causes the corruption.