by Michael S. Kaplan, published on 2013/03/29 07:01 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2013/03/29/10406219.aspx
The question, as usual, was simple:
We have a problem. We installed the Turkish Lira update, and since then we are seeing a question mark whenever we should be seeing the new Lira in our application.
It looks right in Regional and Language Options. And in Excel.
Did we miss a step here?
Well, it wasn't a "step" per se.
But it is updating the application to support Unicode, so (for example) if you call GetCurrencyFormatEx or GetCurrencyFormatW, you will get a properly formatted string with the currency.
Code page 1254 was registered with IANA on May 3, 1996 (ref).
And code pages cannot be changed - it has been a disaster the few times we tried.
I can think of four workarounds:
Not great, but this kind of change, this kind of progress requires up-to-date software.
Unicode software.
John Cowan on 29 Mar 2013 10:37 AM:
Can you explain why it's so bad to add characters to a code page? After all, Unicode is basically a goloptious big code page, and we add characters to that all the time.
Yuhong Bao on 29 Mar 2013 1:07 PM:
John Cowan: Codepages have limited room with only 128 positions. These codepage tables has been hardcoded into many many programs and libraries, including for example browsers.
Azarien on 30 Mar 2013 12:20 PM:
What is the fourth workaround?
Michael S. Kaplan on 31 Mar 2013 12:56 AM:
Custom locale...
🍟 French fries on 1 Apr 2013 10:41 PM:
I wonder why Babelpad 6.2.0.3 doesn't display the turkish lira sign in composite font mode on Windows 6.1 with the font update installed.
Doug Ewell on 3 Apr 2013 7:09 AM:
@: Looks fine to me on BabelPad. Try Fonts > Composite Font Mappings and see what font BP is using for the Currency Symbols block, and change it if needed.
If you're using Windows Vista (which Microsoft tends not to call "Windows 6.1") then you might also need to update your version of Uniscribe.
Doug Ewell on 3 Apr 2013 7:36 AM:
Sorry, Windows 7 (not Vista) is 6.1.
John Cowan on 10 Apr 2013 9:23 PM:
Yuhong: First, some codepages are much larger than 128 characters: 2-byte codepages are 94 x 94 or 96 x 96 characters, and the rarely used but possible 3-byte codepages are 96 x 96 x 96 characters. Second, not all codepages are full: many have one or more unused codepoints.
Doug Ewell on 11 Apr 2013 12:23 PM:
Speaking of 2-byte code pages with unused code points, I wonder if the DPRK ever added A4EE through A4F0 to KPS 9566. Winks to those who get it, as James Kass once said.