[PATCH 1/2] Added a function to check if specific OS features are available

Jef Driesen jefdriesen at telenet.be
Mon Oct 22 10:19:11 PDT 2012


On 22-10-12 18:53, Dirk Hohndel wrote:
> "Lubomir I. Ivanov" <neolit123 at gmail.com> writes:
>
>> On 22 October 2012 14:40, Henrik Brautaset Aronsen
>> <subsurface at henrik.synth.no> wrote:
>>> Den 20.10.12 16:35, skrev Dirk Hohndel:
>>>
>>>> On Oct 20, 2012, at 7:28 AM, Lubomir I. Ivanov wrote:
>>>>
>>>>> On 20 October 2012 08:24, Miika Turkia <miika.turkia at gmail.com> wrote:
>>>>>>
>>>>>> Is there a way to test if the selected font supports the required
>>>>>> character instead of determining this based on the OS version? Would
>>>>>> be more universal approach that way.
>>>>
>>>> In general I agree. But for the last few years the major OSs all have had
>>>> reasonable Unicode fonts bundled.
>>>> So this is mainly an issue with very old OSs. As mentioned, XP is more
>>>> than a decade old.
>>>
>>> Fun fact:  I just discovered that the unicode stars worked for me, out of
>>> the box, in the rc1-localized version in WinXP Pro SP3 (running in
>>> VirtualBox).  In the latest rc3, they don't work.
>
> Yes, we manually turn them off if we detect that you are running XP.
> Fundamentally it appears to be rather unreliable to figure out if a font
> on Windows does have the glyphs that we are looking for. We identified
> several versions of XP (localized and non-localized, SP1 and SP3) that
> don't appear to have the star glyphs. So those people would get odd
> characters instead of the stars. It seemed better to simply give
> everyone on XP (I assume - VERY FEW actual users) the lame "*" version
> than to have it fail badly for some...

How about using an pixbuf instead of utf8 character? Could be done as a 
fallback, or just for everyone...

Jef


More information about the subsurface mailing list