I like this, in that even if it fails to display correctly you'll just get a box or a missing symbol or and equals sign; and all are much better failure modes than something that looks like a minus sign.

But I am not a GUI wizard.

Is falling back to = something that might happen? Also, we need to check the various platforms' default fonts (OS X 10.10, OS X <10.10, Win7, Win8, maybe Unity default, possibly others?) to see if this looks right.

@Michagogo falling back? I don't think so... being visually indistinguishable to? probably, but thats not so severe a problem as the 'negative' values.

OSX 10.10
Before:

After:

Win7:

Are there any OS without correct default font utf8 support with a mentionable amount of market share?

Tested ACK.

Just another thought:

why not s/~/aprox.?

May want to also find someone on Win8+ and an earlier OS X, since I'm
pretty sure 10.10 changed the system font.

On Tuesday, January 13, 2015, Jonas Schnelli [email protected]
wrote:

Just another thought:

why not s/~/aprox.?

—
Reply to this email directly or view it on GitHub
#5651 (comment).

@zander \u is c++11 so we can't use it yet. The current formulation makes no assumption about the compiler, and just creates a UTF-8 string in bytes form. The end-result of \u in memory would be the same.

As for the client we explicitly force all string converstions to be UTF-8 at the start of the program (which is the default for Qt5 higher):

#if QT_VERSION < 0x050000
    // Internal string conversion is all UTF-8
    QTextCodec::setCodecForTr(QTextCodec::codecForName("UTF-8"));
    QTextCodec::setCodecForCStrings(QTextCodec::codecForTr());
#endif

That said, QChar(0x2243) looks a bit more readable so I'd prefer that.

Tested ACK anyway.

utACK

@Diapolo it'd help moving this forward if you test on windows

@laanwj Done, looks good. Tested ACK! Btw. can you help me with pushing the payment request unit tests?

Looks like on Windows 7, it does show properly (0.10.0rc3 with this cherry-picked on top):

I think I have a Windows 10 Preview VM, if so I'll try installing it there too.

Someone mind checking this in some of the more difficult languages? I believe traditional chinese is one that usually ends up causing us trouble with exotic chars like these.

Looks fine in zh_CN on Windows 7:

@Michagogo thanks for testing.

Should still be checked on OS X and the various Linux environments, though.

I did test on OSX10.10 and 10.9 (comparable to Win8 and Win7 in case of actuality):
See osx10.10 screenshot above and here also comes the 10.9 screenshot:

Verified looks ok in Ubuntu with qt4 and qt5. Kde/kwin.

@theuni Uh, I thought Ubuntu is Unity?

On Tue, Jan 13, 2015 at 11:09 PM, Cory Fields [email protected]
wrote:

Verified looks ok in Ubuntu with qt4 and qt5. Kde/kwin.

—
Reply to this email directly or view it on GitHub
#5651 (comment).

Only by default.

We should also test this on OSX over a gitian build.

@jonasschnelli Right - let's postpone this until the MacOSX unicode issues are fixed.

Prebuilt (gitian) binaries for testing are here: https://bitcoin.jonasschnelli.ch/pulls/5651/

Tested over gitian on OSX and windows.

ACK