scripted-diff: Use Courier New as tabular figures font in Overview page

[hebasto]

The bitcoin/bitcoin@8d75115 (bitcoin/bitcoin#16432) introduced a monospaced font for tabulated figures.

But the Qt's font matching algorithm does not guarantee that:

• the monospaced font will be chosen at all (see qt: Add privacy to the Overview page bitcoin/bitcoin#16432 (comment), the actually used font on macOS 11.1 is .AppleSystemUIFont which is not monospaced, i.e., QFontInfo::fixedPitch() returns false)
• the chosen monospaced font has expected style and weight (see Choose monospaced font in C++ code rather in *.ui file #87 (comment) "I think the correct monospace font should be bigger and eventually bold")

This change suggests to use Courier New explicitly which is available on Windows and macOS by default.
On Linux systems the Qt substitutes the missed Courier New font (in case it is not installed) quite well. For example, the Liberation Mono is used on my system.

This is an alternative to #87. But personally I still lean to embedding font (#79).

[luke-jr]

NACK forcing a specific font. We have GUIUtil::fixedPitchFont to workaround Qt limitations...

Furthermore, numbers shouldn't need a monospace font at all.
Replacing the spaces with U+2007 and absent decimal point with U+2008 seems like the correct approach there.

[luke-jr]

Replacing the spaces with U+2007 and absent decimal point with U+2008 seems like the correct approach there.

Problem with this approach is that we're masking with '#' which isn't the same width as digits, and there doesn't seem to be a good substitute. Best I could come up with was https://dpaste.com/2H7QFU9BZ

[laanwj]

Wouldn't this just have to be chanaged all over again when we decide to embed the fonts? (#79)

[hebasto]

Wouldn't this just have to be chanaged all over again when we decide to embed the fonts? (#79)

I really hope we'll move with #79 :)

[jarolrod]

Concept 0 on this

Embedding the font as proposed in #79 is the way to go. In embedding the font we can effectively eliminate the current issues surrounding the representation of monospace characters. This change will not directly address the issue where the font is not available, we assume that linux will choose an appropriate monospace font.

My opinion is that we should focus on #79 instead of devoting review time to this

[luke-jr]

@jarolrod #79 is NACK. And even if we embed a font for backward platforms like Windows/macOS, especially NACK on forcing it by neglecting support of non-embedded builds.

[hebasto]

@luke-jr

@jarolrod #79 is NACK. And even if we embed a font for backward platforms like Windows/macOS, especially NACK on forcing it by neglecting support of non-embedded builds.

Hopefully your comment is addressed in #79 (comment)

[hebasto]

Closing in favor of #79.