[MPEG-OTSPEC] Requesting progress update on COLRv1 in fontTools, FreeType, etc.

William_J_G Overington wjgo_10009 at btinternet.com
Fri Jan 22 18:35:59 CET 2021


Earlier today I wrote as follows.
> On a different aspect of the specification, originally if one used 
> FFFF as the colour index it allowed the colour actually used to be set 
> by the end user as the foreground colour. Does the system now allow/ 
> will you please consider allowing, similar 'tunnel through' colours, 
> so that there would be a feature whereby an end user could also select 
> in an application first decoration colour and second decoration colour 
> so that, say, one could have, for example, a font with some glyphs of 
> holly with berries and the end user could choose to have green leaves 
> with red berries or gold leaves with silver berries or whatever colour 
> combination desired, with the way that the specification is done such 
> that there is a default colour? So, for example, displaying the glyphs 
> in an application that does not have a decoration colour choosing 
> facility, or it is switched off, the holly appears as green leaves 
> with red berries yet the colours can be chosen by the end user in 
> applications where a decoration colour choosing facility is available 
> and switched on.

Thinking further on this, it would appear that there are (at least) 
three aspects to implementing this.

The first is the Specification.

The second is implementation in fontmaking software.

The third is implementation in application software such as a desktop 
publishing program.

I have done some scientific programming, though it was many years ago, 
and my speciality was in providing programming advice to undergraduate 
students so I was reasonably knowledgeable of programming for that 
purpose, but I was not really deeply into things like the advanced parts 
of C++. I have no experience in producing large commercial software 
packages such as fontmaking programs or desktop publishing programs. So, 
within those limitations, and not willing to say that something can be 
done if I cannot do it myself, here are my thoughts on the 
implementations, expressed as if questions which hopefully some 
participants here might answer please.

If my idea of decoration colours were implemented, with, say, decoration 
colour 1 as FFFE default colour green and decoration colour 2 as FFFD 
default colour red, would that be straightforward to implement in a 
fontmaking program please? I am thinking that the colour palette 
dialogue panel in a fontmaking program could have buttons for

dec col 1

and

dec col 2

and they would each show the respective default colour, and that could 
be customized by a font designer for a particular font.

Would such customization be straightforwardly programmable by the person 
producing the fontmaking software package? Is it much the same as 
customizing any other colour in the palette, the only difference being 
that the other colour palette members would have low numbers and these 
two colour palette members would have the fixed numbers as in the 
specification? So to the font, would they just be colours, it is just 
that an application program with decoration colour choosing capability 
available and switched on would recognise them?

In a desktop publishing program, would the software package programmer 
be able to add in the menus for the selection of decoration colours 
fairly straightforwardly by using copies of the software already used 
for selecting text foreground colour? How difficult would it be to 
recognise FFFE and FFFD colour codes in a font and substitute the end 
user selecting colours in the display?

A consideration is whether having FFFE and FFFD colours in a font that 
is used in a COLRv0 system setting would work satisfactorily. Could 
someone who understands how the specification works in detail possibly 
comment on this please?

William Overington

Friday 22 January 2021



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20210122/12cb68aa/attachment.html>


More information about the mpeg-otspec mailing list