# [NTG-context] (fake) bold as font feature

Hans Hagen j.hagen at xs4all.nl
Wed Mar 28 12:29:54 CEST 2018

```On 3/28/2018 11:32 AM, luigi scarso wrote:
> On Mon, Mar 26, 2018 at 11:02 PM, Pablo Rodriguez <oinos at gmx.es> wrote:
>> On 03/26/2018 07:24 PM, Hans Hagen wrote:
>>> On 3/26/2018 4:55 PM, Hans Hagen wrote:
>>>> On 3/26/2018 1:40 PM, Pablo Rodriguez wrote:
>>>>> Hans,
>>>>>
>>>>> would it be possible to add a embold feature (similar to the slanted
>>>>> font feature) that fakes the bold font with
>>>>> \starteffect[both]...\stopeffect and it doesn’t strecht the font?
>>>>
>>>> much is possible if i can motivate myself
>>
>> Good to know in order to propose new features :-).
> Hm.
> Fake bold / slanted / small caps /<what ever> of a base font
> are dangerous, there is an high risk to compromise the
> legibility of the text (this is especially true nowadays with user
> consumer tablets  10" full hd ).
> A font is not only glyphs, but also kernings and  hints and all these
> things must match.
> As Knuth  shown , the right way is to design from the beginning how
> each glyph should be transformed
> instead of apply to all the glyphs  the same transformation  (which is
> faster, this is sure).
> This is the key concept of metafont --- in my opinion better than the
> variable fonts .
> I am not saying that this impossible to do in context-mkiv ( *it is*
> possible: just see  5.3 Virtual fonts of the luatex reference)
> but in the end the result is the same of write the correct
> bold/slanted/small caps/<whatever>  version of the  base font .
it's anyway not true to say that 'boldening a font' is a bold font ..
it's more like: hey, we can do the same as all these font vendors,
provide 25 weights of the same base regular (we can't go thinner btw)

a boldened latin modern regular for instance looks like a real 'modern'
latin modern and looks way different from the metafontish produced bold
latin modern (i actually like the boldened more in this case so we can
have a "modern latin modern" typescript ... of course math is a
different issue but even that can be made acceptable

it anway makes a nice topic / presentation for the ctx meeting

Hans

-----------------------------------------------------------------