[ft-devel] ttfautohint: broken components when ARGS_ARE_XY_VALUES is false

2016-08-05 Thread Denis Jacquerye
Hi,

When composite glyphs have components with ARGS_ARE_XY_VALUES false, i.e.
arg1 and arg2 are matched point indeces, `ttfautohint --composites`
incorrectly increments the second matched point index. It should only
change the second matched point if the component glyph has changed point
indeces instead of all the time.


For example, when Molengo-Reular.ttf (from
https://github.com/moyogo/molengo-fonts/releases/download/v0.10/Molengo-fonts-0.10.tar.gz)
goes through ttfautohint --composites, the composite glyphs are broken.
The gravecomb glyph does not have any point index changes. However it is
used as a component in Agrave, while ttfautohint correctly increments the
first matched point, i.e. the point index to match on the whole composite
Agrave, it also incorrectly increments the second matched point index
(arg2), i.e. the point index being matched on gravecomb. The first point
needs to be incremented since a '.ttfautohint' component with one point is
added as the first component. But the second point should not since no
point index where changed in gravecomb.

Cheers,
___
Freetype-devel mailing list
Freetype-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/freetype-devel


Re: [ft-devel] Lots of over-/undershoot in H&FJ "Mercury SSm A" at 17.6px

2014-05-10 Thread Denis Jacquerye
On Sat, May 10, 2014 at 6:06 PM, Werner LEMBERG  wrote:

>
> > I'm not sure... it displays fine on a Mac.
>
> Maybe this very font is installed on your computer...
>
> > I found something better, have a look at [...]
> > http://www.typography.com/fonts/mercury-text/webfonts/mercuryssm-book.
> The
> > example text at 18px demonstrates what I reported, other sizes look
> > fine.
>
> It only *accidentally* looks fine!  Attached is an image that shows
> what the WOFF file of mercuryssm-book on this web page actually
> contains (using the Unicode cmap).  The WOFF file is located in
>
>   http://cloud.typography.com/759638/73452/css/fonts.css,
>
> referenced in the HTML source of the web page.  As you can see, it
> misses *many* glyphs, and at 18px the difference to your default serif
> font is large enough that the hinting process delivers a different
> x height.
>
> I guess that those omissions are fully intentional so that you can't
> simply grab the font without paying.  However, I believe it is an
> error of www.typography.com that the glyph reportoire of the web font
> isn't properly synchronized with the displayed text.
>
> My conclusion: No problem with FreeType :-)
>
>
Two fonts are being loaded in the CSS: 'Mercury SSm A', 'Mercury SSm B',
then fallback fonts are defined.
The two fonts each have missing characters, but put together they make a
full set.
This occurs whether you select Mercury, Sentinel or Witney. One font has
only EHINXdorsuwz12356890 etc., the other is missing those characters.
It’s possible the two different fonts are hinted differently.


-- 
Denis Moyogo Jacquerye
___
Freetype-devel mailing list
Freetype-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/freetype-devel


Re: [ft-devel] New Microsoft size-specific design selection mechanism

2013-10-22 Thread Denis Jacquerye
The reason given for not using the 'size' feature is the lack of
implementation and the fact that 'size' is a GPOS feature, not the
best place for this.

On Tue, Oct 22, 2013 at 7:52 PM, James Cloos  wrote:
>> "BE" == Behdad Esfahbod  writes:
>
> BE> 
> http://typedrawers.com/discussion/470/new-microsoft-size-specific-design-selection-mechanism
>
> BE> Would be nice to add support for this.  I like to use it in fontconfig.
>
> They couldn't just use the existing size feature, eh?
>
> NIH syndrome can be a pain.
>
> -JimC
> --
> James Cloos  OpenPGP: 1024D/ED7DAEA6
>
> ___
> Freetype-devel mailing list
> Freetype-devel@nongnu.org
> https://lists.nongnu.org/mailman/listinfo/freetype-devel



-- 
Denis Moyogo Jacquerye
African Network for Localisation http://www.africanlocalisation.net/
Nkótá ya Kongó míbalé --- http://info-langues-congo.1sd.org/
DejaVu fonts --- http://www.dejavu-fonts.org/

___
Freetype-devel mailing list
Freetype-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/freetype-devel


Re: [ft-devel] Latin, Cyrillic, and Greek fonts questions

2013-09-22 Thread Denis Jacquerye
Hi,

Some designers have Cyrillic with slightly taller x-height (sometimes by a
minute amount) to adjust the squareness of the script relative to the
roundness of Latin, even glyphs that are usually components have adjusted
outlines. Other designers recommend components but different spacing
(slightly larger side bearings) instead of x-height to address the issue.
On 22 Sep 2013 08:47, "Dmitry Timoshkov"  wrote:

> Hi Werner,
>
> Werner LEMBERG  wrote:
>
> > I have two questions:
> >
> >   1.  Do Cyrillic or Greek outline fonts exist which don't contain the
> >   ASCII characters a-z, A-Z, and 0-9?
>
> I can speek only about Cyrillic here, but I haven't seen any cyrillic font
> without latin letters and 0-9 digits. It simply doesn't make sence since
> often texts have mixed cyrillic/latin-based context.
>
> >   2.  Do you know of any Cyrillic or Greek fonts where the lowercase
> >   and uppercase glyph heights differ from the heights of the Latin
> >   glyphs?  Given that all three alphabets share e.g. characters
> >   `A' and `o', this rather sounds implausible, but who knows...
>
> Again regarding Cyrillic that's not the case, and very often for instance
> truetype fonts (Arial is one of them) have cyrillic glyphs as a composite
> of the existing latin ones with a transform/composition applied: an example
> is: Я -> R with a mirroring transform attached.
>
> --
> Dmitry.
>
> ___
> Freetype-devel mailing list
> Freetype-devel@nongnu.org
> https://lists.nongnu.org/mailman/listinfo/freetype-devel
>
___
Freetype-devel mailing list
Freetype-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/freetype-devel


[ft-devel] 2.3.5 TrueType hinting bug

2007-09-03 Thread Denis Jacquerye
Hi,

There seems to be a bug due to 2.2 to 2.3 changes in TrueType hinting.
See the example aring U+00E5 in DejaVu Sans Book [1] where the ring
above isn't properly rendered at some size with freetype 2.3.5. The
problem seem to occur wiht some MDRP instructions.
This also happens with a few other fonts.

[1] http://prdownloads.sourceforge.net/dejavu/dejavu-sans-ttf-2.19.zip?download

Cheers,
Denis Moyogo Jacquerye
<><>___
Freetype-devel mailing list
Freetype-devel@nongnu.org
http://lists.nongnu.org/mailman/listinfo/freetype-devel