It was the truncation, good call :)
I've changed it to round when converting from 26.6 to pixel space, and
that's fixed the remaining issues I was seeing.
Thanks,
Jamie.
___
HarfBuzz mailing list
HarfBuzz@lists.freedesktop.org
I've continued to investigate this, and it seems to be related to my call
to hb_font_set_scale. The value I was passing was a float multiplier for
the font (you can see it being used in my first email) - I've tried
changing that to use the up-scaled value that we give to FT_Set_Transform,
however
My guess is that you are truncating advances as returned by HarfBuzz whereas
perhaps you need to round them. That explains 99% of issues like the ones you
are seeing...
On 15-10-12 12:17 PM, Jamie Dale wrote:
> I've continued to investigate this, and it seems to be related to my call
> to
Hmm, that's possible...
All the values we give to our font renderer from FreeType and HarfBuzz are
pre-converted into pixel space. For our old font rendering code, this meant
that the glyph metrics and kerning would have been converted into pixel
space separately before being summed together to
On Tue, 6 Oct 2015 18:52:11 +0100
Jamie Dale wrote:
> 2) What is OT::Sanitizer doing, and could it be skipped without
> leading to crashes?
My understanding is that its job is to prevent crashes and other data
corruption. A subtable specified by an offset can
On 06.10.2015 0:27, Behdad Esfahbod wrote:
The version of the font from Jamie *does* have a kern table.
Ooh, sorry for the noise then.
___
HarfBuzz mailing list
HarfBuzz@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/harfbuzz
Did you enable 'kern' feature? Something like that:
const hb_feature_t features[1] = {
{ HB_TAG('k','e','r','n'), !!kerningEnabled, 0, uint(-1) }
};
const int num_features = 1;
bool shapedOk = hb_shape_full(hb_font, buffer,
On 05.10.2015 22:32, Jamie Dale wrote:
The images below are using the Roboto font, and show the output from
each implementation.
The top image is using our "kerning only" implementation, and the bottom
image is using HarfBuzz. Note that the "T" and "e" characters are
further apart in the bottom
I wasn't aware such a thing existed. I've tried turning it on as per your
example, however it doesn't seem to have made any difference.
-Jamie.
On 5 October 2015 at 20:45, Konstantin Ritt wrote:
> Did you enable 'kern' feature? Something like that:
>
> const
I don't know what font data our "kerning only" implementation is using
under the hood, but the kerning used when calculating the advance is from a
call to FT_Get_Kerning with the FT_KERNING_DEFAULT flag. We only do this
when FT_HAS_KERNING reports that the font has kerning data, and the
remainder
I'm not sure what's wrong. Can you send me the font file?
Also, I'm interested to hear more about your caching and load-flags stuff. I
like to upstream them to the extent that makes sense.
behdad
On 15-10-05 03:32 PM, Jamie Dale wrote:
> Hey all,
>
> In our application we have two text
The default is on, so that's irrelevant.
On 15-10-05 04:07 PM, Jamie Dale wrote:
> I wasn't aware such a thing existed. I've tried turning it on as per your
> example, however it doesn't seem to have made any difference.
>
> -Jamie.
>
> On 5 October 2015 at 20:45, Konstantin Ritt
Thanks Jamie,
For loadflags, would it be enough to add hb_ft_font_set_flags() that sets
loadflags, and use that in all places? I'll go ahead and do that, since it's
been long overdue.
Re caching, do you have data you can share on which parts are the slowest,
etc? Would you be willing to test a
hb_ft_font_set_flags sounds basically like what I'm already doing using
user-data, so yeah, that sounds like it would be fine.
I'll see if I can get you some stats (I can't remember off the top of my
head), but I'd certainly be willing to help try out caching patches to
hb-ft.
The main reason we
Yeah, we actually use a slightly older version of the Roboto font.
Here's what I get using the non-GPOS version:
[image: Inline images 1]
Things are a lot more squished than with our old code, but the "T" "e" part
might actually be the same (it's kinda hard to tell though).
-Jamie.
On 5
On 05.10.2015 23:17, Jamie Dale wrote:
I don't know what font data our "kerning only" implementation is using
under the hood, but the kerning used when calculating the advance is
from a call to FT_Get_Kerning with the FT_KERNING_DEFAULT flag. We only
do this when FT_HAS_KERNING reports that the
I remember running into this issue when experimenting with harfbuzz-ng
some months ago.
For reference, take a look at commits
d740c8f78bcbe80a8bcd8a9468830a406da37cc0
and
e70f45eb522bcb41388cc218b79bbd6aaecf8050
On Tue, Jul 27, 2010 at 3:55 PM, Jonathan Kew jonat...@jfkew.plus.com wrote:
Hi
17 matches
Mail list logo