Once again AI export of radial gradients causes problems for
Chrome. What should have been a white-to-transparent gradient
became a white-to-black 90%-to-100% opaque gradient. So
hand-edited the SVG to remove the black orb floating in front
of their faces... pity.
Since this was done using guesswork these are not an exact match
for the png versions, but they seem close enough.
Adobe Illustrator generated these, and displays them properly, but
neither Chrome (Mac, Linux) nor Firefox (Linux) do. MacOS finder has
even more issues. Both of these used a radial gradient with a
negative 'fx' value, for what that's worth, but it appears the proper
value is a (significantly larger) positive value.
Chrome and Firefox are now happy. Don't expect this fixes MacOS, which
wasn't showing the gradient at all.
These were generated from source, then svg_cleaner was run on
them, and then scour_svg was run on those.
As with the png update there are a lot of small modifications that
don't represent real changes in the images. This version of scour is
0.37 which might account for some deltas over the previous versions,
or perhaps svg_cleaner accounted for some of these.
Note that two of these images (1f314, 1f318) fail to display properly
on Chrome, Firefox and MacOS finder, though they do display properly
in Adobe Illustrator (go figure). A later commit will patch these.
Tweaked svg_cleaner a bit to remove display none, add option to
strip whitespace. Stripping viewBox was too aggressive, two of our
emoji have non-default viewBox values, so emit it when x or y
are not zero.
Added script to invoke scour (now must be part of installation).
Note this does not change image files.
- modify emoji_builder so we can use a flag to control whether to use
big or small glyph metrics. android has legacy tooling that expects
small metrics. we leave the default in emoji_builder to use big metrics
but in the makefile default the flag we pass to request small metrics.
eventually this should shake out to big metrics.
- support more flags by default (not all yet). Note we omit some in the
Makefile that we use aliases for-- we don't need their images.
- add aliases for some new flags
- remove new flags from 'unknown flag aliases' list
- unalias some emoji we will have custom images for
- bump version number
- add annnotations file for u11 new and changed emoji
- use existing utilities in nototools/unicode_data, add_aliases
- add check that file names do not use presentation selectors
- include tags in valid cps that can appear in a sequence
- add check for valid tag sequences (for subregion flags)
- separate out check that no source for an alias is present (we
expect to alias this so should not have an image with that name)
- filter data by age (somewhat), provide command line flag, remove
hard-coded unicode 9.0 value
- separate coverage check (for when data is partial), provide
command line flag and don't run by default
- provide command line flag to exclude subdirs by name when collecting
images
- refactor output so each error has a consistent initial text indicating
the check where the error was found, make output a tad less verbose
Previously forgot to update fontRevision when updating the name table
version string. Also, the commit date/id was created before the
actual pushed commits, so were not correct. This fixes that.
The version string remains at 2.004 and now the fontRevision matches.
This updates the svg, png, and html resources from behdad/region-flags.
It also updates the LICENSE (COPYING) and README.md files from there.
Formerly our copy contained only a few subregion flags, but this change
adds them all from upstream.
Also formerly we explicitly used symlinks in our data for flags that
are the same as other country flags. Now however we just use the
upstream data. Note that the emoji_aliases.txt file will still
cause the font to only have one copy of these.
Note this also deletes the copy of IC.png that somehow had found its
way into the region-flags directory.
Use big glyph metrics to support vertical text rendering
Clients who want to identify sequences in text that have image files
would like to use the file names to identify these sequences, and
would like to identify canonical sequences using this method.
Add an option to the add_aliases tool to also create aliases using
the canonical names where they differ from the file names.
Our file naming omits the emoji presentation selector character, so
the image files by default do not include it. This provides a means
to restore that.
Doug Felt