Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

unicode-org/text-rendering-tests could be useful? #1332

Closed
davelab6 opened this issue May 9, 2017 · 3 comments
Closed

unicode-org/text-rendering-tests could be useful? #1332

davelab6 opened this issue May 9, 2017 · 3 comments
Assignees

Comments

@davelab6
Copy link
Contributor

davelab6 commented May 9, 2017

Rod suggested unicode-org/text-rendering-tests (eg https://rawgit.com/unicode-org/text-rendering-tests/master/reports/fontkit.html) could be relevant to font-onboarding. Any suggestions @graphicore @m4rc1e ? :)

@davelab6 davelab6 added this to the 0.3.0 Collection Dashboard Web UI milestone May 9, 2017
@felipesanches
Copy link
Collaborator

I found it weird that all expected/observed SVGs in that page seem to be flipped upside-down...

@felipesanches
Copy link
Collaborator

The link that @davelab6 posted is the generated results page. The actual testsuite project is available at https://github.com/unicode-org/text-rendering-tests

@rsheeter
Copy link

rsheeter commented May 9, 2017

(posting some comments I made internally here too)

We have some degree of testing for things like "fsSelection is wrong" but less on the rendering front. I might imagine a set of tests made up of things to render + indication of what the expected result is / what to look out for in a broken font. Bonus points for training a model to recognize violations automatically :D

To try to spitball an actual example, they have stuff like "If your implementation is correct, each rendered glyph should show ...". We might have stuff like "If your font is correct the rendered glyph(s) should have ". Thinking "aloud", maybe for pairs that are usually kerned we render those pairs, a warning if there is no kerning rule, and a note about how and why the pair is typically kerned.

We might warn if things that are usually whitespace draw ink [a real problem we've seen]. We might warn if things that usually draw ink don't.

If we know what the previous version is, which we often do for Google Fonts, we might warn/highlight for manual review any glyph that changed a lot or perhaps if the rendering of any of a set of test strings has changed a lot.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

4 participants