search

googlefonts / googlefonts.github.io

Google Fonts documentation
https://googlefonts.github.io
Apache License 2.0
58 stars 27 forks source link

added SuperCondensed #113

Open RosaWagner opened 1 year ago

RosaWagner commented 1 year ago

SuperCondensed (25) is in the axis registry, so adding it.

vv-monsalve commented 1 year ago

This is weird since that instance is not included in the usWidthClass. We've seen that many things were not ideal with the first registers for axis, so I would rather validate if we want to keep this instance in the Axis Registry (no family uses that position in the catalog, then it could be deleted from the registry) before making it official here in the Guide.

This would make a case to add "HariLine and ExtraBlack" in the weight axis.

I can ask at the next Axis Reg meeting.

RosaWagner commented 1 year ago

Ah I added it here cause Linefont has a position at 25 that Fontbakery was reporting as a FAIL ^^'

vv-monsalve commented 1 year ago

Yeah, because it follows the Axis Reg, but the last is not following the Spec, on which we declare, here and in the registry itself, that we base our instances. We haven't allowed adding instances like "ExtraBlack" to the wght axis so far, so it would be a good time to review/validate this.

davelab6 commented 1 year ago

https://fonts.google.com/variablefonts?vfaxis=wdth&vfquery=flex has a wdth min of 25. This looks good to merge to me.

vv-monsalve commented 1 year ago

That would be to officialize an inconsistency. In one of our last Axis Registry meetings, Evan said that adding that instance to the registry is another case of the old mistakes around fallbacks as they were initiated. And that he would be willing to remove this instance from the axis registry if it doesn't follow the spec and if it currently is only used by one font (before having added cases).

We agreed to follow up on this in another meeting in sync with the possibility of alternatively adding the "HairLine" and "ExtraBlack" instances in the axis registry. For consistency's sake, we either include only the instances defined in the spec or allow the additional ones for the weight axis.