It would be much easier to check kerning pair adjustment lookups if the positive or zero values were highlighted in Blue. Currently, I copy the kerning pair lookup from the code editor and paste it into notepad, where I do a find and replace to make the positive values obvious. Typically, that might be a dozen or so among about 600.
Green, blue, just about any different color. Gets my vote.
I suppose, if it were possible:
- Green would be suitable for OK (negative)
- Blue would be for dubious (positive)
- Red would be for an error (zero)
What about using the filter edit box:
- to show all positive values
- to show all negative values
0 to show all 0
Hmm. I think color-coding would be best. It’s visual, it doesn’t limit one ot just positive, negative or zero.
And it is called the visual opentype designer ;^)
Not to sully this request thread, but this is semi-related. I would also like that column to remember the width I drag it to.
How long as that been there? That’s very helpful for locating pairs that are known to be problematic, e.g. hyphen or “c”. However, it doesn’t find c.pcap or c.smcp
-
- to show all positive values
-
- to show all negative values (- already used to find pairs with hyphen!)
- 0 to show all zero values
This would work well. I see that colour coding is already used to distinguish classes from glyphs, so maybe colour-coding is not the ideal solution. It would still take a lot of scrolling for fonts without class-based kerning. Is this practical?
- +nn to show all positive values < nn (one could use a large value like 999 to find all)
- -nn to show all negative values > nn (I am only interested in values less than about 20 funits).
- 0 to show all zero values
True, and it is easy to implement.
Consider it done.
Probably quite some time; I can’t remember.
- to show all positive values
- to show all negative values (- already used to find pairs with hyphen!)
- 0 to show all zero values
Actually if the edit box contains an existing glyph name, the listbox is filtered to show all pairs which include that glyph, otherwise it will just show all. We could change it to show none, so you can immediately see the glyph name wasn’t found in the kern pair list.
This would work well. I see that colour coding is already used to distinguish classes from glyphs, so maybe colour-coding is not the ideal solution. It would still take a lot of scrolling for fonts without class-based kerning. Is this practical?
Even though I’m not colorblind, I never like using color that much. That said I think it was my idea to show classes in red…
I’ve implemented the color request, and will now show the class names in red.
- +nn to show all positive values < nn (one could use a large value like 999 to find all)
- -nn to show all negative values > nn (I am only interested in values less than about 20 funits).
- 0 to show all zero values
Implementing the color request is easier, so lets first try that!
Excellent on both accounts (color and the width issue), Erwin!
Mike
This is a great time-saver. I now run Autokern after importing a script from another font, and quickly fix the pairs that need attention. Finding them is now much easier.