Eventually you'll choose one, but in the meantime you've got five or ten or even twenty /a/'s. Where do you put them? How do you make them easily switchable for previewing and testing purposes?
In Glyphs I use the .ssXX naming similar to the above, the OT feature lets you quickly autocompile the alternates code and you can then use the preview window to type out and test swap by using the OT feature.
Yes but isn't reviewing candidates often the basis for making the decision?
I find that's not how I work. If I occasionally save a version of a glyph as .alt, I find that always indicates that I'm not satisfied with that approach, and I am unlikely to return to it. I tend to view each glyph as a piece of molten metal: I go at it with a hammer until I get the shape that works.
I remember many years ago reading a comment by Lida Lopez Cardozo, talking about the difference between the working method she learned in graphic design — in which making multiple options was the norm — and in stonecutting — in which one works at a single design until it is right. My type design approach is more like the latter, which is one of the reasons why I mostly don't make typefaces with multiple stylistic variants.
Not saying this is how everyone should work, of course.
I usually just add .1, .2, .3, or whatever. Usually, it's not because I'm undecided, but because I want to try a different approach and I want an easy way to go back if it doesn't work out. Sooner or later I delete the earlier "alternate" versions. I almost never go back to them, but you never know. (If I include alternates in a font, it's because it was part of the concept, not because I couldn't make up my mind.)
I line them up laterally in the background layer, promptly make a –provisional– selection which I put in the main layer, but leave them all in the background forever.
I'll throw them in ss01-20... then if I keep them, I already have the features started, just need to do variants (such as accented versions of the same letters).
I have been less disciplined than some folks here about going back and killing alternates before release, but with the result that sometimes I see them in use years later and think “you know, that wasn’t really a great idea” or “mostly good, except I went the wrong way in the treatment on that one glyph” (darn unicase r in Hypatia Sans still bugs me).
So I am now a big fan of going back at the end and doing one last serious re-think of such “extras.” Just make sure you still like them all.
Comments
But if I am going to make a font and test it in a layout application, then I use Paul’s method.
I do something similar. I save my alternates in the PUA, never look at them, and delete them all anyway.
I remember many years ago reading a comment by Lida Lopez Cardozo, talking about the difference between the working method she learned in graphic design — in which making multiple options was the norm — and in stonecutting — in which one works at a single design until it is right. My type design approach is more like the latter, which is one of the reasons why I mostly don't make typefaces with multiple stylistic variants.
Not saying this is how everyone should work, of course.
I have been less disciplined than some folks here about going back and killing alternates before release, but with the result that sometimes I see them in use years later and think “you know, that wasn’t really a great idea” or “mostly good, except I went the wrong way in the treatment on that one glyph” (darn unicase r in Hypatia Sans still bugs me).
So I am now a big fan of going back at the end and doing one last serious re-think of such “extras.” Just make sure you still like them all.