My text face wish list

24

Comments

  • …the text string posted didn’t render the angle brackets (U+2329, U+232A), I had to insert them via the glyph palette—why is that?

  • Nick Shinn
    Nick Shinn Posts: 2,182
    edited April 2021
    Andreas, I don’t understand. Please explain.
  • the string given by H. Wollmersdorfer (April 26, see above) contains U+27E8 and U+27E9 for the angled brackets. Apparently your brackets sit in the U+2329 and U+232A positions instead, therefore your glyphs didn’t show up in that text string.
  • Christian Thalmann
    Christian Thalmann Posts: 1,967
    edited April 2021
    Never drew the angled brackets before, but I guess it's not too late to start. :grimace:

    They won't rename themselves from ⟨ and ⟩ in Glyphs — are they too obscure to have a Nice Name...?
    Also: Should my | be shorter...?
  • Bravo, Andreas!
    the italic braces and angled brackets are too dark; 
    A little bit. But the height and vertical alignment is fine and classic style.
  • [...]
    However, the text string posted didn’t render the angle brackets (U+2329, U+232A), I had to insert them via the glyph palette—why is that?
    Which operating system and which application?

    They something do crazy things. Maybe the website of typedrawers or the browser changed the code points.

    Mostly I use TextEdit on Mac because of the best support for OT-features. 
  • Nick Shinn
    Nick Shinn Posts: 2,182
    So I should give the angle bracket glyphs two code points each?
  • They won't rename themselves from ⟨ and ⟩ in Glyphs — are they too obscure to have a Nice Name...?
    Inserting them with Glyph → Add Glyphs… works for me. ⟨ & ⟩ (U+27E8 & U+27E9) have the nice names leftanglebracket-math and rightanglebracket-math and 〈 & 〉 (U+3008 & U+3009) have the nice names anglebracketleft and anglebracketright.
  • konrad ritter
    konrad ritter Posts: 204
    edited April 2021
    Mr Shinn, you're in my book of Very Good People. I absolutely didn't mean to imply you may be less than a decent guy. I had in mind young, barely-out-of-grad school kids in Europe; and various individuals who cannibalize the work of old masters, and regurgitate it at obscene prices. 

    A little-know fact: Mr Carter's Galliard comes with two italics /g/s. Here's a glimpse of the second, better one. 



  • Nick Shinn
    Nick Shinn Posts: 2,182
    edited April 2021
    Thanks Konrad, I was just taking the opportunity for a little font promotion.~
    **
    There was a fellow who used to post a lot on Typophile, Charles, I think, I forget his surname, an old school “compositor”, who encouraged me to put a lot of features in Goodchild, similar to the ones that Joshua likes to see, for book work, serious books with all the bells and whistles. Man, I love that stuff! But I don’t always get around to including it.
  • Possibly Charles Ellertson?
  • Christian: with the obvious caveat that I am not a type designer, I will try to answer the question of whether your vertical bar glyph should be shorter. In general, for serifed text faces, I prefer the bar to be the full ascender/descender height — but I also prefer for the delimiters ()[]{} to be the full ascender/descender height. So, in other words, I would not make the bar shorter; I would make the delimiters taller. I like when nonextending lowercase letters and small caps seem to be vertically centered within the space of the parens or brackets. To me it seems poised and elegant. Others may, of course, disagree.

  • Here is a little experiment done in Štorm's Jannon. I note that the delimiters do descend, although the intention seems to be to split the difference between centering them on the lc height and the cap height, which is understandable. The vertical bar is the same height as the delimiters. However, the virgule is shorter, which surprises me. I would prefer for the virgule to also reach the ascender/descender lines. I like that the delimiters are close to monoline and generously spaced. Obviously this style derives from the Renaissance and would not be appropriate in all typefaces. There are case variant delimiters for small caps, but in many circumstances I think I actually prefer the full-height delimiters, though they might have to be baseline-shifted to center on the small caps. Case variants for all-caps are, of course, essential when the delimiters descend.

    As my typography has come to be slightly more classically influenced, I have increasingly come to think of roman and italic as categories that apply only to letterforms, numerals, and punctuation, not to anything else. In many cases, unless the typography calls for a more contemporary approach, I like to see things like / \ | + = remain upright even in an italic context. Like monoline parens etc, these symbols don't seem to count as letters, and should therefore be somewhat immune to italicization.
  • Nick Shinn
    Nick Shinn Posts: 2,182
    edited April 2021
    A general principle would seem to be that parenthesis style should correspond to the main style of a paragraph.

    For instance, a list of menu items in italic:

    ½ cup (125 mL) sugar
    ¼ tsp (1 mL) pepper
    1 tbsp (15 mL) sesame oil
    etc.

    ½ cup (125 mL) sugar
    ¼ tsp
    (1 mL) pepper
    1 tbsp
    (15 mL) sesame oil
    etc.

    The roman parentheses look wrong here.

    I did put alternate kerned roman parentheses in my Modern Suite italic fonts, accessible via a Stylistic Set. No doubt under the influence of Mr. Ellertson.
  • So, in other words, I would not make the bar shorter; I would make the delimiters taller.
    That seems like a good idea. Since Ysabeau is trying to emulate Garamond in the sans parameter space, I just checked with Garamond Premier and EB Garamond, and sure enough, they have deeply descending delimiters (though they disagree about / and \). I'll give that a try.
    For those who have slashes going all the way to the descender floor: Doesn't that cause kerning problems, e.g. with /g/?
  • Another suggestion: If you produce a font with a nice set of alternates, try to make sure that the glyphs are ordered in some sort of coherent way so that the glyph palette gives a reasonable overview of the font.
  • Another suggestion: If you produce a font with a nice set of alternates, try to make sure that the glyphs are ordered in some sort of coherent way so that the glyph palette gives a reasonable overview of the font.
    That is relevant, especially for me right now that I have to decide?
    Which sorting order would you prefer/suggest?
    So far I more or less kept the "Opentype Standard" encoding (privileging in general the Unicode code points order), and then added the additional accented letters in alphabetical order (Upper and Lowercase), diacritics, additional punctuation, Letterlike forms (™℣℟), currencies, letters from Alternate sets, case sensitive © and ® and numeral sets (including fractions) with other associated glyphs (%, ‰′″), Dingbats (2700 range, inclusing the Angle brackets 276C and 276D), arrows, some historical punctuation (from the 2000 range), combining diacritics (with zero width) and the circledotted (25CC), custom decorative/typographic symbols and Math symbols at the end.
  • I'm not advocating for any particular order -- only that the order make sense. For encoded glyphs, simply sorting by unicode or by encoding works reasonably well. But I've run across lots of fonts where the unencoded glyphs are either in an entirely random order, or are sorted purely alphabetically such that figures and punctuation end up being scattered all over the place.

    I'm simply suggesting that these should be ordered in some coherent way, with alternate figures grouped together in numeric order, alternate punctuation grouped together, and alternate alphabetic forms grouped together, either in purely alphabetic order, or with specific types of alternates (small caps, swash, initials, finals) grouped together and then sorted alphabetically within each group.
  • konrad ritter
    konrad ritter Posts: 204
    I strongly support Mr Isaak's suggestion. It would help a lot of users. 
  • I tried out deeper and more consistent delimiters. Does this work?

  • I tried out deeper and more consistent delimiters. Does this work?

    Considered it’s a "Garamond" they are nice. Better evaluated in text with a lighter weight, maybe, but the height now is better and, I think, closer to what Joshua suggests.
  • Regular weight:

  • Very nice, with little modulation an a bit lighter, it seems they stand out well.
    I added the angle brackets myself, thanks to Joshua’s pointing it out. :)
  • I tried out deeper and more consistent delimiters. Does this work?

    they are all too heavy, compared to capitals and minuscules. Especially the slashes.
    I would not let the slash thickness exceed the horizontal of H, and never let the thicker parts of ( ) etc. get bolder than the stem of h.

  • Like this?



  • Like this?



    Honestly, I thought that in the lighter weight they were already nicely balanced.
    What Andreas suggested, however, improved the overall situation in the heavier one.