Comment on Fonts, and colors, and work skins, oh my!

  1. spqrz Thu 25 Jul 2024 03:03PM UTC

    The reason why your smiley showed as a J was that your Microsoft mailer had turned :-) not into a Unicode smiling face (which is now standard), but into a WingDings-coded smiling face.

    WingDings code was one of the many pre-Unicode things. The WingDings font mapped various symbols onto the same code points as ASCII. So that smiling face literally was a letter J, but with the font set to WingDings.

    Her mailer had then removed the font setting, so she saw the J.

    A blind person with a screen reader would have the same problem: the screen reader reads the underlying text, without regards to whether some font has changed its appearance. That's why these days it's really a good idea to use proper Unicode values for everything: assume somebody somewhere won't be able to display the same font as we have....

    Comment Actions

    1. Charles_Rockafellor Thu 25 Jul 2024 03:23PM UTC

      EDIT: Thank you, spqrz!  (And my apologies: I was so caught up with the plot twist there that I forgot to greet and thank you properly.)

      That explains a bit.  (And I'm now torn between several different emoji here, though at least these would render properly... as long as no one were to read this on any older systems!)

      I know that going from the general to the specific is a bit of an iffy thing at times, but I don't suppose that you'd have any insight as to a similar issue commented on my Green Rain proof of concept re. a Unicode Character “↋” (U+218B) (i.e.: a 180° rotated number 3) being rendered instead as a stick figure / Sherlock Holmes dancing man / Creepypasta Thin Man?

      Last Edited Thu 25 Jul 2024 03:25PM UTC

      Comment Actions

      1. spqrz Thu 25 Jul 2024 04:15PM UTC

        Yes I'm afraid I did end up knowing far too much about character sets :) U+218A and U+218B were added to the Unicode standard in version 8 which was released in 2015. That means any font made before 2015 won't have it, and will display the font's default "no such character" shape.

        And unfortunately neither Apple nor Google has yet fully caught up with Unicode 8 in their fonts, even though we're now 9 years in. (Sigh.)

        (Rant: Google haven't even fixed the issue that's been on their GitHub for years about some of their Braille symbols showing the wrong dots. This affects one of the characters in The M3GAN Files Chapter 3 and it kind-of annoys me that 2 of the dots are too far to the left on Android, even though most readers will probably blast past it without noticing. If I ever find a sighted helper learned Braille wrong because of this I'd be tempted to send Megs to Google HQ to sort them out :)

        You can get an inverted 3 in CSS but it's awkward: you'd need a display:inline-block with zero padding, containing a div with zero padding that can take a transform to rotate or reflect (scale -1) a normal 3. Don't like it.

        But it might be "near enough" to just put in a capital epsilon (U+0190) Ɛ

        Comment Actions

        1. Charles_Rockafellor Thu 25 Jul 2024 04:47PM UTC

          This is why backward compatibility is so important (an IF-THEN tree would help, but also lead to so much cascaded extra crap to track, I know).  <sigh> 😞

          I think that the epsilon is likely the best bet: my diacritics CSS should render it acceptably (normally I'd worry, but the whole Green Rain thing is rather visual, so sighted readers disabling CSS there would be a tad counterproductive), and majuscule epsilon is reasonably uniform in general anyway — it might not exist as-is in any of the canon Green Rain glyphs (don't know), but that's irrelevant for the purpose of a simple Matrix-themed banner.
            I might pursue that transform at some point in the future; I usually try to keep my visual minimal, but just mucking about with it as a toy model for effects in general could be interesting (and exasperating, I suspect...).

          I use a TTS screen reader to doublecheck my typos and the flow, and found that the free ones are mostly crappy (one surprisingly good one, though it was also so feature-heavy that it felt a bit much).  It actually annoyed me enough that, in my series-tutorial (since it starts out with an emergency fix-it for a couple of glitches), I included a subsection on the reader skipping a number of things such as data lists or the rest of a paragraph following a less than symbol, and another subsection on the comparison and contrast of the freebies, for those who want/need to know.

          I don't know if many people (geeks aside, I mean) worry much about these things, and I'm not sure how many people typically pay attention to what the see / hear / feel (as opposed to skimming, catching 10% or less of a paragraph, and considering it to be perfectly sufficient), though actual readers probably pay more attention than social media surfers, but if you were to write up a guide (long or short) on TTS and/or braille reader pitfalls (and/or such issues re. fonts and standards changes over time), I would absolutely link it all over everywhere relevant in my tutorials and metas.

          Last Edited Thu 25 Jul 2024 04:47PM UTC

          Comment Actions