And automation does increase reliability, as far as I can tell. While it's still true that computers can't think, they remain pretty good at following well-defined instructions.
@FaheemMitha -- Yes, that's what I mean by "signature". Depending on the size of the paper used (which can be either in the form of sheets or a roll), current practice will result in signatures of multiples of 4 pages; TUGboat signatures are optimally 16 pages; a smaller book format is optimally 32 pages. So the printing quality should be uniform on the "printing side" of an individual signature. (It should be obvious that ink can be applied uniformly to only one side of the paper in one pass.)
@barbarabeeton I understand. Ideally the printing quality should be uniform across all signatures, but I suppose that is difficult to impossible to achieve in practice.
@JosephWright well it nfc decompses to U+03A9 so .. oh exactly what you said there. On the other hand there may be fonts without Greek support generally that have Ohm so perhaps pragmatism says that at least an option would be good...
@JosephWright -- If I understand the request correctly, I think that sticking with U+2126 (even if the shape is equivalent to the symbol in the Greek alphabet) is a good idea. What if this is embedded in an italic context (say a figure caption)? The ohm shouldn't go italic; it should be "fixed" in shape.
@JosephWright certainly with langle and rangle (which got normalized to cjk angle brackets) the glyphs were never the same, with the one at 3008 being double width but the one at 2329 being just like the new 27e8 which was just added to avoid the normalisation
@DavidCarlisle -- That "normalization" of the angle brackets is just plain wrong. I should have fought harder, or persuaded Murray to come up with a counterexample. But I was new at the mechanics of the Unicode committee. Maybe "bra-ket" would have been convincing, if I had known about it at the time.
@barbarabeeton yes Murray basically confirmed they knew it was wrong but didn't want to change the never change normalisation rule so added two new characters instead
@JosephWright -- Well, my academic technical background was applied math and engineering. Other physical sciences use different notation, and learning it requires more exposure than I had at the time I got involved in the STIX project. (I think I have a wider range now.)
@barbarabeeton I can see how it happens: I don't do a lot of theoretical quantum stuff, but I did do the relevant stuff as an undergraduate, and come across it occasionally in background reading, so I know of bra-ket even though I'd had to revise to use it
@barbarabeeton The italic business is OK: I have all of the font control in siunitx so that's independent of the surrounding material in a defined way