💾 Archived View for bbs.geminispace.org › s › Geminispace › 3228 captured on 2023-11-04 at 13:57:33. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-09-28)
-=-=-=-=-=-=-
My lizard brain keeps on thinking of ways to add formatting such as emphasis or underlines to gemtext. I have to keep reminding myself that a lot of this formatting is a crutch for good writing, and only really exists in an ad hoc sense due to the digital publishing revolution. Some guy just invented bold italic underline and we've been just putting them everywhere since. Just like all caps is possible but has no taste, so is text formatting.
2023-07-18 · 4 months ago · 👍 cquenelle, winduptoy, DocEdelgas
Agreed. Text formatting may serve a purpose in mathematical formulas or in comic strips, but overall, it is overrated.
I just bolted some character translation tables onto ox-gemini so I can go 𝘰𝘰𝘰𝘰𝘰𝘰 𝗮𝗮𝗮𝗮𝗵 o̶h̶ ̶n̶o̶!̶ butˢᵘᵖᵉʳˢᶜʳⁱᵖᵗˢ andₛubₛcriₚₜₛ are patchy and I haven’t been bothered to do bold italics yet.
@totroptof Note that using those Unicode math symbols for regular text is unwise. Your messages will be unintelligible via screen readers, and there is no guarantee a particular client/device has the symbols available. Even in the best case, the font appearance does not match the user's preferences.
I'm actually thinking of forcibly filtering them out in Bubble and replacing them with plain Latin characters.
I think there is a good case to be made for supporting italics. Books have been using it for over a hundred years, and you need emphasis sometimes to transcribe speech. For example this line from Three Men in a Boat from 1889:
"No; you get the paper, and the pencil, and the catalogue, and George write down, and I’ll do the work."
--doesn't work without the emphasis on "you".
In a recent post I suggested adding <italics> and <<bold>> and nothing else to Gemtext, but I don't really think I'd use bold. It was mostly to show my thought about the simplest possible markup: by nesting <> instead of using different symbols you could, if you wanted to, support two or even more types of highlight with only one special symbol and so no ambiguity about how special symbols interact.
if we must, I would just use *stars* for emphasis and have them scoped line wise. 1 bit running variable per line which resets at the ends of the line. rendering it is optional for clients (means text clients work). But I'm trying not to think about it. YAGNI.
Something else I'm thinking of is whether this text formatting is overly Anglocentric. If English is the only language which would use these, then it's a sign that maybe it's not the best pattern. Additionally, even in English you get taught lowercase and uppercase letters and punctuation in school, not emphasis.
Agreed on having it reset each line. A few more thoughts here:
Languages other than English is a good point! This paper is quite fun, about translating emphasis:
— https://www.erudit.org/en/journals/meta/2011-v56-n2-meta1821316/1006185ar/
Emphatic italics are more common in English than in other languages because tonic prominence is the preferred means of marking information focus in English, while other languages use purely linguistic devices, such as word order.
Separately: it does seem like some languages might "need" bold and underline in the same way as English "needs" italics, but I didn't find a good summary.
2023-07-19 · 4 months ago
@skyjake: I know there are drawbacks to this non-standard use (although I hadn’t thought about screen readers, which is easily the most troubling to me). Why filter them in Bubble though? Unless it’s just because it’s a particular gripe of yours, which honestly sounds like a good enough reason.
I would filter them because, as a matter of principle, Unicode characters should be used according to their semantic meaning and not their visual appearance.
The appearance may vary wildly on different devices and apps. Relying on any single visual appearance, the one that you're seeing, will cause some percentage of readers to be unable to decipher your message. The screen reader case is just the extreme example of this.
Right, but filtering those characters by deleting them would deny them any useful semantics. Filtering them by replacing them with lookalikes would only be a violation of the principle by assuming semantics from appearance.
As I said:
forcibly filtering them out in Bubble and replacing them with plain Latin characters.
The reasoning being that if you intended them to be used for visually styling italic and bold text, there's a higher likelihood of them being correctly presented for a given reader if they are just shown as the basic corresponding Latin letters.
That would be more difficult to apply to super/subscripts because those have other valid uses, namely footnotes. As a simple heuristic, one could check if a word consists of nothing but super or subcripts it's likely not a footnote.
Software can't read the poster's mind, so it won't be perfect in all cases, but that's not the goal. It just needs to disincentivize the "bad" behavior while improving the outcome on average.
@gyaradong: There might be validity to the notion that a lot of formatting can be a crutch for poor writing, but emphasis should be placed on the “can be” (it would have been much easier to communicate this succinctly with some way to typeset emphasis!). I suspect that if we take it as a rule that use of typeset emphasis is a marker of poor-quality writing then the pool of quality English literature would shrink considerably. I think the use of italics for emphasis is a pretty entrenched feature of standard English at this point (AFAIK more so than the use of all-caps to convey shouting), and it absolutely was a part of my English education.
I had a whole further spiel here, but it was long and rambly and I wasn’t sure how committed I was to defending my points from counter-argument, so I just want to say this: I think tools for expression are good. Obviously not all tools for expression have the same engineering trade-offs, which strikes me as the central question wrt emphasis in Gemtext. Allowing expression to be dictated by strict necessity (re YAGNI) sounds pretty bleak, and similarly wrt personal taste.
I was really reflecting on my own overuse of emphasis, but overall I regret saying it is a crutch. Having said that, I do stand by YAGNI for the simple reason that most handwriting has no way to represent emphasis but it has been a fine way to communicate.
Also, I'm a fan of the minimal set of features to do a thing. That's why I like gemini. I'm not saying emphasis is bloat but a "why not" attitude to features might get us to bloat.
2023-07-20 · 4 months ago
My gut says to look to print books as precedent. I would favor using * for “emphasis” (typography unspecified). Then stop there.
2023-07-21 · 4 months ago
In hand written text I think underline for emphasis is common?
"No, _you_ do the dishes." definitely works ;)
It occurred to me that eith "unspecified emphasis" clients could use the language in the response metadata; italics for English, maybe bold or underline for other languages.
2023-07-21 · 3 months ago