Fleuron (typography) | Hacker News Reader

Fleuron (typography)

en.wikipedia.org

74 pointssurprisetalk37 comments

In LaTeX you have the pgfornament package [1] that allow you to do nice things.

[1](pdf): https://ctan.math.illinois.edu/macros/latex/contrib/tkz/pgfo...

Is it just me who doesn't like fleurons? For me, they are in an awkward limbo between abstract symbology and realistic illustrations, and draw an unreasonable amount of attention to themselves compared to other forms of decoration.

Generally, I would prefer that whitespace alone be used to separate paragraphs, and more significant sections be demarcated by headings, which can be embellished less intrusively by use of a decorative typeface, underlining or colour.

I find them pretty anachronistic on the web, but I love them in a tastefully typeset book.

They really need to match the typeface though -- a Garamond asks for very old-style designs. But if you're using Times New Roman a short simple geometric line will do much better. Whereas with Bodoni, a trio of big old asterisks feels more appropriate.

Maybe Unicode should add a new code block for these fleurons. They'd all be combining-characters, and you would build effectively a mini-DSL for describing fleurons as geometric compositions of radicals.

So like https://en.wikipedia.org/wiki/Ideographic_Description_Charac... , but instead of describing the compositions, they actually perform them?

Unicode doesn't perform it, but I understand it's straightforward to typeset it in, for example, TeX/MetaPost:

https://news.ycombinator.com/item?id=36809624 ("Typesetting Rare Chinese Characters in LATEX")

Oh wow that’s interesting. Would be interesting if it actually performed the change, and on any character. Pointing emojis in the right direction with the Horizontal Reflection modifier would probably be useful!

That would totally break the basic conceptions of line height and font size though. The fleuron in the top image is probably five lines tall.

Arranging fleurons is great, but that seems it should belong to the realm of word processing layout or drawing program arrangement, not something at the Unicode/font level.

Complex fleurons aren't graphemes; they're illustration.

> That would totally break the basic conceptions of line height and font size though

So do [combining characters], yet they're valid Unicode.

Please don't zalgo on HN. It makes the thread less readable and thus the discussion less interesting.

Edit: I've scrubbed it out of your comment and put the formerly zalgoed text in [square brackets].

For those not familiar with the term "zalgo": https://en.wikipedia.org/wiki/Zalgo_text

No they don't. That hasn't changed the line height at all.

And obviously those combinations don't exist in any actual writing system. It's a (funny) abuse of the Unicode design, not the intention of the design.

You did write "break the basic conceptions of", not "change". To me, a line overlapping with the previous two lines seems to do exactly that.

Also, the original suggestion was for Unicode to add ornamental combining characters. I don't see how that would require any breaking change. You can already stack them vertically and horizontally.

That vertical-stacking combining pattern is similar to how Tibetan script actually works (on a smaller scale):

https://en.wikipedia.org/wiki/Tibetan_script?useskin=vector#...

It's pretty much on the Unicode level. Its goal is to support world's writing systems - current, extinct, and even fictional.

> Complex fleurons aren't graphemes; they're illustration.

The line between writing systems and illustrations is pretty blurred, if you look closely enough. Especially now, with composable and even AI-generated emoji. (I wonder if they regret including emoji now...)

Any complex combination or arrangement of fleurons or other type ornament is not really possible with any contemporary word processor for one good reason: type is not set by hand anymore. Precise but free placement of glyphs is tedious if not impossible. It would require a whole new paradigm of setting type, one that would be based on modularity and strict typographic measurement system. A digital letterpress system if you will. The closest we've had that digitally was probably textmode and ASCII art.

I ADORE this idea.

I wonder how far one could take it with just ligatures?

> I wonder how far one could take it with just ligatures?

Very, very far...

https://blog.erk.dev/posts/anifont/

https://www.youtube.com/watch?v=GF2sn2DXjlA

Font shapers would definitely have one of the moments of all time with this one.

French version of Canadian national anthem (O Canada), first verse:

  Ô Canada!
  Terre de nos aïeux,
  Ton front est ceint de fleurons glorieux!
                         ^^^^^^^^

https://www.canada.ca/en/canadian-heritage/services/anthems-...

I don't get it.

> glyph, used either as a punctuation mark or as an ornament for typographic compositions.

The gallery contains only examples of general flowery decoration, but not the much more interesting use case of these glyphs as punctuation marks. I can't find any such example googling either. Or am I misunderstanding what punctuation means in this context?

They mention using it as a punctuation mark, but surely it must be distinct from other punctuation marks to deserve its own unicode point, further they must differ from each other to deserve their own unicode points.

Surely this deserves to be on the font rendering side, not the unicode side?

I've seen fleurons used in blog posts as end of article mark. It's sort of a generic sign-off / chop, stuck inline to the last paragraph of the text. Usually, when used this way, they'll be styled in a different, lighter color than the text before them.

Obviously, an "end of article" marker isn't a thing that needs to exist — you know the article is over because there's no more article! But people putting extra "stuff" inline at the end of the last paragraph of an article, has a long history in printing anyway. Because people can be extra, and I suppose Unicode should support encoding the extravagancies of existing historical documents.

It's not always obvious the article is over these days. Sometimes they run sneaky ads afterwards.

Ok, an end of article marker is fine.

But it doesn't mean we need multiple code points meaning the same thing.

An a is an a regardless of whether it serif or sans, whether it has the thing on the top like a floppy d (edit: the unfortunate comparison wasn't intentional).

They probably “deserve” to have unicode code points because they existed in pre-unicode character sets.

For an interesting sense of "backwards compatibility" with the planet's past encodings, Unicode decided in concert with Microsoft to encode all of Windows' classic Wingdings 1/2/3 in various places in the Unicode planes. Some of them combined with Variation Selector 16 now double as emoji, so there's an interesting through line between ancient Wingdings and modern emoji. But also yes, some of Wingdings was far more decorative than most of what Unicode prefers to encode, it only gets a pass for how ubiquitous Wingdings is/was.

Looks like quite a few of these arrived with the addition of the historical Manichaean script in which “there are several punctuation-marks to indicate headlines, page-divisions, sentence-divisions, and others”.

The term "fleuron" describes something more like the term "emoji" than it does a particular punctution mark. Yes, fleurons are often used for ending sentences. But they function more as a decoration that happens to go there than as a period or question mark.

Sure, I can imagine a single emoji in all of unicode, just pick a different font for whatever particular one you want to use. But much easier and more logical to have a wide variety of them available to you that roughly matches your current font.

A smiley emoji has a distinct meaning to sad emoji. So each should have its own code point. If you want your emoji to look should be left to the font.

Do fleurons have distinct meanings?

Do all the block elements and geometric shapes have their own distinct meanings, or do they just happen to be useful for different situations?

https://www.w3.org/TR/xml-entity-names/025.html

The custom of a font including several fleurons has existed for centuries, dating almost to the invention of fonts themselves.

I think it's mostly to handle backwards compatibility with a handful of common pre-existing fonts including Symbol and Zapf Dingbats.

Because of those fonts, these particular marks already had widespread digital usage that needed to be maintained.

So it's basically a flower emoji?

Clearly we need a kisrhombille-tessellation based fleuron building system.

only the first three render properly in firefox / linux

Only the first three are covered by at least one of the fonts on your system. If you want the rest to render, install more fonts.

My Firefox on my Linux uses DejaVu Sans for the first three, Noto Sans Palmyrene for the next two, Noto Sans Manichaean for the one after that and Noto Sans Symbols 2 for the rest.

Most popular Linux distros have one or two packages bundling all Noto fonts.

shouldn't have to. it should just work.

Reply on news.ycombinator.com