Comment by thaumasiotes | Hacker News Reader

Comment by thaumasiotes | Hacker News Reader

Parent Full thread

> As in "We don't have a clear idea what this rune was for, or what it means, but we see it in documents and so added it to Unicode."

Documents? I had the strong impression that there are no documents written in runes. A rune we only know by its occurrence in documents would be far more interesting for the existence of a document than it would be for its own sake!

Compare what the page about Anglo-Saxon runes says about the corpus:

> The Old English and Old Frisian Runic Inscriptions database project at the Catholic University of Eichstätt-Ingolstadt, Germany aims at collecting the genuine corpus of Old English inscriptions containing more than two runes in its paper edition, while the electronic edition aims at including both genuine and doubtful inscriptions down to single-rune inscriptions.

> The corpus of the paper edition encompasses about one hundred objects (including stone slabs, stone crosses, bones, rings, brooches, weapons, urns, a writing tablet, tweezers, a sun-dial,[clarification needed] comb, bracteates, caskets, a font, dishes, and graffiti). The database includes, in addition, 16 inscriptions containing a single rune, several runic coins, and 8 cases of dubious runic characters (runelike signs, possible Latin characters, weathered characters). Comprising fewer than 200 inscriptions, the corpus is slightly larger than that of Continental Elder Futhark (about 80 inscriptions, c. 400–700), but slightly smaller than that of the Scandinavian Elder Futhark (about 260 inscriptions, c. 200–800).

So across every runic system we know, we have under 600 texts, all of those texts are short inscriptions, and even to reach that number of samples we need to include texts that we aren't even sure contain any runes.

> Documents? I had the strong impression that there are no documents written in runes.

One of the original goals of Unicode was to be able to computerize every document. I still have some old linguistics books in which characters have been handwritten into typed or even typeset text. So these are the types of documents being referred to: academic papers.

Some fancy books have photographs of ancient writing; I’m not sure if Unicode tries to encode such sources and I pretty much doubt it (how would you even know what to call the symbols? You touch on this in your comment). However often they are attached to treatises that order the characters in some way (I.e. index an alphabet) in which case the first case above would apply.

In other words: thanks to some scholars who wrote down and ordered runic alphabets, you can now discuss runes with your friends and colleagues through email.

> One of the original goals of Unicode was to be able to computerize every document. I still have some old linguistics books in which characters have been handwritten into typed or even typeset text.

That's a weird goal for Unicode to have. We've already accomplished that; a PDF file does the job better (note: PDF documents already support every character existing in the past, present, or future!) while being less complex.

I don’t understand. If there is no computerized way to represent the script, all you can do would be to include photographs in your pdf. The point of computerization is not simply storage and retrieval (and retrieval is hard if you can’t represent the script) but automated processing, which is meaningless if you can’t represent any semantics).

Separately, PDF felt like a step backwards on the day it was announced and sadly nothing since then has changed that.

How do you search for non-unicode characters in a pdf document?

How do you search for them in a book?

ctrl-F once you have digitized it.

And how do you type the character you are searching for?

I don't think that not having a physical key on the keyboard has ever stopped anybody from inputing unicode symbols.

On Ubuntu: l-ctrl+l-shift+u, <codepoint>, <enter>

Of course, that sucks, so I've programmed a nearby key to act as l-ctrl+l-shift+u.

Several characters can also be typed with Compose Key.

For characters I use regularly (in my case, generally the elder and younger futharks), I've created a keyboard out of an Elgato StreamDeck XL so I can type any of these runes with a single button press.

The question was

> How do you search for *non-unicode* characters in a pdf document

Eh. That was the original question, but I think the implication of some of subsequent questions can be interpreted in multiple ways.

This is interesting. I’m comparing this to how musical notation is encoded in unicode. I mean, there is a block dedicated to the symbols, so the symbols are encoded, but you can’t document music using only unicode. But musical documents are being composed and written all the time. To write music you need an additional software which arranges these symbols in a certain way so that they express the authors intention.

I guess math has a similar representation in unicode as well.

All that said, I think people use runes to express magic and spells (even to this day). I don’t think all the magical runes are expressed in unicode (and perhaps they shouldn’t). If you want to use a rune in that way, you might have to draw it out in SVG or something and then email it to your friends.

> I guess math has a similar representation in unicode as well.

It's an ongoing project. As you seem to have guessed, Unicode math symbols are just about as useless for representing math as Unicode music symbols are for representing music. Producing mathematical documents is done using dedicated software, generally LaTeX.

(And what you get is a PDF, because, as I noted in another comment, PDFs already support every notation there is, was, or ever will be.)

> (And what you get is a PDF, because, as I noted in another comment, PDFs already support every notation there is, was, or ever will be.)

I now have a perverse urge to invent a sculptural writing system just so I can break this completely reasonable claim.

> I had the strong impression that there are no documents written in runes.

There are. Such documents are called runestones and thousands survive to this day, most in Sweden.

https://en.wikipedia.org/wiki/Runestone

Huh. https://en.wikipedia.org/wiki/Document says:

> Documents are also distinguished from "realia", which are three-dimensional objects that would otherwise satisfy the definition of "document" because they memorialize or represent thought; documents are considered more as 2-dimensional representations.

I think "realia" - a term I had never heard before - describes runestones better than "document".

Runes continued to be used long past the Elder Futhark period and from the medieval period manuscripts survive that fit the modern conception of a "document", most famously the Codex Runicus https://www.e-pages.dk/ku/579/html5/ (202 pages)

https://www.youtube.com/watch?v=2yWWFLI5kFU describes another side-effect of encoding old scripts/runes.

>Documents? I had the strong impression that there are no documents written in runes.

If a clay tablet counts, why not a runestone?

I'm not knocking runestones for being the wrong medium. I'm knocking them for not being documents. A typical cuneiform record might be analogized to an invoice for delivery of a crate of shirts or whatever. (And of course we also have textbooks, dictionaries, literature, correspondence, business reports, mathematical treatises, and every other type of written work.) A typical runic record would be more like the text "Made in Taiwan" printed on the shirt labels.

One of the biggest problems in the study of these cultures is that they left no written records. We know they had a writing system, the runes, but as far as we can tell they almost never used it for anything. Quite the opposite is true of Mesopotamian cultures, where we're buried in more records than we have the manpower to translate.

I suppose it also matters what you think a rune is. Does futhork count? There's parchment with that written on it. Elder Futhark, none as far as i know.

Reply on news.ycombinator.com