<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">On 2/17/2024 12:02 PM, Christoph Päper
via Unicode wrote:<br>
</div>
<blockquote type="cite"
cite="mid:E3B09251-5CF6-49BB-BD34-D86908C45705@crissov.de">
<pre class="moz-quote-pre" wrap="">Asmus Freytag via Unicode <a class="moz-txt-link-rfc2396E" href="mailto:unicode@corp.unicode.org"><unicode@corp.unicode.org></a>:
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">
We usually don't encode characters intended for use in handwriting, except if they are needed to digitally archive manuscripts. Not sure grade school papers pass that bar.
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">
Every piece of writing might be digitally archived nowadays, even more so in the future. Therefore, every _established_ literal atomic sign should be encodable, so it can be unambiguously read by machines. I strongly believe this includes paralinguistic signs, whereas nonlinguistic signs (e.g. much of ISO 7000) would require an extension of the scope of Unicode (although several graphic symbols from that and other standards already have a codepoint assigned to them).
This one is clearly well established, i.e. has at least one canonical form and meaning, even if its use is geographically limited. It cannot be represented by a combination of other, already encoded characters.
</pre>
</blockquote>
<p><font face="Candara">That's an argument a proposal could make,
but I'm not sure I'm ready to agree with that analysis.</font></p>
<p><font face="Candara">Even if we approach 100% digital archiving,
not everything can be, will be or needs to be archived as *plain
text*. (Or even rich text). <br>
</font></p>
<p><font face="Candara">Manuscripts are a good example of
handwritten text that benefits from conversion to digital text,
because they are subject of intense scholarship that would
benefit from having the usual array of digital text processing
available, such as search, and convenient rendering of excerpts.</font></p>
<p><font face="Candara">People are studying the marks accompanying
cave paintings, such as lines, circles or dots. One even
resembles a hash mark #, making that arguably the oldest
uniquely recognizable symbol ever encoded as a character.
(Aside: dots and lines don't count, because we encode many
different dots and lines).</font></p>
<p><font face="Candara">For those studies, there's no overriding
need to place the symbols into running text, or to attempt to
show sequences of them as plain text. Therefore, such use alone
is not sufficient rationale for deciding the question what
constitutes an abstract character and to provide a standardized
encoding, plus assign properties such as line breaking behavior.</font></p>
<p><font face="Candara">The Dutch mark in question is interesting in
that it's clearly associated with a well-defined concept and has
a recognizable (and conventional) shape. Neither of those two
aspects present any obstacle to encoding. However, the need to
represent it in plain text needs to be established and any
successful proposal will have to provide an argument that is
specific and to the point. <br>
</font></p>
<p><font face="Candara">The mere claim of a general principle as
suggested above is not sufficient to make a persuasive argument
for a specific encoding.</font></p>
<p><font face="Candara">A./<br>
</font></p>
</body>
</html>