Ligatures (revisited)...

Say, this might be a sore point (I've only been reading this list
about 5 months, so I might be rehashing old stuff), but has much
thought been put into ligatures that aren't represented in Latin1?

Keep in mind, Latin1 was never meant to be an all-conclusive set
of Western European glyphs.  It was a *reasonable* amount that
could be fit into 8 bits (and hence limited to <=256 symbols)...

Since we are using mneumonic names like "&egrave;" we can have
many more than 256 symbols...

Hence, presentational glyphs, like the French oe, OE, the Dutch
ij, IJ, etc. *can* be included.  In Latin1, they were excluded
because it was felt that rendering software should be able to
recognized the two octet sequence "o" and "e", and render it as
a ligature.

It turns out that this is dead wrong.  You can't always render OE
or oe as a ligature.  It depends on the word (in fact, you need a
dictionary to figure it our correctly).  For instance, "oeuf"
(egg in French) is ligaturized, but "Boenhoffer" (the French
spelling of b&ouml;nhoffer, the German Catholic philosopher), would
not (and must not!) be ligaturized.

This is further confused by the issue that some French publishers
(though they probably aren't alone) use the "AE" ligature (&aelig;),
which (at least in Danish, I believe) *ISN'T* a ligature, but a unique
letter/vowel.  This is seriously confused...

In other words, for sorting and comparison reasons, "AE" == "&AElig"
in French, but not in Danish.

In French, &aacute; should sort (and compare) to "a", but in
Vietnamese, they are completely different letters/vowels.

I also brought up the question of hyphenation (at least off-line),
but this seems to have provoked an "underwhelming" response...

Too bad.  Hyphenation isn't that hard to do correctly, and it can
dramatically increase the quality of presentation, as well as make
for better use of the screen.  Even with my 21" colour screen, it
seems that I never have enough space for all my windows...

As always, flame me *off-line* if you must.

-Philip

Received on Wednesday, 26 April 1995 16:31:02 UTC