Pim Blokland wrote as follows, responding to Doug Ewell. quote
> I suspect it would end when you start talking about combinations like qj > and fþ that are unlikely to appear in natural language text. At least > gj exists in Hungarian. fb, fh and fk are very common in Dutch (much more so than fj). fþ exists in Icelandic; at least I've found "arfþegi". However I don't speak Icelandic, so I've no idea if this is a combination of two "subwords". end quote During the spring and summer of 2002 I produced a number of web pages about encodings for ligatures, the encodings using the Private Use Area. Some of the characters mentioned are encoded within the golden ligatures collection. http://www.users.globalnet.co.uk/~ngo/golden.htm I will try to add qj gj and f thorn in due course. Where I have an f ligature I have added an ff ligature into the encoding scheme, so I expect to add ff thorn in as well, just in case it is needed, though I have no knowledge of whether it is ever used, though I was unaware of the possibility of an f thorn ligature until reading this thread. While I am adding some more ligatures to the collection, if anyone wants any other characters added in, please email me privately. I found that encoding the golden ligatures collection led to me learning about a number of interesting aspects of typography of which I was previously unaware, so it was an educational experience for me as well as being fun and useful in practice within its limits. Naturally, my production of the golden ligatures collection does not of itself produce fonts which contain these ligatures, yet it does help a little in making the possibility topical, so maybe a few of the font designers who read this list might perhaps include more ligatures in their fonts. An interesting aspect of my codification of ligatures is that any documents produced using them will not be standard Unicode documents. However, the encodings might be very useful so that someone may make artistic typography fonts using a font production program such as the Softy shareware program and be able to produce pages of hardcopy print out locally using such a font, where a ligature character such as ct may be encoded as U+E707. Naturally, there is nothing to stop anyone encoding a ct ligature however he or she chooses within the Private Use Area, yet my collection of encodings is a published, consistent set which would help with interoperability of fonts from various artists. I am currently producing a typeface which I am calling Quest text so that I can have a typeface available which has whatever ligatures I choose. I have so far produced all of the lowercase letters, the digits, full stop and twelve capitals and also lowercase long s, ash, eth and thorn. I am hoping that the font will be useful for English, Old English and Esperanto in particular, though I can add characters where I choose, using both regular Unicode code points and Private Use Area code points, both from the golden ligatures collection and from other published Private Use Area encodings. I am producing Quest text using the Softy program and am finding it a very effective program. More recently, a new development, designed primarily as a means to produce displays of languages of the Indian subcontinent upon the screens of interactive televisions using the font format capability of those interactive televisions using the ligatures of those languages, may be a very useful way to use the ligature encodings of the golden ligatures collection as well. http://www.users.globalnet.co.uk/~ngo/ast03300.htm So, a document in which one wishes to have a ct ligature would have the ct ligature encoded as ct or maybe c ZWJ t depending upon the circumstances, and a .etf file would have one or both of the following lines, depending upon the application. ct U+EBEF U+E707 (that is, four characters) c ZWJ t U+EBEF U+E707 (that is five characters) Thus the combination of the golden ligatures collection, an .etf file and various software tools to use them could be an effective way of allowing people to use ligatures on a wide variety of platforms while having the documents containing the original texts encoded using regular Unicode characters only. A text file containing codes from the golden ligatures collection would thus only be used locally on a temporary basis for a current task, though to useful effect. Some of my small fonts produced using Softy are available at the following web page. http://www.users.globalnet.co.uk/~ngo/font7001.htm William Overington 10 March 2003
