Even worse is the Pashtu scripting style commonly used in Pakistan which effectively so intertwines the characters that Newspaper headlines are still handlettered today and stripped into the typeset copy.
See the thread on Apple Advanced Typography. The thread cites the Monotype Recorder for 1981. This was the year Monotype Nastaliq was introduced for the Farsi and Urdu markets.
The implementation was developed, as the thread says, in co-operation with the Pakistan association of lithographic printers.
Before a team of Cambridge phycists headed by Prof Dr John Billingsley in 1976 finished developing a laser imaging system with funding from the Monotype Corporation, and before that system was applied to the most complex Arabic calligraphy, Nastaliq was calligraphed, the calligraphy photographed, and the reproductions photographically copied onto the printing surface. The first metal composition of Monotype Naskh for the western Arabic market was in 1976 and IBM did not introduce a daisywheel Selectric for Naskh until the mid-nineteen seventies. Dr Joseph Becker's article on Arabic composition in the Xerox Star operating system appeared in Scientific American in 1984 (the table-based transform approach had by then been sold by Monotype for four years, which US writers tend to 'forget' as Monotype was not perceived as a US supplier).
The rub with the Arabic script and with Indic scripts is indeed that, "Aesthetics can play such a large part of what is found acceptable that mere mechanical methods of laying out text fails." Where does the border lie between incorrect orthography, correct orthography with lowend typography, and correct orthography with highend typography. The Unicode/TrueType idea was to leave this to the type maker who in the obligatory simple CMAP shaping would support correct orthography with lowend typography and in the optional smart MORX shaping would support such level of sophistication as suited the design and the market. If people expect to get highend Arabic TrueType with endless ligation and intonations for recitals and much else for free, then this is not going to happen.
Incidentally, Adobe's Arabic contains PostScript. It was developed by John Hudson who had to hop through hoops backwards since Microsoft VOLT does not support PostScript, so the shaping logic had to be copied and pasted. If memory serves Apple's tools do not support PostScript either. Anyway, what this discussion needs is a few folks with advanced knowledge of Arabic calligraphy. And I'm absolutely not one such.
hh