Unicode brought order to chaos. Its definition reaches more than 140,000 characters, originally spread across many incompatible sets that are now a click away, or at worst, a code away. That makes Unicode an almost inexhaustible source of strange symbols and mysterious representations. The most recent example keeping the Web on edge is ⍼, code U+237C, "Right angle with downward zigzag arrow", or simply 'angzarr'. Student Jonathan Chan from the University of British Columbia decided to dive into the well of its origins, and to be honest, I don't think he managed to come out…
Unicode and STIX
The adventure doesn't start off well: the section 22.7 of Technical Symbols in the Unicode standard under the category Miscellaneous Technical doesn't even make a basic mention of the symbol. Jonathan's research led him to the Proposal to Encode Additional Mathematical Symbols from March 14, 2000, where we can see 'angzarr' in complete solitude, with a certain air of mystery.
However, that wasn't his starting point. The roots of the proposal extend back to the beginnings of the STIX Project, which aimed to introduce additional symbols into Unicode for technical and mathematics-focused publications. Here Jonathan discovered the work of Barbara Beeton, who collected and combined some of the first character tables in June 1997. Studying those tables, our friend 'angzarr' appears again as D97C, with an association to the technical report ISO TR9573. A later table confirms 'angzarr' as part of that report, and even assigns it a temporary Unicode code, E248.
ISO Archaeology
The original version of TR9573 (unfortunately behind a paywall, like many ISO documents) dates back to July 1991. There, 'angzarr' is defined as D97C/55676. Its later editions already identify the symbol as U+237C, but Jonathan suggests that happened after its formal inclusion in Unicode.
And there's more! Section 13 of TR9573 (where 'angzarr' is) serves as a replacement for Annex D of ISO 8879:1986. However, examining that document, Jonathan discovered that 'angzarr' is conspicuous by its absence.
Back to TR9573, its Section 5 reports that the symbols contain "an identifier for glyphs, registered according to ISO/IEC 10036". Fortunately, the 10036 registry is online, and if we search for 55676 there…
A Symbol for $5
The most interesting thing is that 10036 describes the glyphs as 'inherited' from AFII. What was AFII? The Association for Font Information Interchange, the registration authority for ISO 10036. Under AFII, anyone could register a symbol for a modest sum (between 5 and 50 dollars at the time), but the industry soon favored ISO 10646 (the Universal Coded Character Set) and Unicode. The International Glyph Register of AFII should have more details on 'angzarr'. Unfortunately, that document seems to be lost.
A Wild Barbara Appeared...
Jonathan's research concludes (for now) with a participation by Barbara Beeton herself on StackExchange, who confirmed that TR9573 indeed precedes AFII and the STIX Project, but added nothing more about the origin or use of 'angzarr'. So… all we have is a mysterious symbol with a generic description, inherited from documentation dating back to the early '90s, and manifested at some point in the second half of the '80s. Do you have a better idea? Leave a comment!
Sources: Jonathan Chang, Hackaday, Stack Exchange
