On the microagression of U+1F4A9

This model view culture article i can text you a pile of poo, but i can't write my name is making the rounds right now and might be of interest to folks here.

For my part, i think the problem is an input-ui issue rather than one of the unicode spec (not that unicode isn't without fault). In an ideal world, we wouldn't much care whether or not a glyph is (or is not) represented by an ideal code point because our input mechanisms would make that nuance irrelevant. In a way, i'm reminded of the opentype input gripes voiced by others; ideally, this would be a nonissue but our input software and devices seem to rub in our faces just how secondary the implementation is.

Although the article is effectively arguing that the spec's favoritism of conjuncts and software-level implementation (both os and type rendering and layout engine) disregards the legitimacy of the components in her written name, the issue feels like an uncomfortable intersection of the myriad nuances of calligraphy butting head-to-head with a a rendering system that favors abstraction and composition. I'm no unicode expert, though, so i likely have an imperfect view of this situation. If anything though, it is a stunning acknowledgement of how popular phones are that in what appeared to have been a matter of months, i already have skin tone modifiers available in my os. 

in what is also a minor surprise, the first thread on hacker news has a remarkably cogent and insightful analysis of the situation and a really interesting back and forth between the commenter and the article's author. my favorite part was this botched attempt to call out the commenter

> Is Bengali your first language?

curious to hear others' thoughts


  • Nick ShinnNick Shinn Posts: 1,534
    edited March 2015
    I think it is the responsibility of national governments to advocate for the languages spoken by their citizens and to that end, to regulate multi-nationals and participate in NGOs such as the Unicode Consortium.

    I think that one must expect a certain amount of triage (and hence discrimination—and I mean that in a neutral manner) from multinationals; at some point they make decisions, based on economics as much as anything, on where to draw the line of language support. Adobe’s approach, as described some time ago by Thomas Phinney in his Adobe blog (sorry, can’t find a link) seems reasonable.

    As an independent foundry owner-designer-publisher, I view the Unicode Consortium as rather like the WWW consortium or the Microsoft-Apple consortium that produced OpenType. In other words, I go with the flow, and do not have the time or inclination to get involved as a stakeholder and present my concerns.

    I think Unicode and OpenType are awesome, despite the flaws. Nothing’s perfect. 

    Based on my experience designing typefaces with Latin-Greek-Cyrillic encodings, I favor alphabets over graphemes. K, Kappa and Ka, for instance, will often require being represented by different glyphs in the same font.

    Nonetheless, here is a scheme I developed for a reduced glyph set, unicase, of Latin, Greek and Cyrillic:

  • James PuckettJames Puckett Posts: 1,689

    Mukerjee’s essay conveniently fails to mention that people don’t always agree on things like spelling, pronunciation, or how their writing systems work. So no matter how a language like Bengali is supported someone is going to be unhappy about it.

    I think it is the responsibility of national governments to advocate for the languages spoken by their citizens and to that end, to regulate multi-nationals and participate in NGOs such as the Unicode Consortium.

    Agreed. I’ll go further and add that national governments should document their local writing systems, in western languages, so that the foreigners who have to engineer typesetting solutions can have some native sources to base their work on. Public universities around the world provide free corpora of local languages, something similar should exist for writing systems.

  • attarattar Posts: 209
    Was a proposal even submitted to the Consortium? He says membership is paid but you don't have to be a member to submit a proposal. It there was a proposal that's been rejected for no particular reason, then I'd say he has a point.
  • Dave CrosslandDave Crossland Posts: 1,052
    edited March 2015
    you don't have to be a member to submit a proposal.

    You don't have to be a man to write software, or make a career in type design, but not very many women have done so; for type this may change as according to UoReading their MA Type Design course is now an even 50/50 split for graduates.

    The point of contemporary social justice is that equality is on the books, is here in theory, but not enacted, not here in practice.

    Could the UC could be more welcoming to people who don't know anyone at a member organization? Yes, it could.

  • attarattar Posts: 209
    > You don't have to be a man to write software,

    Should I infer from the analogy that you are basically saying it is impossible in practise for a non-member to have a proposal be accepted?  
    Please show me a proposal for what the author speaks about that was refused without any clear technical ground and then we'll talk.
    Meanwhile, here is a counter-example: the IEC Power proposal was submitted by a small group of non-members without the massive support a country's institutions may back it with and it was accepted into Unicode.
  • Thanks for sharing this piece, Marcos. My admittedly shallow understanding of the Bengali unicode situation was that the support was pretty good. Apparently not.
Sign In or Register to comment.