Well it is still not so universal as there are wide ranges of glyphs excluded for now to encoding as characters: - many icons used in cartography (there are progresses now, but in their emoji form for use in talks/instant messaging/SMS, where they are colorful but do not match the simple glyph used in maps that will use them in multiple distinctive colors and sizes; often the 3D effects are undesirable, most of the time shey should be monochomatic, and color/sizes and other styles applied conditionally by some stylesheet) - country flags have been included but many regional emblems are excluded (as they don't match any ISO 3166-1 code) - common road signs/street signs and signs for indoor facilities & services - various symbols used in software UIs: many OSes have to provide an additional font encoding them as PUAs or using some encoding specific to the font containing them (much like it was with most dingbats in older Adobe Postscript fonts) - various box drawing characters used in legacy terminals (notably in Teletext and on older 8-bit systems): a few of them were added from DOS/OEM codepages.
Of course corporate logos used in proprietary fonts for specific OSes cannot be encoded for legal reasons (not as long as there's no licencing permitting its inclusion in other fonts for other OSes): e.g. logos from Apple and Microsoft for MacOS and Windows., but as well other logos for various Unix editions and even Linux distributions, including the green bot for Android), and other logos registered as trademarks, and logos used to identify some national technical standards and indicating a conformance (usage is restricted by the standard defining these logos, many of them being supported by private organizations selling their licences). All these logos have to be encoded transported as embedded or linked images carrying their own copyright (which must be also transported along with their graphic definition). As well we cannot encode glyphs representing physical persons (e.g. based on a photo of Barak Obama), or containing biometric data (e.g. fingerprints, DNA sequences, personal handwritten signatures), or some protected artitectural designs (even if these are old historic designs such as Greco-Roman designs), or logos representing some coin faces. As well we cnanot represent precise taxons (animalia or flora are very roughly represented, but we don't go up to the species level, or even just the gender) 2016-08-28 2:34 GMT+02:00 Asmus Freytag (c) <asm...@ix.netcom.com>: > On 8/27/2016 10:15 AM, Doug Ewell wrote: > >> Ken Whistler wrote: >> >> I would contend that encoding wildly popular and extensively used >>> little pictographs as characters makes a whole lot more sense in the >>> abstract than encoding box-drawing graphic pieces for completely >>> obsolete screen technology ever did. >>> >> >> Though to be fair, the screen technology was a lot less "completely >> obsolete" in 1991, when the box drawing characters were encoded (Unicode >> 1.0), than it is today. >> > > They came into the draft in the period from 1988 to 1990; during that > period, dialogs using "text mode" displays were common for many > applications, not just pure terminal emulation. > > To demonstrate that it was "universal" Unicode had to show that it could > be used to replace the entire range of actively used character encodings. > Just as the same universality argument is what drove the initial acceptance > of emoji. And will drive acceptance of a whole host of other symbols and > characters, no matter how well they stack up against the purity of > principle. > > A./ >
