2012/6/1 Asmus Freytag <[email protected]>: > The chances that any form of meta encoding for symbols (including ligation) > will ever reach critical mass in support is less than for > Latin/Greek/Cyrillic accents, because - as of today - there's no established > use for any of these schemes. > > All of these things remain solutions in search of a problem.
No, my poposal gives something that is immediately usable, and does not create any ambiguity. It is simple to implement even without the presence of a technical ligaturing solution. Those flags will be immediately usable, without any of the political complications created by the case of flags. It will avoid prolieferations of proposals, and infinite debates for encoding or not some flags, or for changing the representative glyphs. It ensures a stable solution for the long term, and still allows further developments (that will take many years) to support a larger set of flags (and also without the complications created by the fact that these flags have semantically significant colors, that cannot be ignored). I won't support a proposal that wants to encode even just the basic flags of UN member countries, because we will immediately fall into the trap of the representative glyph, and the fact that they are monochromatic (this is a MAJOR barrier against their encoding, as the representative glyphs will create new confusions). My solution allows distinctions immediately for ALL flags. It does not limit the encoding. And then it allows renderers to use a better form as seoon as they wish (provided that the alternate glyphs they will generate for the composition preserves the initial semantic of the encoded flag, and is not used to designate another country or flag, or create confusions about which country is intended, notably if the renderer only supports monochromatic outputs). My solution will be reasoannably short : most flags for modern UN member countries will be encoded with onle 2 symbolic characters. Their behavior is clear, as well as their semantic, or behavior for word-breaking and line breaking, aven if these symbols do not extended the current definition of default graphme clusters (in my proposal, **each** encoded symbol is a separate default grapheme cluster, even when they are not replaced by a single "ligatured" glyph). This is NOT a *meta* encoding, as it immediately has a concrete realisation, not *requiring* a complex solution. A basic font containing the few symbols (maximum: 192, divided in three subsets of 64 each) will be enough. This means a fast and easy deployement as well (showing the actual graphic and colorful flags is left to implementations and external specifications).
