I've often lamented that the leap from ASCII (or its 8-bit extensions) to Unicode is so big and feels like fully understanding Unicode and its implications requires a degree in linguistics; for many applications it might not be practical to support all the bells and whistles of Unicode. At the same time especially ASCII is too limited for even plain English, never mind other European languages. So I was very happy when I discovered that there is a well-defined sane subset of Unicode for European usage. This subset allows writing pretty much all European languages, but does not include much of the weirder parts of Unicode so developers can handle data with more basic understanding of Unicode.
I understand that some people might have more maximalist opinion on Unicode, that programs should try to support all of Unicode. But to me personally correctness is more important, and having this small subset means that it will be easier to define what is the correct behaviour and ensure that the programs actually behave as expected. Of course ideally programs would handle also non-european languages correctly, but in reality I see it difficult for developers to ensure correct behaviour for writing systems that they do not know and as such there is high risk of having bugs.
Also having this subset defined means that it should be more practical to have a common subset that different fonts support. Currently afaik it is far more of a wild west in terms of what glyphs are or are not included. I think most people agree that it is completely unrealistic to expect fonts to cover all of Unicode, so this subset could be the sort of thing that would be more reasonable to expect fonts to support.
I really hope that this sort of thing becomes more common and we could see more well-defined Unicode subsets pop up to cover other languages/regions/uses. This way eventually it would be easier at a glance to identify what parts of Unicode different things support, and also for developers to prioritize what they should work on.
1 comment
[ 2.7 ms ] story [ 19.8 ms ] threadI understand that some people might have more maximalist opinion on Unicode, that programs should try to support all of Unicode. But to me personally correctness is more important, and having this small subset means that it will be easier to define what is the correct behaviour and ensure that the programs actually behave as expected. Of course ideally programs would handle also non-european languages correctly, but in reality I see it difficult for developers to ensure correct behaviour for writing systems that they do not know and as such there is high risk of having bugs.
Also having this subset defined means that it should be more practical to have a common subset that different fonts support. Currently afaik it is far more of a wild west in terms of what glyphs are or are not included. I think most people agree that it is completely unrealistic to expect fonts to cover all of Unicode, so this subset could be the sort of thing that would be more reasonable to expect fonts to support.
I really hope that this sort of thing becomes more common and we could see more well-defined Unicode subsets pop up to cover other languages/regions/uses. This way eventually it would be easier at a glance to identify what parts of Unicode different things support, and also for developers to prioritize what they should work on.