The Case Against Autodecode

[H. S. Teoh via Digitalmars-d]

On Fri, May 27, 2016 at 07:53:30PM +0000, Adam D. Ruppe via Digitalmars-d wrote:
> On Friday, 27 May 2016 at 19:30:53 UTC, Andrei Alexandrescu wrote:
> > It seems code points are kind of useless because they don't really
> > mean anything, would that be accurate? -- Andrei
> 
> It might help to think of code points as being a kind of byte code for
> a text-representing VM.
> 
> It's not meaningless, but it also isn't trivial and relevant metrics
> can only be seen in application.
> 
> BTW you don't even have to get into unicode to hit complications. Tab,
> backspace, carriage return, these are part of ASCII but already
> complicate questions.
> 
> http://stackoverflow.com/questions/6792812/the-backspace-escape-character-b-in-c-unexpected-behavior
> 
> came up on a quick search. Does the backspace character reduce the
> length of a string? In some contexts, maybe.

Fun fact: on some old Unix boxen, Backspace + underscore was interpreted
to mean "underline the previous character". Probably inherited from the
old typewriter days. Scarily enough, some Posix terminals may still
interpret this sequence this way!  An early precursor of Unicode
combining diacritics, perhaps? :-D


T

-- 
Everybody talks about it, but nobody does anything about it!  -- Mark Twain