Fix Phobos dependencies on autodecoding

[H. S. Teoh]

On Fri, Aug 16, 2019 at 04:41:01PM +0000, Abdulhaq via Digitalmars-d wrote:
[...]
> It's no coincidence that those people who can read, write and speak
> more than one language with more than one script are those who think
> Unicode is beneficial.

To be clear, there are aspects of Unicode that I don't agree with.  But
what Walter is proposing (1 glyph == 1 character) simply does not work.
It fails to handle the inherent complexities of working with
multi-lingual strings.


> It seems that those who are stuck in the world of anglo/latin
> characters just don't have the experience required to understand why
> their simpler schemes won't work.

Walter claims to have experience working with code translated into 4
languages.  I suspect (Walter please correct me if I'm wrong) that it
mostly just involved selecting a language at the beginning of the
program, and substituting strings with translations into said language
during output.  If this is the case, his stance of 1 glyph == 1
character makes sense, because that's all that's needed to support this
limited functionality.

Where this scheme falls down is when you need to perform automatic
processing of multi-lingual strings -- an unavoidable inevitability in
this day and age of global communications. It makes no sense for a
single letter to have two different encodings just because your user
decided to use a different font, but that's exactly what Walter is
proposing -- I wonder if he realizes that.


T

-- 
Written on the window of a clothing store: No shirt, no shoes, no service.