Marcel Hendrix wrote:
> Peter Knaggs <pkna...@bournemouth.ac.uk> wrote Re: RfD: Escaped Strings S\" (version 5)
> [..]
>> Translation rules:
>>     Characters are processed one at a time and appended to the
>>     compiled string. If the character is a '\' character it is
>>     processed by parsing and substituting one or more characters
>>     as follows, where the character after the backslash is case
>>     sensitive:
>>     \a   BEL (alert, ASCII 7)
> [..]
>>     \\   backslash itself
>>     \    An ambiguous condition exists if a \ is placed before any
>>          character, other than those defined in 6.2.xxxx S\".
> [..]

> Why was it necessary to make this an ambiguous condition?
> S\" is not used by any systems not represented in the Forth 200x effort.
> IMHO it is a bit silly (for a standards effort) not to mention all \<char>
> codes in use today, and/or to allow future vendor-specific extensions that
> will break portability of code and require work-arounds.

> A general solution could be to require that a deferred (and standardized)
> hook word is executed in case an unknown code is encountered. This would
> guarantee that any future S\" problems can be fixed by user code.
> Thinking this through should convince most people that simply forbidding
> non-standard \-codes is by far preferable.