The discussions about raw strings probably burnt some magnitudes more time than their implementation, so I try to keep it short...
Summary
The design that currently has the most support includes custom delimiters for "regular" strings as well (using the hash character).
Example: "example\\\"backslash\"" == #"example\\"backslash""# // true
example\"backslash"
Idea
I'd like to build a tiny piece on top of that concept and not only treat double quotes as regular characters unless they are accompanied by the right number of hashes, but apply the same logic to backslashes.
Example:
print(#"The "backslash" has a special meaning, thus \(1 + 1) is replaced with #\(1 + 1)#")
would output
The "backslash" has a special meaning, thus \(1 + 1) is replaced with 2
Explanation
Current string literals have no # in front of them, and nothing would change here. But if you want to use a literal that contains either a double quote and/or a backslash, you prepend the opening and closing quotes with #, and leave the string itself untouched.
If you need #\
or #"
in the string itself, you keep adding more hashes.
Downsides
Well, if your string contains long sequences of hashes followed by backslash, "real raw" strings would be slightly more comfortable. I don't think that scenario is common, though.
Also: Source files with many number signs... but if you don't like those, you probably should avoid strings containing them ;-) (and it should also be possible to use \ or ` instead).
Benefits
No need to escape backslashes or double quotes inside strings while retaining the ability to perform interpolation. Imho this is a critical feature for most kinds of templating, where you have a stream of characters with small parts that are generated.