Interpreter conflict between '<space>' and ' name

[SirWumpus]

Reading Forth 200x Draft 19.1 I think there is a conflict in 3.4 The Forth text interpreter with respect to the handling of 3.4.1.3 Text interpreter input number conversion. In 3.4.1.3

 <cnum> := '<char>' 

where '<char>' can be any printable ASCII character including space (as defined in 3.1.2.1 Graphic characters). So in theory I should be able to put a space on the stack or define BL with:

ok ' '
ok ' ' CONSTANT BL

but they appear to fail because the manner in which the interpreter is defined it tries to lookup a word before parsing a number, so single-quote space single-quote is treated like

ok ' <word>

where <word> is the single-quote. Should the standard alter the 3.4.1.3 to exclude space from '<char>'or modify the interpreter behaviour to try to parse numbers before word lookup?

The reason this came up was that I stumbled on this discussion https://forth-standard.org/proposals/bl-rationale-is-wrong#reply-249 and had a play with my implementation testing stuff and noticed a "bug". After some reading it appears to be a conflict in the standard draft.

[ruv]

You are right, and it's a known problem. Somebody should submit a proposal, and it will be most probably introduced into the next version of the standard.