On Wednesday 31 December 2003 06:04 am, Alex Shinn wrote:
>..
> Two minor comments:
>
> This doesn't have the fix for properly handling newlines in ~A.  It
> should pass the test
>
>   (format "~A~A~&" "\n" "") => "\n"

The text of ~&, freshline, is:

"output a newline character if it is known that the previous output was not a
newline"

An implementation is not required to detect all instances of newline output.
So, while non-optimal, the implementation is correct.  [I will change the
code to detect this case, however.  Thanks].

> Also ascii-tab is written first "ascii-tab" then "ASCII-TAB", which
> doesn't work on Schemes that default to case-sensitive readers.  It
> wouldn't be worth supporting such non-standard behavior in general,
> but since most people consider it good style not to change the case of
> identifiers the reference implementation should probably use a
> consistent case.

Sorry.  I thought I had caught all of these.  [The standard still specifies
case insensitivity, but I am aware of the problem].

Thanks for catching these,
-KenD