Ready to finalize SRFI 38. bear (01 Mar 2003 00:07 UTC)
Re: Ready to finalize SRFI 38. Marc Feeley (01 Mar 2003 00:21 UTC)
Re: Ready to finalize SRFI 38. bear (01 Mar 2003 01:39 UTC)
Re: Ready to finalize SRFI 38. Per Bothner (01 Mar 2003 00:51 UTC)
Re: Ready to finalize SRFI 38. bear (01 Mar 2003 01:40 UTC)

Re: Ready to finalize SRFI 38. bear 01 Mar 2003 01:39 UTC


On Fri, 28 Feb 2003, Marc Feeley wrote:

>I took a quick last look and found something unexpected:
>
>   <indexnum> --> <digit 10>+ | <radix R> <digit R>+
>
>Why does <indexnum> allow a radix?  When is such a thing useful?  This
>is a problem for Gambit because it allows symbols to start with a "##"
>prefix.  So for example ##o12# is a symbol, but SRFI-38 would treat
>this as if it were #10# .  Is there any rationale for this?

I didn't think it was useful, but I got a request for allowing
it to be formatted with different radixes, etc.  Since R5RS
doesn't specifically mention symbols starting with # as a
possibility, I figured this was the simplest way to handle it.

If it will actually break stuff in Gambit, would the person who
asked for radix variants, etc, please explain why it's a good
idea?  Because if nobody convinces me, I'm willing to yank it out.

			Ray