Email list hosting service & mailing list manager

Truth in advertising Alan Watson (24 Jun 2004 00:05 UTC)
Re: Truth in advertising campbell@xxxxxx (24 Jun 2004 01:43 UTC)
Re: Truth in advertising felix (24 Jun 2004 05:07 UTC)
Re: Truth in advertising Robby Findler (24 Jun 2004 12:07 UTC)
Re: Truth in advertising Alan Watson (24 Jun 2004 17:33 UTC)
Re: Truth in advertising Felix Winkelmann (25 Jun 2004 04:58 UTC)
Re: Truth in advertising Alan Watson (25 Jun 2004 16:12 UTC)
Re: Truth in advertising Robby Findler (25 Jun 2004 16:39 UTC)
Re: Truth in advertising Alan Watson (25 Jun 2004 16:59 UTC)
Re: Truth in advertising Bradd W. Szonye (25 Jun 2004 18:51 UTC)

Re: Truth in advertising Felix Winkelmann 25 Jun 2004 04:58 UTC

Alan Watson wrote:
> Robby Findler writes:
>
>
>>>  This SRFI cannot be supported by all Schemes. Therefore, by using it
>>>  you are restricting the portability of your code.
>>
>>Isn't this true, in principle, of all the SRFIs?
>
>
> No and yes.
>
> It is certainly true that not all SRFIs *are* implemented in all
> Schemes. However, most SRFIs *can* be implemented in all R5RS-compatible
> Schemes, often trivially. This SRFI (and possibly certain others)
> *cannot*.
>
> The following two examples illustrate the difference:
>
>   If I write code that uses SRFI-1 and then discover that I want or need
>   to run the code in a Scheme that doesn't implement SRFI-1, I can
>   simply load Olin's reference implementation.
>
>   If I write code that uses SRFI-55 and then discover that I want or
>   need to run the code in Scheme48, I am hosed.
>

If we restrict ourselves to SRFIs that can be implemented on
any R5RS system without changes, we would get nowhere.

cheers,
felix