current input & output ports
Taylor Campbell
(17 Jun 2005 01:32 UTC)
|
Re: current input & output ports
Alex Shinn
(17 Jun 2005 01:51 UTC)
|
Re: current input & output ports
Taylor Campbell
(17 Jun 2005 04:46 UTC)
|
Re: current input & output ports
Per Bothner
(17 Jun 2005 06:46 UTC)
|
Re: current input & output ports
Michael Sperber
(17 Jun 2005 21:23 UTC)
|
Re: current input & output ports Taylor Campbell (17 Jun 2005 23:53 UTC)
|
Re: current input & output ports
bear
(19 Jun 2005 18:51 UTC)
|
Re: current input & output ports Taylor Campbell 18 Jun 2005 00:03 UTC
On Fri, 17 Jun 2005, Michael Sperber wrote: > The more I think about it, the more I like it. There are two > downsides: > > - The last vestige of R5RS compatibility goes out the window. Does > anyone care? I don't. > - This SRFI is then no longer suitable for ad-hoc debugging output > (which, I think, provides the rationale for CURRENT-OUTPUT-PORT). As I suggested, slots in the dynamic environment could be added for ports with more specifically meaningful applications; e.g., there could be a DEBUG-OUTPUT-PORT added. But as it is the current input & output ports are not specifically intended for random debugging output, or I've seen them misused a great deal, and even if they were that would be hardly reason enough for the havoc wrought on signature consistency for them (not to mention an unclear name).