In article <··············@eho.eaglets.com>,
Sam Steingold <···@goems.com> wrote:
>`read-sequence' doesn't specify whether it reads bytes or chars, leaving
>that for the stream to decide.
It reads whatever the STREAM-ELEMENT-TYPE is. Which needs to be compatible
with the type of the sequence it's storing into.
> what if the stream happened to be a
>socket? most implementations provide this extension, but they disagree
>on what `read-sequence' should do with them (clisp refuses to do
>anything and instead provides `read-byte-sequence' and
>`read-char-sequence' for sockets; allegro insists on using chars).
This seems pretty arbitrary. At the level of the read() system call, Unix
doesn't distinguish between sockets and other types of streams; they're all
just byte streams.
>Any chance X3J13 will ever specify this?
Seems unlikely. Language specifications generally avoid dealing with
OS-specific details like networking. Although, now that TCP/IP stacks
are as almost as ubiquitous as hierarchical file systems, it might make
sense to have some generic interfaces. Java, as a language specifically
directed towards networked applications, is leading the way in this.
--
Barry Margolin, ······@bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Don't bother cc'ing followups to me.