[boost] Re: (Another) socket streams library

From: pedro.lamarao_at_[hidden]
Date: 2005-04-22 13:16:05

Iain Hanson wrote:

[SNIP]

> I agree that the socket class may be private detail for an iostreams
> implementation but I do not agree that all network i/o should be done
> this way. I don't think that iostreams as a model works well for packet
> protocols. 

Do you mean packet protocols as in AF_DGRAM or AF_PACKET?
Why do you think it wouldn't work well?

> BTW do you use path MTU or MSS to set your streambuf size?

You mean, the output buffer?
I haven't gotten to that kind of detailed fine tuning.
But the current default value for the buffer is much larger than 1500 bytes.

I'm mostly interested in the syntax and semantics offered to the user
and the kinds of new expressions it allows him to use; these days most
of the work I do on this library involves producing useful examples to
help me fine tune the design.

--
 Pedro Lamarão