On Fri, 7 Feb 2003 22:27:51 +0200 (FLE Standard Time), Arne Ansper
wrote:
>I'm sorry. It seems to me that you made a small mistake in your
>original
>post or your definition of "non-blocking socket" is different than
>other
>peoples.
>
>You said:
>
>"It is almost always an error to use 'select' with non-blocking
>sockets."
>
>All your arguments given in this letter and the next one indicate
>that you
>wanted to say:
>
>"It is almost always an error to use 'select' with blocking
sockets."
>
>(And I completely agree with this one)
Hahaha. I can't believe I did that! Sorry about that.
>> So what's the scenario where calling 'select' on a non-blocking
>>socket is useful? All that happens is if you encounter an unusual
>>situation where something changes in-between your call to 'select'
>>and your call to 'read', 'write', 'accept', or whatever, your
>>program
>>stalls in the I/O function, possibly for a very long time.
>This description fits exactly the situation when you call select on
a
>blocking socket. Kernel can remove the aborted connection from
accept
>queue and then accept will block. That why you want to use non-
>blocking
>sockets.
I don't know what circuit mis-fired in my brain. Yes, I meant that
using 'select' followed by an I/O operation on a *blocking* socket is
a problem.
Sorry for the noise and confusion.
DS
______________________________________________________________________
OpenSSL Project http://www.openssl.org
User Support Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
