Posted by John Navas on 01/22/08 22:25

On Tue, 22 Jan 2008 21:47:53 +0000 (UTC), Ilya Zakharevich
<nospam-abuse@ilyaz.org> wrote in <fn5oa9$1g4b$1@agate.berkeley.edu>:

>[A complimentary Cc of this posting was sent to
>Nomen Nescio
><nobody@dizum.com>], who wrote in article <834de9d5e1f905d8a386067692958f4a@dizum.com>:
>> > Could you explain how latency is related to polling?
>>
>> Could you explain how it's NOT?
>
>Apparently, you never wrote any code which does polling without
>latency...

True. Because it's not possible, even in a real-time OS, which Windows
is most definitely not. There is always some latency. The issue is how
much. The design of Windows makes it sometimes large enough to matter
in the case of USB.

Apparently (clearly) you've never written any Windows USB code.

>The key is to know when data is available. This requires an adaptive
>algorithm in setting up a wake-up timer, nothing more complicated.
>After 5-6 cycles, you get the timer tuned-up, and run with minimal
>overhead and no latency.
>
>Hope this helps,

It doesn't, because like all simplistic answers, it doesn't work that
way in the real world.

I'm done with you on this thread too. Think what you want. I'm
guessing you're distantly related to Podge.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)

[Back to original message]