On Wed, 1 Sep 2004, Steve Calfee wrote:

> Yes, not much difference.
> 
> The strange thing is that even though windows knew exactly how long the 
> configuration packet was (after reading the first 9 bytes), it asked for 255 
> bytes! This implies to me that some early devices incorrectly returned the 
> configuration length and windows figured out the actual length by analyzing 
> the data.

Yes, the Zoom modem trace shows that.  On the other hand, the keychain
trace shows different behavior.  The inescapable conclusion is that
Windows has multiple code paths for retrieving the configuration
information.  That's not too surprising; I've seen two different
strategies used (even within a single trace!) for reading string
descriptors.

By the way, I forgot to mention it before, your traces don't show a port 
reset occurring after the first Get-Descriptor.  Is this a result of your 
post-processing the HTML trace document?  It's practically certain that a 
reset did take place.

Alan Stern



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to