On Tue, 25 Oct 2005, Glenn Maynard wrote:
> On Tue, Oct 25, 2005 at 10:44:16AM -0700, Greg KH wrote:
> > I saw that you mentioned that we don't need the address, as it's in the
> > directory name, but I would argue that we should have it anyway, as it's
> > hard for some shell languages to parse out the address from the
> > directory name.
>
> It's also much cleaner, from a user standpoint, to treat a sysfs directory
> as a set of named properties with a path, than special path-specific unnamed
> properties only accessible by parsing the path. "get_property(path,
> 'bAddress')"
> is better than string parsing in just about all languages.
I feel argumentative this morning, so here goes...
> "get_property(path, 'bAddress')"
> is better than string parsing in just about all languages.
"Better" is a matter of opinion. "get_property" requires more computer
resources (transfer from user mode to kernel mode and back to perform the
I/O).
Besides, you've ignored a fundamental point. Where did "path" come from
in the first place? In any reasonable scenario, you will have something
like this:
path="${intf_path}/ep_${address}"
So then how is "get_property(path, 'bAddress')" any better than simply
"address"?
Alan Stern
-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.
Get Certified Today * Register for a JBoss Training Course
Free Certification Exam for All Training Attendees Through End of 2005
Visit http://www.jboss.com/services/certification for more information
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
