"Dunlap, Randy" wrote:
>
> at http://www.usb.org/developers/usb20/
Finally!
I'm glad I wasn't putting money on when it was going to be released - I
would have guessed the USB 2.0 developers conference in May. Maybe this
way people will know what it is that is being discussed at the
conference.
Some interim observations:
1. I think I know why no-one has announced a 2.0 hub yet. What a
nightmare!
2. A less than scrupulous manufacturer could start rebadging devices
right now a USB 2.0 device, given the statement in Chapter 7 that "USB
2.0 devices are not required to support high-speed mode." USB 1.1 is
superseded by USB 2.0, and the magic words to look for on the box are
"high speed" (and of course "full Linux support").
3. The magic part about how high-speed devices are identified ("well
since full speed uses a pull-up on one data line(D+), and low speed uses
a pull-up on the other line(D-), so how will high speed be detected?")
is revealed - attach as full speed and then drop off the resister on D+
after negotiation.
4. More test procedure stuff (e.g. the signal quality test is now part
of the spec) and built in test features.
5. Changes to the protocol layer that I wasn't expecting (e.g. see split
transaction clause 8.4.2 and NAK limiting clause 8.5.1)
6. Fairly minor changes to the device framework
7. I note the requirement for a programming guide per Clause 10.6:
"As noted previously, the actual interfaces between the USB System and
host software are specific to the host platform and operating system. A
companion specification is required for each combination of platform and
operating system with USB support. These specifications describe the
specific interfaces used to integrate the USB into the host. Each
operating system provider for the USB System identifies a compatible
Universal USB Specification revision"
8. Smarter hub power management. See 11.13
9. Hub complexity will / should mostly be hidden behind the ASIC silicon
magic. Still need some changes to make use of the new features.
10. Generally looks like the existing architecture can be adapted to the
new requirements. Probably needs a new host controller driver, new (or
substantially adapted) hub driver access to device, hub and host
hardware, and about 6-10 weeks of development and test.
Has anyone established a link with NEC or NetChip (or any other
organisation in http://www.usb.org/developers/data/usb20/bbvendors.pdf)
to establish initial sampling / product availability?
Mainly for Randy: is there some way that this can be structured so we
look professional and well organised? Perhaps a Linux International
approach?
Brad
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
