> On Wed, Nov 28, 2001 at 08:17:53PM -0000, David Laight wrote:
> > Not read 802.1Q - but you can't (surely) use a protocol that adds
> > 4 bytes to the maximum length of an ethrnet frame.
> 
> Really ?
> 'man vlan' on a NetBSD system shows you:
>      To be compatible with others 802.1Q devices, the vlan interface shall
>      support a 1500 bytes MTU, which means that the parent interface will have
>      to handle frames that are 4 bytes larger than the ethernet MTU.

Humm... the things that sneak through standards these days......
Sounds like a ploy to force you to throw out all those ethernet switches
and routers you only bought last year :-)

802.3-2000 standards.ieee.org/reading/ieee/std/lanman/802.3-2000_part1.pdf
contains no notes about the sudden appearance of 'tagged' MAC frames.
Just the text (which I think goes way back):

> The contents of invalid MAC frames shall not be passed to the LLC or MAC
> Control sublayers (15). The occurrence of invalid MAC frames may be
> communicated to network management.

> 15: Invalid MAC frames may be ignored, discarded, or used in a private
> manner by MAC clients other than LLC or MAC control.  The use of such
> frames is beyond the scope of this standard.

In the old days 'invalid' frames included all those with a 'length'
greater than 0x5dc.  Now it seems we have to 'know' that the maximum
frame length depends on somw bytes the MAC layes has never looked at.

Wonder which 'invalid' frames they will de-invalidate next.

    David