Manoj Srivastava <[EMAIL PROTECTED]> wrote: > Imagine of evryone started tweaking the header sizes of IP > packets. Heh.
As long as it's not represented as text which is standard, what's the problem? Here: ip headers should consist of 9999 10 bit bytes. Now, try sticking that into some standard with the provision that you can't represent the result as the standard. If you distributed the result, how would it be any more damaging than you distributing some piece of code which tried to implement this bogosity? In either case, some people would be hurt, which is a good reason not to do this particular thing -- and why a philosophy and practice of doing the right thing serves in this case. Once again: yes, if it's legal to distribute standards which don't allow revision then we should do so, but not as part of main. But standards which merely ask that modified results be identified as not that standard should receive preferential treatment (and be distributed as a part of main). I'm not sure why you have a problem with this? -- Raul
