From: Charles Lindsey (chl@clw.cs.man.ac.uk)
Date: Fri Sep 27 2002 - 05:23:53 CDT
In <3D939097.7050509@alex.blilly.com> Bruce Lilly <blilly@erols.com> writes:
>Charles Lindsey wrote:
>> In <3D8F950C.5050500@alex.blilly.com> Bruce Lilly <blilly@erols.com> writes:
>>
>>
>>>Starting from some baseline, the steps might be:
>>
>>
>>>1. MUST NOT generate, SHOULD accept (no damage caused, prepare
>>> for generation in next release)
>>>2. SHOULD NOT generate, MUST accept (notice to developers that
>>> full implementation is imminent)
>>>3. MAY generate, MUST accept (full implementation)
>>
>>
>> I don't see why it should take three steps. Why is step 1 necessary?
>For interoperability, which cannot be expected if some
>software generates something corresponding new syntax before
>other developers have had a chance to prepare receivers for
>that syntax. The "SHOULD accept" notifies the latter to
>prepare to receive, while the "MUST NOT generate" prevents
>problems during the transition period.
A "SHOULD NOT generate" is adequate for that purpose. See the
correspondence between Henry and John for the meaning of "SHOULD NOT".
Essentially it means that you use it at your own risk. If you know that
your message is only going to places that have already upgraded, or if you
are prepared to have it dropped at whatever percentage of sites you
believe are still not upgraded, then you generate the new form.
In practice, it means that the vast majority of sites continue to generate
the old form.
-- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clw.cs.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5