On Fri January 14 2005 04:38, Arnt Gulbrandsen wrote:
[G]roup syntax has been so little used in the past years that
IMO it's better to kill the syntax than to try to support it.
Bad premise; group syntax is used quite a bit.
As a
feature, group syntax simply isn't carrying its load.
The capabilities have been somewhat limited since RFC 822
discontinued the provision for nested lists which existed
in RFCs 724 & 733.
IMO, the best option would be to issue a short RFC saying:
1. Mail senders SHOULD NOT generate group syntax.
I would not support that.
2. Mail receivers are urged to test that they can parse empty groups,
since several current senders generate that, e.g. "To:
unknown-recipients:;" and "To: ietf-announce:;".
Of course implementers should test their parser
implementations against legal content. RFCs 724, 733, 822,
and 2822 have all spoon-fed prospective implementers with
examples containing list syntax among other things.
Comments?
If you want to write a "Testing Considered Useful" draft,
go ahead; but I doubt that any implementers who have
ignored the examples contained within the specification
RFCs will wake up and take notice.