ietf-nntp Thoughts on renaming X commands
Evan Champion
evanc at synapse.net
Thu Oct 3 12:33:36 PDT 1996
Ben Polk wrote:
>
> What do other people think about renaming XOVER->OVER in
> the spec?
>
> I'd prefer to document it as XOVER. There are millions of
> installed programs that use this, and no technical reason I
> can see to change it.
>
> Same for XPAT.
One of the great problems with protocol extensions is this
convention/requirement that extension name be prefixed by an X, which
makes it very difficult to bring experimental extensions in to the
mainstream.
I understand the reasoning behind prefixing the name by an X, but really
all it becomes is a nuissance (and sometimes much more than a nuissance)
later on.
I have no problem with the names being changed to remove the X (servers
can keep accepting the X'd versions), but I do have a problem with new
extensions requiring the X flag once they have been integrated in to a
product of at least open beta status.
Especially in the case of a very popular product like INN, as soon as a
new feature is available everyone races to support it, and thus it
doesn't really seem like a protocol extension any more.
[This is sort of analogous to Netscape with web browsers... Sure, the
HTML codes may not be "standard" but within a month of any new browser
release, everyone is using the new codes without even a second thought.
And in this case you wouldn't expect Netscape to name the tages <XTABLE>
and then rename them when the <TABLE> tag is accepted.]
So, my vote would be to rename the attributes and change the section
regarding naming extensions with an X.
Evan
--
Evan Champion * Director, Network Operations
mailto:evanc at synapse.net * Directeur, Exploitation du reseau
http://www.synapse.net/ * Synapse Internet
More information about the ietf-nntp
mailing list