From: "Marshall T. Rose"
vern - i hope we can agree that i don't fall under the categorization of
those who have not yet found an opportunity to Contribute To The Standards
if we disagree on this, then discard this message.
No, you're one of those others who've written tools that might support
a transition to XML and whose silence I found relevant.
i will tell you why i think it is a good idea that the I-D repository accept
XML versions of drafts, whilst retaining the .txt versions as the "official"
ones (assuming that the term "official internet-draft" isn't oxymoronic).
so, i suggest a simple experiment: let the I-D repository store alternative
versions of drafts in addition to the .txt versions. try this out for 9
months and see if people find it useful or not. is this really asking so
Before the experiment of XML in parallel with .txt for RFC's, please
consider this one. Let the .txt file continue to be the only version in
the ftp.isi.edu/in-notes directory, but allow the .txt version to contain
a "pointer to an XML version that was used during the preparation of the
.txt version. " (weasel were words carefully chosen to allow the XML
version to be anywhere and the pointer to go stale.)
There is a use for XML or nroff versions of I-D's (but not RFC's) that
has not been mentioned much (maybe first in your mention of "ASCII memos
can't be reformatted"). It saves lots of work to exchange editorial changes
as deltas to a mark up language version. Also, if the mark up version of
some drafts had been public, a very bitter war or two in a working group
or two would have been short circuited, or at least fought on the real
grounds instead of stuff about theft of nroff under false pretenses.
Perhaps in other words, allow XML in ftp.isi.edu:internet-drafts but
not in ftp.isi.edu:in-notes
The danger is the slippery slope leading to the full panoply of
Microsoft document forms. Within weeks, there'd be demands that since
XML is allowed for drafts, full MSWord with animations must be
allowed--no--required for RFCs and who needs the crufty old ASCII?
(This would not be Microsoft's fault, at least not directly.)
Vernon Schryver vjs(_at_)rhyolite(_dot_)com