Re: [xml2rfc-dev] [Ext] v3 <seriesInfo> vs IMPNOTES vs canonical XML

["John R Levine" <[e-mail]>]

> The current canonical RFCs in XML don't match any published spec. This 
> leaves us in the situation of having to publish a new spec that might 
> have escape clauses ("the XML might contain X or Y for this element") 
> and then later publish a more definitive spec that says what we really 
> want. Or, we can admit that we made a mistake with the current canonical 
> XML and fix them when the revision to RFC 7991 is finished. This is 
> related to the decision about what the version number for the revision 
> to RFC 7991 will be, given that the current canonical XML is not 
> honestly reporting the version number.

One of the major reasons we moved from free text to XML is that it lets us 
write a well defined specification for what's in the XML.  That makes it 
possible to write translators into display formats, and write future 
translators for display formats that haven't been invented yet.

I would be dismayed if we had the XML equivalent of early Postscript RFCs 
where the Postscript is missing fonts and modern software can only guess 
at what it was supposed to look like.

Ten years from now when our successors have some new haptic accessible 
output devices, they should be able to write a converter and run the old 
XML through it to support the new devices.  If the XML changes from one 
RFC to the next, we'll have made that gratuitously much harder.

At this point I have no strong opinion about the best way to fix it, but 
we do need to fix it.

R's,
John