Subj : The Nodelist Revisited
To   : Jasen Betts
From : Scott Little
Date : Wed Jan 01 2003 08:15 am

[ 30 Dec 02 16:57, Jasen Betts wrote to Scott Little ]

SL>> I'm not "stubbornly championing one particular form" - I simply
JB> Are you prepared to consider any other forms?

HRN is still an option, and I agree there are valid reasons to use it.  It's
just that I haven't seen them argued - people have come up with all sorts of
nonsensical paranoid excuses of why XML is bad instead.

JB> I'd have hoped for some improvement atleast :)

Well, for one, SLF segments that originated from XML (by *Cs using XML
processors) will obey the rules coded into the conversion algorithm.  Plus, *Cs
may opt to convert incoming SLF to XML, merge in their local segment, then
convert back to SLF (if their uplink is SLF-only), rather than run an SLF
processor in parallel (by converting their XML to SLF and injecting it into the
SLF processor).

Done at a high level, these conversions will likely clean up much of the
nodelisted data, perhaps at the cost of pissing some lower *Cs off who are
currently doing their own thing.


-- Scott Little  [fidonet#3:712/848 / [email protected]]

--- FMail/Win32 1.60+
* Origin:  Cyberia: All your msgbase are belong to us! (3:712/848)