Dear Mike,
I thought it was a wonderfully valuable workshop as it brought me in contact with 
a wide group of people with whom I wouldn't otherwise have interacted. I was 
particularly amused by different perceptions of education and training communities. I 
believe that these areas have much in common even though they have different tradeoffs. 
I certainly recommend that you continue and extend breadth of involved communities.
I thought that in general, the major role played by IMS was good as it gave a 
reasonable focus to discussion. This focus was actually helpful even when in some of 
technical sessions, I though that the IMS approach was somewhat flawed or perhaps 
better incomplete
I hope you find it not too impertinent if I suggest some refinements in the 
organizations of the working groups. Here I can comment on the 4 technical working 
groups that I attended.
I would consider the current "learning object" group has rather flaky. The kick-off 
leader of this (M.Pettit) is obviously brilliant technically as I think he had a major role in 
very interesting interoperability demonstration (Perl, Java, ActiveX, and CORBA) on 
first day. However I thought he did not show enough breadth to be able to be effective as 
leader of this group. I think in fact that IMS's great strength appears to be more in the 
"information science areas" (as in Metadata and Learner Profiles) and less in areas such 
as learner objects where some of the key issues of interactivity and where some of most 
rapidly changing technology is relevant.
More generally, I suggest that the key groups be led by people without 
preconceived agenda and with broad perspective. Of the four technology groups, I 
thought the one led by P. Dodds (outside IMS I think?) was clearly the best as he came to 
the meeting with an open mind and was willing to question approach of IMS. The other 2 
group leaders (Durrance and Rhodes) were very professional and on top of their material. 
They did not stray too much from IMS party line but this might in fact be appropriate, as 
especially for metadata, the IMS work here is quite thorough and did not seem too 
controversial. 
I also thought the last presentation by Wes Region also illustrated difficulties with 
working group chairs with their own agenda. His summary was presented well but 
seemed to me to be largely his personal interests.
So please keep up the good work. I would advise ADL to maintain some 
independence from IMS and only rely on them to develop specifications in areas where 
IMS have clear strengths.