My post-GTTSE'2005 paper is ready. Abstract:
Hierarchical data could be viewed and processed as XML using the SXML format and Scheme language. We introduce a symmetry constraint on this approach, reveal the weak points of the SXML representation, and discuss mapping between XML and SXML.
The main goal of the paper is to be a checklist and food for thought for implementors. It summarizes my experience in this projects:
At the moment, the paper isn't available online. I'm waiting for acceptance of rejection. But I can e-mail a copy, if you are interested.
Surprisingly, it was very hard to write the paper. It took a month, and it could take more unless deadline. As result, I haven't used a kind proposal of Martin Bravenboer to send him a draft for review.
I'm not sure the paper will be accepted. But it's not very important. Important is that I described unexpected surprises of implementing XML virtual machine, and a next implementation will be better.