drafting “x:value-of”
Thursday, March 31st, 2005Thinking about testing XML to SXML conversion to pass the second milestone, I found out that I need functionality that was planned for the future. So I temporary switched to the phase three.
Thinking about testing XML to SXML conversion to pass the second milestone, I found out that I need functionality that was planned for the future. So I temporary switched to the phase three.
(This post was written yesterday, but due to internet outage I’m posting it only now.) Exactly as I expected, implementation of conversion of attributes and namespace definitions has taken one working block. The second milestone is now near.
Other two bugs (see “bug list” and “bugfixing 1“) are interrelated. Use of the map is powerful, but dangerous.
In the previous entry I mentioned three bugs to be fixed. In this post I describe the second issue.
Surprisingly, conversion from libxml2 tree to SXML seems a trivial task now. Only attributes and namespaces are left.
Using a global variable is a bad programming style, isn’t it? But I’ve just used it, and I’m sure it’s a right thing.
Recently I got a complaint from an user that my FrameScript code did not work. Understanding the problem was a valuable experience.
Now I’m the owner of the quite good domains “tohtml.com” and “2html.com”. I’m surprised why they were free.
I’ve submitted an application for participation in the Summer School on Generative and Transformational Techniques in Software Engineering.
While communicating about making TeXML distribution archives, I’ve got examples of powerful use of “setup.py”.
I wanted to work on generative XML, but I had to answer a letter from Paul Tremblay about new release of TeXML. Answering developed into working on TeXML.
All tests are passed, and now I think that conversion from SXML to libxml2 is working and stable.
Namespace support in the SXML to libxml2 conversion is completed.
Yesterday evening, trying to write the previous post, I found that my blog isn’t working. It’s the third computer failire in the last days (here are the first one and the second one). Have I bad carma?
Before starting fixing the default namespace problem, I’ve created a test framework and added several tests. The results are as usual.
And although not supporting the whole Scheme standard, the compiler supports fully optimized proper tail calls, continuations, and (of course) full closures. http://www.iro.umontreal.ca/~boucherd/mslug/meetings/20041020/minutes-en.html
I never understand why, having such a victimable system as DNS, Internet is still alive. Now my understanding is giving up.