Linas Vepštas

Opened #3351 to cover openpsi documentation. Unrelated to the documentation is "dynamics" -- this was the part that was tangled up, back when I worked on it. In particular, the...

Using the "value" field, on the atom? I thought I went through rest-api very throughly, and removed all usage of value, a while ago, because I knew it was broken....

ugh. well, I believe that the python operator== and != should work just fine for atom compares. I hope it does not need a less-than/greater-than?

Perhaps `SequentialAnd` instead of `ListLink` might be more obvious. I too had needed this, at one point, for language.

There's another thing openpsi doesn't currently do: allow variables in the contexts. I briefly had a version that allowed that (I needed them for language) but apparently this caused problems...

OK .. so `AndLinks` are evaluatable. This one has a problem though: not all of it's contents are. So its kind-of syntactically broken. (The equal link is evaluatable, the inheritance...

BTW when I said "OpenPsi doesn't allow variables in contexts", I meant "it doesn't know how do do anything useful with variables in contexts, and will probably generate garbage as...

> But sometimes one may want to reuse some of the variable groundings in the action, and it wasn't possible unless they were all evaluated at the same time. Um,...

> let me try to explain the situation... Those paragraphs belong either in a README, or (better yet?) the wiki. That way, you can just say RTFM.

This does not make sense, as constants cannot be variables. ``` (VariableList (VariableNode "x2") (ConceptNode "Required constant for DualLink-2") (VariableNode "z2") ) ``` Note that VariabeLists are often optional: if...