search

zippy / ceptr

(a recomposable medium for distributed social computing) || (semantic self-describing protocol stacks)
http://ceptr.org
GNU General Public License v3.0
89 stars 17 forks source link

fix the bogusness of LIST& TREE structured symbols #48

Closed zippy closed 8 years ago

zippy commented 9 years ago

All symbols whose structure is LIST (like SIGNALS and PENDING_RESPONSES) are an indicator that we've got something going wrong. It maybe that these are instances where we are using Symbols to manage collections which should be managed by scapes, or and indicator that we should be using optional element structures (see #8)

zippy commented 8 years ago

In some cases LIST & TREE structured symbols are just ones missing optionality being added by #8

zippy commented 8 years ago

Also, one of the types of optionality we may need, is, like in process signatures, the optionality of an unknown symbol type, but a known structure type. e.g. the current implementation of PROTOCOLS would require declaring a LIST of PROTOCOL structured symbols.

zippy commented 8 years ago

as of 133449ab00 we can define SYMBOL_OF_STRUCTURE in structure defs (with % symbol in docs and base_defs)

note this leads to lots of fixes of ASPECT definitions including converting the aspect symbol in signal envelopes to ASPECT_IDENT of structure SYMBOL

zippy commented 8 years ago

we still need to finalize how we will specify ANY, currently we just use the TREE structure

zippy commented 8 years ago

as of 0a59413a8 you can specify ANY with STRUCTURE_ANYTHING (or ! in the base_defs language). So this allowed full removal of LIST. Still to clean up : TREE

zippy commented 8 years ago

as of 2a09dec151482 TREE is now only used to specify orthogonal tree structures as it should.