On Mon, Apr 24, 2017 at 8:25 AM, Martin Bjorklund <[email protected]> wrote:
> Hi, > > Andy Bierman <[email protected]> wrote: > > Hi, > > > > https://tools.ietf.org/html/draft-bjorklund-netmod-yang-tree-diagrams-00 > > > > > > The introduction in this draft does not really reflect the goals that > need > > to > > be addressed. Perhaps: > > > > sec 1, para 1: > > OLD: > > > > This document > > provides the syntax used in YANG Tree Diagrams. It is expected that > > this document will be updated or replaced as changes to the YANG > > language, see [RFC7950 <https://tools.ietf.org/html/rfc7950>], > necessitate. > > > > > > NEW: > > > > This document provides the syntax used in YANG Tree Diagrams > > > > for version 1.1 of the YANG data modeling language [RFC7950 > > <https://tools.ietf.org/html/rfc7950>]. > > > > It is expected that this document will be updated or replaced > > > > for future versions of YANG. It should be possible for > > > > multiple independent tools to generate the same YANG tree diagrams, > > > > using this specification. > > Ok. > > > Sec. 2: > > > > The vertical bar (|) and indentation is not defined. > > Ok. > > > The tab alignment is not defined > > Isn't this covered by indentation above? > no -- the type names are not printed a fixed distance from the last field but rather they align with some column (not sure the rule for picking the column) > > > The general recursive algorithm is not defined. > > Yes, on the TODO list. > > > The mandatory vs. optional parts of the diagram are not defined. > > Nothing is mandatory; this will be an Informational document. > > So 6087bis is expected to define guidelines on what IETF documents should contain? OK I guess, but it would be better if the guidelines were for all diagrams (in the same draft). > > (e.g., groupings, notifications, rpcs) > > Yes, on the TODO list. > > Examples of all field types should be provided (e.g if-features) > > Ok. > > An ABNF for the grammar should be provided. > > I don't understand what you refer to? > > Perhaps a grammar (in ABNF) that matches the recursive algorithm, but maybe this is not possible. More issues: - node order. Document order needs to be defined. Need to construct subtrees from submodules in "include-stmt" order. - can a tree diagram ever include augmenting nodes from external modules? If so there is no real document order for multiple imported modules. Also no syntax to support it. (I would say no, this is out of scope) - is the module name line part of the diagram? > > /martin > Andy
_______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
