Talk:BPFK style guidelines for Lojban dictionary definitions (proposal draft)

• Each meaning definition should be clear, easy and comfortable to read for everyone, and not cluttered with slashes, obscure abbreviations or anything else hindering readability or comprehension. Lists of appropriate translations or glosses are preferably put into a dedicated dictionary entry field, separate from the main definition.
  • This can't be done for complex terms describing math, physics etc.La Gleki (talk) 01:27, 26 djulio 2015 (PDT)
• Each meaning definition should be simple, made from simple terms, even if that leads to more verbosity.
  • This immediately leads to the problem of where to put synonyms. We can remove "combine" from the definition of jmina considering it not a simple word but then newcomers will ask "how to say "combine" in Lojban?" and they won't have this opportinity to search for this word. This problem can be solved by changing the database model allowing for production of Natlang=>Lojban dictionaries. This will also stop unproductive way of relexing English in Lojban via creating zillions of new brivla.
    • Synonyms should go into the gloss fields. And indeed, having Natlang→Lojban dictionaries would be great, but it sounds like a huge work. --mi'e la .ilmen. mu'o 07:13, 28 djulio 2015 (PDT)
      • Consider removing "compare" form jmina. You can't put the whole "x1 combines x2 with x3" into as glossfield since synonyms have place structures. La Gleki (talk) 07:29, 28 djulio 2015 (PDT)
  • Also it's highly subjective what is simple and what is not. La Gleki (talk) 01:27, 26 djulio 2015 (PDT)
    • Yeah, I've pretty much copied this point from Naours' proposal; it should probably be either clarified or removed. --mi'e la .ilmen. mu'o 07:13, 28 djulio 2015 (PDT)
• The appropriate gloss words are recorded in the gloss fields.
  • I'd rather shut down glossword completely. They lead to complete confusion of newcomers who start thinking that {mlatu} means "cat" whereas instead it means "to be a cat". {mi viska mlatu} is a normal consequence of this glossword system. Instead, a natlang-specific system has to be developed. For most SAE language this is just solved by creating two types of translations: for verbs and for nouns. {lo mlatu} becomes a noun, {mlatu} becomes a verb. Again this is about Natlang=>Lojban dictionary, it shouldn't have any relation to the current JVS's Lojban=>Natlang direction as the only option.La Gleki (talk) 01:27, 26 djulio 2015 (PDT)
    • Well, gloss words should be split into argument slot glosses (e.g. dunda-x1: "donor") and predicate glosses (e.g. dunda: "give"). The former corresponds to noun forms, the latter to verb forms. {lo dunda} are the things that satisfy dunda's x1, so they can be glossed as "donor". pei --mi'e la .ilmen. mu'o 07:13, 28 djulio 2015 (PDT)
• The etymologies are presented along with the definition in order to help retaining the term and its origin.
  • Completely disagree. They should be presented along with meaning. Why should we copy etymology for every definition in every language??? More on this: https://github.com/lojban/jbovlaste/issues/131 La Gleki (talk) 01:27, 26 djulio 2015 (PDT)
    • I actually agree with you there (see the page's footnotes); this "along with the definition" was a careless mistake. I'll fix it immediately. --mi'e la .ilmen. mu'o 07:13, 28 djulio 2015 (PDT)
• The various sumti types must be specified between parentheses: proposition (du'u), event (nu), property/relation (ka), quantity (ni)…
  • If so these parentheses must be used only for sumti types and nothing else and come immediately after "xN"