[This is a try for summarizing discussions]
@all
* Living language *
I've read a bit though more pages and messages. And found that historically speaking, and according to what's in the FAQ and other LLG pages, Lojban is not immutable at all. Else it would be a "dead language" like ancient Greek (to paraphrase la .stevo.) or a pure experiment for a specific point in time. However, it is considered as "stable", even if changes may occur, because else it "would require people to re-learn things" (as some of you pointed out). Thus, changes are possible but must be CAREFULLY validated.
So in two words: "Lojban evolves". Right.
That's my point of view too. And that's where issues arise!... Indeed, validating changes is a huge job requiring time, structures, procedures, and language skills (if not more). I feel la .pycyn. is right saying it "is a very delicate balance". And la .stevo. too, talking about very difficult reintegrations of forks to the core.
* Pending organizational issues *
Moreover, la .guskant. clearly told us that current organizations are dead or almost dead. So here is a small and macroscopic summary of what I feel like pending issues (feel free to complete/fix):
- Missing info about status of people (ex: no more time to spend on the project)
- Missing "roles" (about what to do) and assignations (which one does what)
- Missing well defined and easy-to-use protocols for:
- Submission, reviewing, voting, integrating (of new official content)
- Certification (of existing content)
- Low notoriety
- More would bring new Lojbanists (and thus more manpower)
- Non-Lojbanists generally don't even know about it (= word not spread)
- Not enough time for skilled Lojbanists to do all what would be necessary to improve
- A few persons "on their own mind", blocking stuff
However, to make things move, there must be a consensus on the basics, so...
* Lojban basics *
What defines Lojban? (Or "What should never change in Lojban?"). So, if I try to sum up again from various sources, the answers are below (please tell if you disagree):
YES!
- Non-ambiguous grammar (~= monoparsing, machine parsable)
- Unicity (~= forks are experiments or proposals)
- Cultural neutrality
- Partially based on logic
- Rather stable ("rather" has to be refined)
- Usable (speakable, learnable, writable, readable...)
- Usage defines the language, as long as these "basics" are untouched.
- However the "usage" is rather a blurry situation for now...
NO!
- Immutable base grammar, and immutable base vocabulary (cmavo, important gismu...)
- Targeted to be TOTALLY based on logic
- Ideal (or "perfect") language (la .pycyn. pointed out that issue)
Maybe desirable, but impossible
- Backward-compatibility of versions (added after what la .timoteios. said)
- Fully formal and complete grammar (as Gleki and And Rosta said)
- Keep strictly what original designers wanted (because they could not foresee all potential issues)
[That concludes this first step. Please share your thoughts!]
la .sykyndyr.