[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] {cukta} as a case study for why the dictionary needs clearer, more detailed entries



I'm honestly not entirely sure I understood everything in the Issue you linked, gleki... But apparently we're already using GitHub for la jbovlaste -- at least for tracking the code for the website itself. So that's good. Whether GitHub tracks the actual database or whether it's only used to tracking ongoing issues, I think it will help.

What format is the database in right now?

The suggestion before had been to consider an XML-based format actually used by linguists, because we could then also use any other tools designed to work with that XML spec. I know HTML well enough that picking up whatever XML format wouldn't be too hard. And there are plenty of tools for working with the usual website databases (like MySQL, or CouchDB, or whatever). But I'm not really expert enough to say which database format would be better for our purposes. I'm just trying to facilitate the discussion, so we can move forward.

~Andrew / cemjig


On Monday, February 15, 2016 at 10:36:55 PM UTC-8, la gleki wrote:


2016-02-16 1:59 GMT+03:00 Andrew <summerf...@gmail.com>:
I forgot to mention that the proposals I mentioned in my last email were actually not my ideas, that I was just summarzing some of the discussion I saw while "lurking" on IRC. I do like the idea of using GitHub, or something similar, though. I think we need some kind of online tool to help us collaborate on this project, something which will help us keep track of ongoing issues and so on. It doesn't have to be GitHub necessarily, but we are already using GitHub to track the cleanup and development of the CLL, so I think GitHub would be a sensible choice.

gleki said:
 
Please note that jbovlaste has wonderful features that Github itself won't support out of the box...

That's a fair point: Whether we ultimately decide to replace or merely revise la jbovlaste, it would be helpful to make note of the good features we would like to keep. For instance, I think it sort of interesting that the different dictionary interfaces -- like vlasisku and sutysisku -- are able to share common data on the backend but present a different user interface. Which features were you thinking of, gleki?


This is a dirty explanation of how i envision the db.
If you want to use github as a database that would be fine of course although strange. This of course wouldn't eliminate the need in an API backend interacting with both frontends and the database.
 

I think Timothy Lawrence and la lojbab are both right about cmavo in the dictionary: Learning cmavo and how they really work is essential to learning the structure of the language, and a detailed explanation of the grammar of the cmavo is probably outside the scope of a dictionary. However, the cmavo dictionary entries, just like the gismu, could still use some clearing up, and links to relevant sections of the CLL seems appropriate. Some entries, in fact, already link to the CLL. I'm sure online English dictionaries would do the same, if there were a such thing as a Complete English Language book!

Thanks,
~Andrew / cemjig

On Monday, February 15, 2016 at 1:16:02 AM UTC-8, lojbab wrote:
On 2/15/2016 12:20 AM, Timothy Lawrence wrote:
> As a user, I would like links to the relevant CLL page (or other
> webpage) that explains each word in context with its grammatical
> category.* (For me, the smaller "grammar words" are a higher priority
> than gismu etc.)
>
>
> For example, the jbovlaste description of "pu" or "zi" really only makes
> any sense if one has already read
> http://www.lojban.org/publications/reference_grammar/chapter10.html
>
>
> I'm struggling trying to read real-world Lojban text because I can't
> understand the grammatical constructs that aren't explained well on
> jbovlaste. It's too difficult for me to read through the entire CLL (and
> I've been discouraged by it being out-of-date) front-to-back.


When byfy was started more than a dozen years ago, its primary purpose
was to do precisely what you are asking for.  For a decade before that,
I as leader of the organization had been stymied in trying to produce a
dictionary because I couldn't figure out how to write dictionary entries
for the cmavo (the grammatical words you refer to).

byfy did do a lot of work on this, and the web pages are still out
there, and many have precisely what you are asking for: an explanation
of each word in context with its grammatical category.  Find the
relevant pages and your needs are met.

Not all the words were done, because in some parts of the language, it
is simply too difficult to explain and/or find examples of usage.  And
the volunteers became bored and distracted.  byfy slowly declined in
productivity and eventually stopped for months or years at a time.

And that sort of text still is not a dictionary entry or definition.  If
jbovlaste wants to pretend to be a dictionary, it needs dictionary
definitions and not grammatical explanation.  It's not an easy problem
to solve, and after a couple of decades of trying, I'm not sure what
solution would satisfy everyone.  Take a look at a dictionary definition
for "the" or "of" (and actually "or" might be a could one, because
English "or" sometimes is inclusive and sometimes exclusive and
describing that definitionally is difficult.

in any event, CLL's "out-of-date-ness" is not so bad that it should
discourage you.  If you learn the language as described, you'll almost
certainly be understood. There might be some confusing spots, but they
are mostly minor.  Remember that most of the people who have learned the
language have learned using the original CLL edition, and that will
likely be true for years even after a revision is published.  Partly due
to my age and health, I don't ever expect to learn any version of the
language significantly different from that described in the current CLL.

lojbab

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.
To post to this group, send email to loj...@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.