On Sun, Sep 12, 2010 at 3:17 PM, Robin Lee Powell
<rlpowell@digitalkingdom.org> wrote:
On Sun, Sep 12, 2010 at 09:26:05AM +0200, Arnt Richard Johansen
wrote:
> On Thu, Sep 09, 2010 at 03:31:42PM -0600, Jonathan Jones wrote:
> > Does jbovlaste warn about the sense field during word
> > submission, or only on the automatic notice as below?
>
> I tried that, but unfortunately no. As long as the word “accretion
> disk; astronomy” exists, if you try to submit a definition that
> points to just “accretion disk”, you can't get past the “Check
> Definition” part of the submission.
>
> This is a Jbovlaste bug, of course,
It's not actually; it's behaving exactly as designed. The idea was
that natlang words should almost always have senses, because natlang
words are almost always polysemous.
Furthermore, if you had to put a sense for the word once, *clearly*
it's polysemous, and having it without a sense is wrong.
Now, since then, we've moved in the direction of not caring about
the senses as much, but the second half is still true: if "accretion
disk" is not specific enough by itself, such that someone had to
enter the sense "astronomy", then "accretion disk" must be ambiguous
and should not be allowed.
Why do you think that this behaviour is bad?
-Robin
It could also be that the creator thought that the "sense" field was a required item, and filled it in for that reason, and not because of any ambiguity.