RE: [edict-jmdict] pos/misc entities

[Stuart McGraw <smcg4191@********>]

Jim Breen wrote:
> 2008/6/13 Stuart McGraw <smcg4191@frii.com>:
>
> > I don't mind dealing with inconsistencies if they are present
> > for some good reason, but the inconsistent representation
> > of POS values seems to be historical accident with no other
> > redeeming value.
>
> For me the "redeeming value" is that it converts more easily
> into a readable form. To me:
>
> 甘い （うまい） (adj-i) (1) skillful; clever; expert; wise;
> successful; (2) delicious; appetizing; appetising;
> (3) fortunate; splendid; promising
>
> looks much better than:
>
> 甘い （うまい） (adj-i) (1) skillful; clever; expert; wise;
> successful; (2) (adj-i) delicious; appetizing; appetising;
> (3) (adj-i) fortunate; splendid; promising

Yes, I agree for human presentation including EDICT that
repeating POS are undesirable.  But for JMdict (primary
purpose machine parsing) I don't think that ease of UI
generation should be the primary concern.

Re the more serious (IMO) (in)consistency issue, as I pointed
out in my response to Jean-Luc, it currently seems to bleed
through to wwwjdic (and I presume edict) producing precisely
the effect you say you don't want -- see この間, or 忠実 for
example.