Certain applications have been blessed with 9fans' ignorance for too long.
dict(7) and friends seemed to be becoming over-joyous.

For those interested in Japanese on Plan 9, I provide here (edict2.tar) scripts
to convert Jim Breen's extended Japanese dictionary, EDICT2, to a format
usable with dict(7). However, the simple.c format was not appealing to me,
so I decided to use the format Andrey proposed for converting DICTD
dictionaries, thus, a patch[1] to dict(7), and the appropriate indexing
application[2], are necessary in order to perform the conversion and
use the resulting dictionary (the path to the indexing app
must be specified in leginx.rc).

Parts of adict, the Acme interface to dict(7), still presume some incapabilities
of Acme that are no longer present. One I've found is the assumption that
Acme tags cannot contain non-alphanumeric characters, so adict goes through
the trouble of converting all such characters to underscores.
The simple fix:
comment lines
  /acme/bin/source/adict/adict.c:288 through /acme/bin/source/adict/adict.c:294
-- by the way, the source distributed with Plan 9 by default is not what the
binary adict was built from; updated sources (though still lacking the
above change)
are in /n/sources/patch/applied/adict-fix2

However, I find Russ Cox's idea of making Acme interfaces entirely
scriptable, even
more appealing. For that, his acmeevent[3] from P9P is entirely
portable to native
Plan 9 -- no changes required. I believe the prime example he's used is a
scripted version of adict, also available in P9P -- for this, I've made a few
changes to the script for use on native Plan 9; see nadict.rc.


by the way, has anyone looked into the venti problem Oleg F. posted about,
on my behalf, earlier this month?
ak


[1] http://mirtchovski.com/p9/dict/dict.patch.tgz
[2] http://mirtchovski.com/p9/dict/mkdictd.c

[3] http://swtch.com/usr/local/plan9/src/cmd/acmeevent.c