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

Bug#697421: [debian-devel:18459] Bug#697421: ITP: unidic-mecab -- free Japanese Dictionaries for mecab


I thought about this package at one point but I did not since its real
upstream Japanese government agency had very restrictive license.
  http://www.tokuteicorpus.jp/      (Japanese pages)
  http://www.tokuteicorpus.jp/dist/  Version 1.3.12
  UniDic Copyright © 2007-2010 DEN Yasuharu, YAMADA Atsushi, OGURA Hideki, KOISO Hanae, OGISO Toshinobu

It states: personal use only, no distribution, ...

So this was not even good for non-free.

On Sat, Jan 05, 2013 at 09:27:08AM +0900, Hideki Yamane wrote:
> Package: wnpp
> Severity: wishlist
> Owner: Hideki Yamane <henrich@debian.org>
> X-Debbugs-CC: debian-devel@lists.debian.org, debian-devel@lists.debian.or.jp
>    Package name: unidic-mecab
>         Version: 2.1.1
> Upstream Author: The UniDic Consortium
>             URL: http://sourceforge.jp/projects/unidic/

This is kind of new site.  Seems to be provided by one of the upstream
author of original government supported one.

>         License: BSD-3-cluase
>                  LPGL-2.1
>                  GPL-2

With much better licenses :-)  But considering "Upstream Author" is
still Govenment one, we better make sure the current license situation
to keep clean hands.  Sourceforge license review may not be good enough
for Debian.

> Description: free Japanese Dictionaries for mecab
>  unidic-mecab is a Dictionary for MeCab, Japanese morphological analysis
>  implementation.
>  .
>   * All entries are based on the definition of "SUW (short-unit word)" that is
>     specified by NINJAL (The National Institute for Japanese Language and
>     Linguistics), which provides word segmentation in uniform size suited for
>     linguistic research.
>   * It has three-layered structure with
>      - lemma
>      - form
>      - spelling
>     And it can provide a clear distinction of two types of word variant: 
>     spelling variant and form variant.
>   * It is useful for research of Speech processing since it can be added
>     accent and shift in sound information.

It does not mention anything about efferts taken to put these as above
license.  Or is it written in the package.

Please check with Mr. Toshinobu Ogiso how he got clearance from
Government agency first.  Initial discussion may be better done first in
debian-devel@lists.debian.or.jp in Japanese if we discuss situation in
Japanese ...


Attachment: signature.asc
Description: Digital signature

Reply to: