On Sunday 30 September 2007, Marcin Miłkowski wrote:
> Maybe for some languages fonts are harder than dictionaries but for
> Finnish, we haven't got a single good open source dictionary yet - it's
> the language that is too hard for most spell-checking.

Not related to the discussion you were having, but I have good news for you on 
this particular matter. As of December 2006 there has been an open 
source "dictionary" for Finnish that is of pretty good quality:

http://kaino.kotus.fi/sanat/nykysuomi/kotus-sanalista-v1.tar.gz

This contains 94 110 entries and has been released under the LGPL by the 
Research Institute for the Languages of Finland. As far as I know (which is 
not much), this is the same stuff they license commercially for language 
technology companies, so having it available under the LGPL is something I am 
quite happy about.

Unfortunately that is not a hunspell dictionary but a XML file with words and 
inflection classes. This is not yet enough to build a spell checker. For this 
purpose we are developing (under the GPL) software and dictionary database 
that also takes advantage of the above mentioned data and our own additions:

http://voikko.sourceforge.net/

This software is already recognising over 99 % of words in typical Finnish 
text and is starting to slowly replace the previous non-free spell checking 
alternative for OOo. It is shipped in some Linux distros too, at least in 
Debian, Ubuntu, and Mandriva.

Harri

[I also replied to the mail from Robert Ludvik where I explained how we are 
handling user feedback when developing Voikko. But the mail seems to have 
gone missing, maybe I sent it from a wrong address. I'll resend if it does 
not appear here soon.]

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to