Subject: Re: AbiSource, Ispell, Aspell and beyond...
From: Kevin Atkinson (email@example.com)
Date: Fri Feb 25 2000 - 21:47:32 CST
On Fri, 25 Feb 2000, Paul Rohr wrote:
> At 09:29 PM 2/25/00 -0500, Kevin Atkinson wrote:
> >No, no, no. Do YOU maintain the session and personal word lists or does
> We do.
> >When coming up with near misses can the personal and session
> >words appear as near misses?
> They are recognized as valid words. They are not included in the suggestion
Ok, this is not very good thing to do in my view. One of the very
things I DID not like about Word was that they did not appear in the
suggestions. Aspell does a VERY VERY good job at using ALL words in its
suggestions as IF they were in the main dictionary. But in order to do
this aspell needs to manage those words list. That is what a lot of my
functions are doing.
> >This is an important question so please
> >don't ignore it.
> Be nice, Kevin. I spend a *lot* of time on this list attempting to give
> thoughtful design feedback to people working on a wide variety of features.
> Sometimes I fall behind:
> Sometimes I misunderstand a question. Sometimes I run out of time and don't
> give anyone the answers they need. Sometimes people have to rephrase stuff
> so I know what they're talking about.
> It's nothing personal.
No I'm not mad at you. I have a general dislike for the world and
sometimes my hidden frustration shows. It is important due to the above
> >Lookups will defiantly be as fast as ispell if not faster.
> >A full second is only on bad cases with short words. There is also a
> >suggestion mode in aspell which is over 5 times faster but doesn't give
> >quite as good results.
> Cool. That sounds entirely appropriate. As long as we get fast, accurate
> suggestions for "easy" misspellings, I suspect users will be willing to wait
> a bit longer for the "mind-reading" mode to work.
Ok, than you can use the fast mode for the popups but you need to some how
let the user know about the better, but slower, suggestion mode. See
http://aspell.sourceforge.net/test/ for how the results fair in various
> >Will an example of how to use it help?
> Yes. That'd be a *big* help.
> It's very easy to understand how to use the 4 C calls in our existing API.
> I can easily envision adding a fifth call, for example, to report
> misspelling/suggestion pairs back to the spell engine. (Until you explained
> that feature, I had no idea you did that.)
> However, there are a *lot* more than 5 entry points in your API. ;-)
Ok, will do.
---- Kevin Atkinson firstname.lastname@example.org http://metalab.unc.edu/kevina/
This archive was generated by hypermail 2b25 : Fri Feb 25 2000 - 21:44:48 CST