Skip to main content.
home | support | download

Back to List Archive

Re: Fw: Re: 8-bit chars

From: John Angel <angel_john(at)not-real.hotmail.com>
Date: Tue Dec 16 2003 - 07:24:35 GMT
Yes Dave, it will make swish-e better tool. ht://dig already does this
perfectly.

The problem you are mentioning is theoretical only. Proposal is just simple
upgrade of TargetCharset implementation (locale attribute in ht://dig) and
will not be default setting. The users should be warned for cat&dog case,
but since we don't have utf-8, that is the ONLY multilanguage solution for
now.

Of course, if utf-8 support will be implemented soon, forget about all this.


----- Original Message ----- 
From: "david moreau" <davidjmoreau@hotmail.com>
To: "Multiple recipients of list" <swish-e@sunsite.berkeley.edu>
Sent: Tuesday, December 16, 2003 02:08
Subject: [SWISH-E] Re: Fw: Re: 8-bit chars


> Open source means if you want a feature, you can implement it. But I think
> Bill probably has more urgent things to fix. Time is a limited resource
and
> every feature implemented involves opportunity costs.
>
> The main problem I see is that search engines should send relevant and
> complete results. In order to get such results using the scheme proposed,
> you need to know numeric representations for each letter in each encoding
> and map them. Otherwise, when a user types 'dog' in you web site and
clicks
> on search, you might miss many relevant 'dog' documents while retrieving
> many irrelevant 'cat' documents (I'm alluding to the earlier example).
>
> Is that going to make swish-e a better tool?
>
> dave moreau
>
> Bill wrote:
> >> We agreed that utf-8 is the right thing, but who knows when it will be
> >> implemented.
> >>
> >> I repeat the question - what is the alternative until utf-8 support is
> >> implemented? You don't have one. Proposed solution is something which
can
> be
> >> used in the meantime.
> >
> >Ok.  Send the patches.
>
>
>
Received on Tue Dec 16 07:24:41 2003