[WEB4LIB] Re: Restricting launches to IP-restricted licensed

[Nancy Sosna Bohm]

It might be both more polite and aesthetically pleasing to include a bit of
javascript with a mouseOver for the links to the database sites that pops up
a self-closing window containing a message something like: "you cannot
access this database from an off-campus computer."  This would prevent
non-ISP-authenticated users from having to load and leave an entire page
with the same message.

----- Original Message -----
From: Walt Crawford <Walt_Crawford at notes.rlg.org>
To: Multiple recipients of list <web4lib at webjunction.org>
Sent: Tuesday, January 04, 2000 3:38 PM
Subject: [WEB4LIB] Re: Restricting launches to IP-restricted licensed


>
>
> Fascinating discussion, but could it be getting off-point? The question
was,
> more or less,
>
> If the vendor is using IP authentication, then should the library also be
doing
> IP authentication?
>
> The only "vendor" response so far is the same as mine would be--noting
that I'm
> not a lawyer or part of our licensing group--which is to say:
>
> No. (Don't we have a consistent answer here from almost everybody?)
>
> If you're using IP authentication and the vendor has agreed to do the
> authentication, then you shouldn't need to do additional
authentication.(If
> there are vendors who disagree, I'd be interested in hearing why.)
>
> The other question is also worthwhile: that is, should IP authentication
be the
> only option? Here again, I'd say "no"--but that may be because our (RLG's)
> preferred method is to use WebScripted access (Thanks, OCLC). In that
case, to
> be sure, it's up to the institution to assure that only legitimate users
are
> getting into the database via the Webscript. Then again, we do use IP
> authentication when that's what the institution needs...
>
> The worst case, from a vendor's perspective (and, I suppose, from the
> institution's perspective if you want databases to be available) is when
neither
> the institution nor the vendor is authenticating users. It does happen
(not
> often, I think), at least until a vendor's testing reveals it.
>