[WEB4LIB] Re: WebSPIRS 5 bug alert

[Dobbs, Aaron]

Correct, logging out avoids the problem.
We should be encouraging folks to log out anyway 
(to dispose of search histories if nothing else)
-Aaron
:-)'

-----Original Message-----
From: Jeff Newman [mailto:jeff.newman at utoronto.ca]
Sent: Thursday, June 13, 2002 11:29 AM
To: Multiple recipients of list
Subject: [WEB4LIB] Re: WebSPIRS 5 bug alert


The University of Toronto had the same problem and realised that the
problem with moving between databases is easily solved if you encourage
people to use the "logout" button in the top-right corner of the
screen.  This ends the session and allows the next index to load without
incident.  Not perfect, but it'll get you through until the patch shows
up.

Jeff Newman


Stacy Pober wrote:

> Links to Webspirs 5 versions of Silver Platter databases only
> work correctly the FIRST time the user clicks on a Webspirs
> database within a session.  If the user then goes back to our
> database menu page and clicks on the link for any WebSPIRS 5
> database different from their original choice, it will simply
> send them to their initial choice.  The only way to prevent this
> is to completely close the browser down and then re-open it
> between clicking on different databases in version 5 of Webspirs.
>
> Worse yet, if the database used has multiple files (as is the
> case with most of the Silver Platter databases we get), the
> patron may not even realize they are in the  wrong database.
> The only place that the name of the database would show up is
> the database choice pull-down menu, and it will read:
> "View databases being searched".
>
> Unless the user opens that menu, they may not realize they're
> still in their initial database, not the one they selected
> subsequently.
>
> This bug does not show up if you use Microsoft's Internet Explorer
> v. 6.0 but it does occur in Netscape 4.79 and Mozilla 1.0 Release
> Candidate 2.  I don't have Netscape 6 on this machine, so someone
> else will have to experiment.
>
> Our main menu of databases is:
> http://www.manhattan.edu/library/palsgate.html
>
> We're currently linking to all three versions of WebSPIRS.  Other
> than the lack of a good signal for the database choice, I was
> relatively pleased with v.5 until finding this bug. (And of
> course, it showed up when I was demonstrating the use of the
> database menu to an entire class.)
>
> Silver Platter says this will be fixed in October.  Considering the
> amount of user confusion it could cause, I would really like it if
> they would act a bit sooner on this issue.  I tried posting this to
> the SPIN-L database, but I cannot seem to post directly to it.
> I see non-SilverPlatter employee posts there every once in a while,
> so I don't think it's an announce-only type of list.  Is there some
> trick in posting to that list?
>
> At this point, someone's bound to pipe up with advice that we should
> ditch Netscape entirely.  Certainly it looks like the world is going
> in that direction.  I maintain a database of free online journals
> < http://ejournals.manhattan.edu > and find that a significant
> percentage of web developers don't seem to be testing their pages
> in Netscape at all. I've come to some sites that have javascript
> pop-ups that warn you the site won't work at all or correctly in
> Netscape, and more sites don't work in some essential way but
> give you no warning of this.
>
> Still, I think that as long as a significant portion of our
> users have Netscape on their computers, our sites should be usable
> in Netscape.  I don't care so much if they're not as attractive,
> but all the essential navigation should be available.  And even
> if Netscape is not completely compliant with all web standards,
> isn't Mozilla supposed to be pretty good in that respect?  When
> a page doesn't work in Mozilla, I suspect the problem is in the
> page design rather than the browser.
>
> --
> Stacy Pober
> Information Alchemist
> Manhattan College Libraries
> Riverdale, NY 10471
> http://www.manhattan.edu/library/



*********************************************************************
Due to deletion of content types excluded from this list by policy,
this multipart message was reduced to a single part, and from there
to a plain text message.
*********************************************************************