On 6/26/06, Bernhard Eversberg <ev_at_biblio.tu-bs.de> wrote:
> Mike Rylander schrieb:
> >
> >> Mike Rylander wrote:
> >> >> >> >
> >> >> >> Again, who does it, and why not?
> >> >> >
> >
> >
> > If you're going to complain about something not being there, let's
> > quote the entire block of text, eh?
> >
> You misunderstood. No offense was meant. I was not complaining that
> Evergreen doesn't have a certain feature. You said it does indexes
> and I couldn't find them. More likely I misunderstood you - nothing
> happens easier with list postings.
Truer words were never posted ... unless /I'm/ misunderstanding. ;)
>
> >
> > Look, I'm as guilty as anyone about having pet features, but we all
> > need to be careful when saying X or Y does or doesn't do Z, because
> > there may never have been a claim of such. And always remember the
> > sage wisdom of one (according to the #code4lib IRC channel bot, in any
> > case) David Fiander, "Working code always wins."
> >
> And I wasn't saying Evergreen isn't winning, to the contrary: I was
> saying that browsable indexes are probably unnecessary. How else could
> Evergreen be winning, not having them? How could I blame Evergreen
> for not having something I proposed for NG catalogs when it isn't one.
> (There's none around yet or this list wouldn't exist.)
>
I didn't mean that Evergreen, in particular, is winning, just that
there's no code at all for an NGC, so there are no winners as yet.
And, perhaps just a bit, it was an offhand challenge to all on the
list to bring some working code. So we're on the same wave length, I
think, just different words.
> > However, I will follow your track off course if you like. If you
> > search for "asdflkjdfqw" in Google, or any library catalog, you will
> > never get any results -- including a USEFUL browse list.
>
> Sure, but Google doesn't give you a useful suggestion for *every*
> mistake you make, and its suggestions are sometimes plain silly.
> But I liked to think that displaying an a-z index at the
> position where your term should be sitting, that might just
> sometimes be helpful. Which led me to think it might be a useful
> feature of NG catalogs. Lack of support in this forum seems to
> indicate otherwise. So be it, to the relief of all those SQLers
> out there who can never figure it out and who seem to think that
> "No results found" or whatever is good enough.
>
This is the only point that that I think we disagree on. As an
"SQLer", I have no problem with browse lists in general, and I
"figured it out" some time ago. In fact, I spent a while last week
refining and speeding up the call number browse, and I think it's very
useful.
My point is that, unlike classification numbers (Dewey or what have
you) or even local call numbers, being dropped into a browse list for
author (or even worse, title) is going to be much more confusing to a
standard patron than just trying to provide more suggestions based on
authority data and spell checking. In addition, saying "Your term
wasn't found. Please see the reference desk for help in finding what
you're looking for." -- which, obviously and admittedly, would be
better text than "No results found." -- is the fastest way to get
someone to the resource they need when the system doesn't have enough
information to provide an answer within a couple clicks.
As an example, say a user is looking for something by Mayflower
descendants. They don't know that the controlled term for the author
they actually want is "General Society of Mayflower Descendants."
Also, the library they are searching doesn't have anything by that
author, nor the authority record because the have no need for it (I'd
argue that having a full authority file and browsing it would be
useful, but that's not how things work in libraries today, and that
would be expensive). Because they did an author search for
"Mayflower", they would be dropped into an author browse list
somewhere in the Ms -- nowhere near where they need to be
So, browse lists are a GoodThing(tm) when the user knows that's what
he is getting /before he gets it/, but dropping an unsuspecting user
into a browse list after he has been happily trolling along in
relevance-ranking land is simply asking for trouble -- and
disappointed users. I would not, however, be adverse to a link that
drops them into a browse list based on authority data look ups.
> B. Eversberg
>
--
Mike Rylander
mrylander_at_gmail.com
GPLS -- PINES Development
Database Developer
http://open-ils.org
Received on Mon Jun 26 2006 - 11:30:31 EDT