[OPEN-ILS-GENERAL] SPAM: Some questons about Evergreen

[Don McMorris]

I'm going to answer specifically on #11, and will also "throw in" a
little bit of information that may be useful.

You do not _have_ to contribute to development to use Evergreen.  The
(majority of the) system is licensed under the GNU General Public
License.  If you want to use it yourself without return, that's
acceptable.  Of course, if you _could_ make contributions, that would
be cool! ;)

It seems I can't say this enough... You don't _have_ to be alone with
Evergreen if you don't want to.  You're certainly free to maintain
everything yourself, but if you would prefer there are companies that
will assist in supporting your particular installation.  A great thing
about Open-Source is, you aren't locked into a proprietary support
system.  If you, for some reason, aren't satisfied with your support
provider, you can use a competitor!  I know of 2 companies right now
that market support for Evergreen...

Given a sufficient configuration, this system should be able to handle
your load! I forget the stats of the PINES operation, but I do
remember that it's supporting the items of >250 libraries!  The
Evergreen system was built to be _very_ scalable.  In the blog
history, there was a datacenter tour.

The PINES installation had (and this is just off the top of my head...
I probably screwed up some numbers) 4 database servers, 30 "worker"
servers, 2 load balancers, and a couple other systems.  The system
would need to be customized to your specific needs, but I would feel
quite confident in saying that a well-designed server system could
easily handle your load!

As far as separate collections.  The Evergreen system has something I
believe is fairly unique, and that is customizable site-types and
levels.  An example that may work for PINES would be as follows:
+System Root (PINES)
--+Consortium
----+Library
------+Branch
--------+Collection
--------+Bookmobile

A sample hierarchy for a multi-branch academic may be as follows:
+University
--+Campus
----+Branch
------+Collection

The sample location setup could be:
+Timbuktu University
--+Uptown Campus
----+Foo Library
------+General Items
------+Graduate Lounge
------+Storage Room
----+Baz Library
------+General Items
------+Storage
----+Blimey Study Center
------+General Collection
--+Mid-Town Campus
----+Dewey Library
------+General Collection
------+Computer Center
------+McMorris Center for Excellence in Supporting Open Source Software
--+Downtown Center for Social and Economic Studies
----+Lecture Hall (Rm. 101)
----+Bibliotheca du Espanol (Rm. 200)
----+Faculty Support Services (Rm. 207)

So, as I tried to demonstrate, you _could_ have a collection by making
a "sub-location".  This would allow an easy way for your patrons' to
search specific collections, search a whole library, search the entire
campus, or even search the entire university network! It also provides
a way to make it more obvious that separate collections are in
different physical locations.  For now, I _believe_ it is relatively
common practice to use the collection name as part of the call number:
STORAGE
005.6
Ern
1997
C.1
as an example.

I hope this (and the others') answer some of your questions.  Please
don't hesitate to write back with more!

--Don McMorris


On 5/4/07, Frances McNamara <f-mcnamara at uchicago.edu> wrote:
> We have put up the Gentoo image on VMware and are using it single user,
> via VNC.
>
> I was able to go in and immediately Z39.50 to LC and OCLC (using our
> passwords) and to Import bib records.  I figured out how to Create some
> bib records.  I figured out how to create some items and check them in
> so they show up in the web catalog.  I figured out how to Edit a MARC
> record after it was added.  I added some books, sound recordings,
> scores, dvds and can see them in the pac.
>
> I mostly did this without documentation, only using the help button in
> MARC Edit and create, so I have some questions about what I am seeing.
> If there is user doc I should look at, let me know.  I can see this
> works fine for some collections.  I purposely looked into some of the
> things important to a large academic like us.
>
> 1.  When I Import a MARC record there seems to be a capability to
> Overlay.  Is there any doc that describes how that works?  (actually a
> lot of our real work would require a loader for files of MARC records,
> but I would think that could be developed or the scripts to batchload
> initial loads could be adapted.  We do use such utilities a lot and they
> allow for some data modification during import including deleting or
> moving data and automatic creation of items.)
>
> 2.  When I Create a MARC record while typing in the 100 Author field it
> gave a message about how there was no authority record for that name.
> Are authority records used in this system?  If so, how do they get in
> there?  Can you Create MARC an Authority record as well as a bib
> record?  We actually send out bibs and load authority records.  Is that
> done in this system?  Are the cross references and see also references
> used in any way?  Seems like all the indexes are Keyword, so I'm not
> sure how references in authority records would be used.
>
> 3.  When I Create or Edit a MARC record, if I save it with empty
> variable fields they stay in the record.  That was disconcerting as most
> systems get rid of them on save.  Not a big deal, but you might want to
> fix.  I had a lot of difficulty editing the Fixed Field.  The 008
> displays as a labelled display at the top, and then as a plain field
> down in the variable fields.  When I did Ctrl/F8 as per Help it wouldn't
> let me edit in the labelled part.  It let me edit down in the 008 but I
> couldn't figure out how to edit it so it was correct.  I was unable to
> add Date1 and place of publication code to the 008 when I created a new
> bib.  Seemed like I should have been able to, so maybe I was doing it wrong.
>
> 4.  When I create a MARC record, is there somewhere that data about when
> that bib was created, or changed or who did it is kept?  Systems do this
> in various ways.  In my current system there is also a "bib_status" and
> a "staff_only" flag.  We use those because we add and display bibs that
> are on order or in process but then we don't send them to set holdings
> in OCLC until they reach a "full cataloging" type of status.  We also
> export records for various reasons based on such a status.  Is there
> anything like that in Evergreen?
>
> 5.  After creating a MARC record I went to "Holdings maintenance" to add
> what I would call an item record.  I could add item or items and save
> them.  It seems like there is the idea of a "location" in the sense of a
> "Branch" that I had to choose before Creating the items.  But there does
> not seem to be an idea of a "collection".  So if in that branch they
> have a separate "Reference Collection"  Where do you put that
> information so the user knows to go to a special place, not the regular
> stacks to get the piece?
>
> 6.  When I create a single item it nicely grabs the Dewey and LC call
> numbers from the MARC bib and lets me choose one and "Apply" it so I
> don't have to retype it.  It defaults to the Dewey number.  Can I assume
> I could default to the LC number if I wanted to?  Or not pick up the
> Dewey number?
>
> 7.  When I created an item, for a single piece it was OK except that we
> mark each book "c.1" on the spine even if we only ever get one copy.
> It's not clear where that should go.  It is obviously much more
> important when you get to multiple volumes.  I created a record for a 2
> volume set.  There is nowhere to put the volume numbers except in the
> call number field after the call number.  I input a more complicated
> example by putting in PC world, a magazine, and making numerous items.
> So they have call numbers like   QA76.8.I26P3 c.1 v.22 (2004:
> Jan-Jun).   I put all that in the call number fields for v. 20-23.  When
> I looked at this in the pac the items did not sort in correct order even
> though I had put them in in order, which seems odd.  Also when there are
> multiple volumes with the same call number you get a separate listing
> for every single volume or copy in the Browse call number list.
>
> Is anyone having problems with putting copy and volume information into
> the call number field?  Are there any plans to have separate fields for
> that type of information?  It might be needed.
>
> Also, in a lot of systems between the bib record and the item records
> there is a layer of "copy" records.  This becomes important with Serials
> where there might be one copy in the main library and another in the law
> library.  The items get grouped together by the "copy".  They also get
> summary holdings based on that copy because eventually listing all the
> items isn't useful and you want to say v.20-25 instead.  We tend to have
> separate copies for the bound volumes which shelve in the stacks, as
> opposed to the unbound issues that shelve in the "Current periodicals"
> section.  Is there any plan to put a copy record between the bibs and
> items?  Or does such a copy record actually exist in the background
> somewhere?
>
> 8.  When I create an item, the bib did not show up in the pac until I
> checked in that item.  We don't do that.  We have too much stuff going
> through, it just goes to the shelf, they cannot stop and check it in
> before doing that.  Is there an option to not have to check it in?
> Actually we have some items where we have a flag "staff only" so it
> won't display in the pac and we have some bibs with that sort of "staff
> only" flag too.  Is there anything besides the status of the item that
> controls that in Evergreen currently?
>
> 9.  When creating items there is an option to print spine labels.  I
> looked at that briefly.  It seemed like it did not work well with LC
> call numbers.  Is there somewhere that you set up a bit of code to break
> up LC call numbers?  Also, can you edit directly into the little box
> that displays how the label will look?  I wasn't sure if I could do that.
>
> 10.  Is it true that there are only Keyword indexes.  Do people think
> there is a need for the type of Browse indexes in current library
> catalogs?  Usually that is where the cross references and see also
> references in authority records have been used.
>
> 11.  One other thing.  Of course we would assume we would need to
> contribute to development if we were thinking of trying to use this, but
> do you think this software could work for a database of 5.2 million bibs
> and 7 million items?
>
> I'm sure we'll have other questions but as I start to look at this, I
> just want to make sure I am not missing some documentation or
> misunderstanding any of the basic functions.  Thanks.
>
> Frances McNamara
> University of Chicago Library
>