Ignore the email I sent just some hours back: the changes I had made
were should not be necessary for Z3950. It appears the Library of
Congress response to the Z3950 download request is in a different
protocol/format called SRW or SRU. Dr Bainbridge suggests that you use
the same details to connect to the Library of Congress server, but fill
them out in the "SRW" part of the download pane rather than in the Z3950
Other details that you can try to use in this part of the download pane
when trying to connect to the Library of Congress:
Find: digital library technologies
Max Records: 2
The above URL and port were the ones originally used by the Library of
Congress for Z3950 and SRW/SRU communication. It still works.
Barbara King wrote:
> Yes, I already contacted Larry and got the correct information to
> input in GS3, but no go. I believe I need to be on the GLI web server,
> which I am not connected to.
> Thanks again.
> On 08/09/11, *Greenstone Team * <email@example.com> wrote:
>> Hi Barbara,
>> I think the max records field in the Z3950 Download panel of your
>> snapshot looks correct, but I believe the other fields you filled in
>> are not. There are specific values for host, port and database that
>> one needs to fill in for each Z3950 server. Since I didn't know what
>> these are for your question either, in the case of the Library of
>> Congress, I resorted to searching for Google on the words: library of
>> congress z39.50 port
>> Which led to this helpful link: http://www.loc.gov/z3950/lcserver.html
>> Since it seems you want to download MARC records, I used the
>> information in the above link to fill in the Z3950 Download fields as
>> Host: http://lx2.loc.gov
>> Port: 210
>> Database: LCDB_MARC8
>> Find: digital library technologies
>> Max Records: 2
>> Unfortunately, while the above information is correct, the Library of
>> Congress is now returning information in XML format, which the
>> Greenstone code that retrieves data using the Z3950 protocol did not
>> anticipate. It is therefore broken at present and will not work,
>> despite correctly filling in the fields as above. I am in the process
>> of (hopefully) fixing it.
>> Thanks for bringing this bug to our attention.
>> > I clicked before attaching 'snapshot'. Could you send me a snapshot
>> of your suggestion please?
>> Your snapshot shows that the maxdocs field has the right value (2)
>> for testing. Are the other settings perhaps what's causing the
>> trouble downloading? Have you confirmed that the Z3950 port of the
>> LoC is indeed called "marc"? Have you tried other Z3950 clients and
>> were they able to download from LoC using the details you filled in
>> into Greenstone's Download pane?
>> Barbara King wrote:
>> >I clicked before attaching 'snapshot'. Could you send me a snapshot
>> of your suggestion please?
>> >That would really help.
>> >On 08/03/11, *Greenstone Team * <firstname.lastname@example.org>
>> >>Hi Barbara,
>> >>While I don't have any experience in this, I would try the
>> following out, to see if it accomplishes what you want:
>> >>In GLI's Download pane, use Z39.50 as the Download Setting and
>> point its download location to Library of Congress by filling in the
>> fields for Host, Port and Database. Then tick the Max Records field,
>> setting it to something low such as 2. If that worked well enough and
>> you feel ready to take the leap to 100, try it again with a max
>> records size of 100.
>> >>Barbara King wrote:
>> >>>Hi once again Anupama,
>> >>>Your response below gave me courage not to abandon GS3 and to
>> continue building a DL for my project.
>> >>>Question: Is it possible to download 100 bibliographic records
>> from the Library of Congress (LOC) catalogue/Z39.50? If so, I would
>> appreciate it if you could inform me what steps I need to do in order
>> to accomplish this task please.
>> >>>Thanks in advance for any support you can provide.