On 10/14/05, Gregory S. Williamson <email@example.com> wrote:
> Shawn -- you asked:
> > 1) If I setup a user for the local use, and then want to move it
> > to a web server. Is it hard to copy the data over and put it on a
> > permanent web server?
> I've moved an entire collection (a grouping of materials in gsdl) from a
> windows server to a linux server by copying the correct directory structu=
> and tweaking a couple of files; this would not work if you were moving
> between machines with a different architecture (32 vs 64 bit; big-endian
> vs little endian) I suspect, although I have not ever tried it. In the
> cases where it doesn't you need to regenerate a collection from the origi=
Machine architecture is not a problem. You just have to re-build when you
are on the new server.
I do this all the time as I move between apple mac's and PC's all the time.
My comments on rebuilding below.
You can also easily script moving data to a permanent web server; using a
'drop box' type arrangement. Another option is the GLI applet facility.
> > 2) One of the uses I have in mind is for a Plant Identification
> > database for a professor on campus. Is it robust enough to do this, and
> > does anyone have any examples of someone using it for this purpose?
> I don't know for sure, but I would think so, depending on the specific
materials to be stored and how
> volatile they are. Keeping track of lots of documents and adding lots is
no problem; changing
> existing ones can be an issue.
Greenstone is certainly robust! It is being used for rare books and a
'e-reserve' collections at my University, for the past 3 years and has never
gone offline. (still using 2.39 on the server)
It also plays nice with the university libraries authentication systems.
It can also use 'fastcgi' if the traffic starts getting a bit high.
Greenstone has the facilities to use existing metadata schemas (like dublin
core) or create new ones, so your professor can use an existing metadata
schema for plant identification, or you can create one to his
Greenstone also has the ability to use thesauri via the PHIND classifier
links to the relevant files and an example files if you which to use a
plant identification thesauri.
If your professor prefers text descriptions to metadata, then it is worth
knowing that FAST full-text search of large collections is one of
Greenstones greatest strengths. (and this works well on low cost hardware)
Changing existing collections is easy with the GLI, and the necessary
rebuilding does not take your collection offline with Greenstone. You can
also do incremental rebuilds and/or use collection sharing to ease the load
of doing full rebuilds.
The sites www.greenstone.org <http://www.greenstone.org> and
www.nzdl.org<http://www.nzdl.org>have numerous example collections,
but I am not aware of greenstone being
used for plant identification.
I suggest using the GLI to create a pilot collection with just a couple of
plants. Maybe an image a description, and some appropriate metadata to start
with. If you get any specific problems the members of greenstone-users are
quick to help - I have been helped many times. :-)
Once you can get it going you can try it with whoever you will have doing
Greenstone also excells in automating the import of data from a variety of
sources, assuming your professor already has this data in an electronic
form. (database, text files, images, web pages etc.)
I hope this helps.
Stephen De Gabrielle
-------------- next part --------------
An HTML attachment was scrubbed...