Update on progress

0 views
Skip to first unread message

Ross Singer

unread,
May 23, 2008, 6:16:19 PM5/23/08
to jangle-...@googlegroups.com
The prototype Jangle core was migrated to the Sinatra framework. With
both the prototype connector and core in Sinatra, it should be fairly
simple for an outside developer (even a non-Ruby developer) to look at
these and have a pretty good idea of what is going on (although they
need documentation). I think this was a good move since it makes the
inner workings of Jangle much more transparent and much less
obfuscated by Ruby frameworks.

The prototype OpenBiblio catalog was populated with mock patrons
(pulled from the contact information from the U.S. Congress, giving
the Dilettantes Library 535 patrons) and a script was run to assign
item records to bib records and attach a status to them.

For each bib record, there was a 70% chance it would have an item
record associated with it (all of the records came from Project
Gutenberg, so, realistically, they could represent an bib record
without holdings). There would then be a 50% chance of a second item,
30% chance of a third, 10% chance of a fourth.

For each item there was a 20% chance that one of the following
statuses might be set:

* a 1 in 10 chance it would be damaged
* a 1 in 10 chance it would be lost
* a 1 in 10 chance it would be on order
* a 1 in 10 chance it would be on the cart to be shelved
* a 1 in 10 chance it would be on loan (presumably via ILL)
* a 1 in 10 chance it would have a hold

If none of those were met, it would be checked out to a congressperson
at random.

This allowed for development on the Item entity. Now, when passed a
barcode, the Jangle connector will return Available/Not Available as
the item title, more specificity in the Item summary (such as lost,
checked out, location for available items). If the item is checked
out, an iCal record is sent in the content tag to indicate when it
will be available.

I also wrote two posts on Jangle.org. One is a bit of explanation of
Jangle. The important part of it is, Jangle is not an open source
project (although it will produce at least one), but rather an open
specification on applying the Atom Publishing Protocol to library
services (although not exclusively). The other is an introduction to
the Jangle core.

The next steps are to document the pieces of the API already written
and write up some commentary on the motivations behind the choices
made.

For now, my OpenBiblio catalog is living at:
http://dilettantes.code4lib.org/openbiblio/opac/

Jangle core:
http://anvil.lisforge.net:6767/services/

Jangle connector:
http://anvil.lisforge.net:4567/

Examples of Item responses:
List of items associated with resource:
http://anvil.lisforge.net:6767/openbiblio/resources/2968/items/

Checked out item:
http://anvil.lisforge.net:6767/openbiblio/items/000003114
(view source to see iCal)

Lost:
http://anvil.lisforge.net:6767/openbiblio/items/000001397

-Ross.

Reply all
Reply to author
Forward
0 new messages