 |
 |
Hi, following up on our discussion about how to extract holdings information from a III catalog, and following David's generous offer to share his code, I went ahead and started a sourceforge project: libxess. libxess is intended to bridge the gap between "Library 2.0" services and legacy ILS such as III Millennium and others. See the picture at http://libxess.sourceforge.net/ It is intended to provide a simple, yet clean web interface to a set of proxy scripts which in turn handle the specific access to those legacy systems. We have an immediate if mundane use for libxess in LibX: we'd like to show people what their library holds. Although we could have included this functionality via screen-scraping into the client, we felt for various reasons that it may be better to keep this as an optional, server-side service. Our vision is that if a library installs libxess, then it can benefit from all services that are "libxess-enabled." - LibX being just one of them. lnstalling libxess should be a low overhead operation, such as uploading a few php scripts to a server. Our target audience are not programmers, they're librarians who may have limited access to their ILS (*), but who usually have access to a web server and a place to run php scripts. I feel that many library-related projects, at least in as much as they interact with existing ILS, would benefit from such a facility, and, to my knowledge, there doesn't seem to be such a facility as of know. For this reason, I'd like to invite interested people to either contribute or give their input. You could contribute by, for instance, providing code for your catalog or whichever system you'd like to see supported. If so, give me your sourceforge account and I'll set you up with CVS access. Let me propose to use the sourceforge forum at http://sourceforge.net/forum/forum.php?forum_id=693773 for discussions. A related question is what the exported interface for this service should be. Right now, there's none. David's script returns MARC-XML. I'm wondering if this may be an application for unapi (and for me finally reason to read through its spec ;-). I had also considered OpenURL v0.1, but as an outsider I really know too little about library APIs and formats, so I'd be grateful for input here. Simplicity is paramount. - Godmar (*) Keeping in mind that in some III-based libraries, you have to have reached a status that's comparable to OT4 just to gain the privilege to contact the vendor's helpdesk.