ACAT Linked data

On 8/10/2015 3:38 PM, Brian Briscoe wrote:
> On Mon, Aug 10, 2015 at 8:28 AM, McDonald, Stephen > >wrote:
>> >The concern about OPAC display failing when Internet connectivity is >> >broken is unnecessary. No one is suggesting relying purely on remote >> >links. Such a system should have alternative lookups at backup sources, >> >and local caching of important data, and optionally local customization to >> >supplement or modify the data. It could even be a purely local system, >> >with the original data downloaded once or very infrequently and no remote >> >linking at all; you could still get some of the benefits of linking by >> >downloading the linked data at the same time as the base data. Like >> >libraries who currently download the NAF one time and do not maintain it, >> >it would not stay up-to-date and would not be as flexible, but it would >> >still have benefits beyond the current system.
>> >
> I don’t understand this. If the data is linked, it will be a link to a > remote source. While it is possible that that other source could be > internal, doesn’t that somewhat defeat the purpose of having links? If the > net connection is down, then there is no linking possible. While I am > espousing linked data, I think ignoring this drawback is naive.

Actually this is a different matter. In computer speak, it is “The Graceful Exit” or “Failing gracefully”. The idea is, when a program breaks down (as happens all the time), the program should not just crash but continues on. From the developer’s point of view, it should be graceful enough so that the user is not even aware of the potential troubles in the background, which can be major. (A lack of awareness of something gone wrong, a site not being searched, could have serious consequences for library catalogs. But a program can fail and continue, but let the searcher be aware that something hasn’t gone quite right)

In practical terms: the world lives with this reality of the internet every day. But it is not cheap: redundancy must be built in by using mirrors and/or rerouting or something, making everything synchronous and so on. It can and has been done, but it requires a real infrastructure. There is also this thing called “cooperation” which would become really vital for something like this to work.

I would suggest that once libraries switch to Bibframe, our old catalogs will become mostly obsolete and I don’t really know what smaller libraries will do. It won’t be free and it probably won’t be cheap.

At least there are open source catalogs now!

James Weinheimer
First Thus
First Thus Facebook Page
Personal Facebook Page Google+
Cooperative Cataloging Rules
Cataloging Matters Podcasts The Library Herald