Information Takes Over

I have been watching the conversation between Richard Wallis and John Blyberg with interest. Most of what both have said seems to me like common sense. You can’t be everything to everyone (wallis’ view on support much generalised), and as one of the everyone, you want to be able to do everything (Blyberg’s view).

I only see one problem regarding access to the data in our ILS. We do have access to our data (theoretically). I can (theoretically) create scripts on our unix box that use it, BUT, the level of documentation that explains the relationships within the database is inadequate. I know there is data in there, I know what that data is. I know what I would like to do with that data. What I don’t have any clear picture of is a way to get that data out of there in the most efficient way possible. (and I am guilty of writing really inefficient SQL now and then)

If my library vendor (talis) provides me with a set of standards based APIs that I can use to get at all of my data, I will happily pay them to provide this in the most efficient way possible (so long as they deliver the goods!).

I do not want to be an expert in database structure, nor do I want to be an expert in writing SQL or any other form of code. This costs too much in time. I want lego blocks that I can build together, giving me access to my data as and when I need it, within the framework of a standards based, well documented, SUPPORTED API.

Richard Wallis: Imagine looking at it from as a support analyst’s point of view. From her end of the telescope she can see [in Talis’ case] potentially 100+ Johns hacking away on their systems every day – a thought to drive you straight toward the caffeine in the morning!

John Blyberg: Those support analysts need to be retrained (reprogrammed?) to be able to think about the caller on the other end of the phone as an equal–that way, they become an advocate for the customer, not just a problem solver. I just can’t buy into the thinking that we can’t be granted the essential right to access our data because that might make it harder to support.

Richard Wallis: What if – Ann Arbour’s citizens really like their Library patron history shown on the web site and it becomes a prime source of information for them; and then Mr Blyberg gets head-hunted to work for another library, or even an evil ILS vendor; and then for some reason the facility stops working. Who would John’s successor phone for help, and how would that conversation go?
What if – John built even more services on his access to the database and; for reasons commercial, performance, or just a change in coding style, his vendor’s next software update made a major change to the database shape.

If I am using the API correctly and able to get at all data AND DOCUMENTATION, support analysts WOULD be able to help because the 100+Johns would all be using the same API and reference. If I was replacing Mr Blyberg in the scenarios above, I would be able to do my bit, research the implementation of the API, check for updates to the API, and be able to modify what had been previously created, before ringing support to ask then if there was anything from their end that looked wonky.

I would concede that there could be a problem in making an ILS that would run on multiple platforms, different databases have different characteristics, and you would have to draw a line somewhere. However, if we are able to access our data with an API, there could be nothing stopping us (in theory) from extracting that data into our own favoured database software or software system. We couldn’t expect our vendor to support that, sure they could help if there was a problem extracting the data with their API, but after that…

Technorati Tags: Library 2.0, library, Talis, Blyberg, Wallis, vendor support

This entry was posted on Tuesday, December 6th, 2005 at 13:45 and is filed under Everyday Life. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.