[WEB4LIB] Re: open source and librarianship

[Jeremy Frumkin]

On Sat, 29 Jan 2000, Eric Hellman wrote:

> As a software developer, my perspective is that the most important 
> goal for the library community ought to be the opening of software 
> interfaces. Coming from a non-library background, I perceive that 
> library software vendors are not very oriented towards making one 
> software package work with other software packages. As libraries 
> become e-Libraries, this becomes less and less acceptable.

Your perception is my perception, and a lot of other people's. The current
economic model that most library software vendors follow promotes
proprietary mechanisms, not open standards. The few current standards that
are out there (MARC, Z39.50, perhaps Dublin Core and RDF, although these
aren't yet widely implemented) aren't enough.

> As an example; linking systems at publishers need to interact with 
> holdings information at libraries; our LinkBaton system (which you'll 
> hear more about later this week) is a small step in this direction. 
> Ideally these systems should interact with procurement systems and 
> aggregator systems.

This would be ideal. I'm interested in hearing what this LinkBaton system
is; I'd also like to add that another huge problem is one of
authentication/authorization. Different aggregators and publishers use
different mechanisms for allowing access to licensed materials. Many times
this limits the kind of linking / searching that can be done with their
products.

> 
> In an ideal library network environment, premium information services 
> ought to be able to work in harmony with free or lightweight 
> components. You should be able to buy an Encyclopedia Java Bean and 
> drop it into your Apache-based ILS server. When MySQL runs out of gas 
> on your database, you should be able to plug in Oracle, and when 
> Oracle squeezes all the money you have out of you, you should be able 
> to switch to Informix.

Exactly! Modularity is key to building the next generation of ILS. 

> 
> I don't think a big GPL project is likely to encourage vendors to 
> open their interfaces. I'd like to see public-domain interfaces 
> established first. These would then define many smaller projects, 
> some of which would be well suited to GPL development, others to 
> commercial development, and others to public-domain contributions. 
> (GPL forbids the use of code in proprietary derivative works, public 
> domain does not.)

I'm not so sure about this; I think we need both approaches. A few big GPL
or even public domain projects can help look at standards that smaller
projects might not; they also could incorporate the smaller projects as
components. An example might be taking the Open Source Course Reserves
System (OSCR), and working it into a module for the OSDLS (Open Source
Digital Library System), which right now doesn't have a reserves
component. 

> I think it's a mistake to look to Open-Source strictly for cost 
> savings; we use Linux and Apache even though they've cost us more $$ 
> than commercial solutions would have. It's the stability, security 
> and performance that make them compelling solutions.

I totally agree. The statement "One should use Open Source software
because it is cheaper" reminds me of the time when people said "We should
use computers because they'll make us more efficient". It's not
cost-savings that we gain; it's a better product in the end.

> Openly publishes interfaces for its software, and I'm reminded that 
> I've been meaning to release our java SICI classes for free; I just 
> haven't had time! Feel free to give me a nudge if you're interested.

*Nudge* *Nudge* :-)


-- Jeremy

-------
Jeremy Frumkin
Metadata Librarian
University of Arizona Libraries
jfrumkin at bird.library.arizona.edu

The Open Source Digital Library System
http://osdls.library.arizona.edu