WebJunction

Libraries have been making inroads on OSS designed specifically for their own computing needs. The [url http://webjunction.org/do/DisplayContent?id=1216 ]OSS focus page at WebJunction[/url] features some library-specific case studies and starting points for fostering and developing OSS applications to solve library problems. Eric Raymond, one of the leading advocates for OSS, has pointed out that [url http://www.openresources.com/documents/cathedral-bazaar/ ]every good work of software starts by scratching a developer's personal itch[/url]. Many of these projects started as in-house effort, often by one person, to deal with an issue where either there was a lack of commercial options, the options that existed didn't completely deal with the problem, or it was simply too expensive to buy a solution.

The best starting point for finding out what's going on with OSS in libraries is at <A HREF="http://oss4lib.org">oss4lib</a>. It is also worth hanging out here and at [url http://usrlib.info ]/usr/lib/info[/url]. Start posting information on applications that you are working on, or even that you would like to see built. The [url http://access2003.lib.sfu.ca ]Access conferences[/url] have started holding annual hackfests, which are meeting points for people from all aspects of the library world to get together and create solutions to real-world problems. How many library conferences are held each year? What if hackfests became more common, and were also held regionally? We have had great results from working with our local public library system, most of us probably have potential partners for OSS development closer than we might think. There are also options for commercial support and installation. [url http://www.redhat.com ]Red Hat[/url] illustrates this model in the Linux world, but there are also companies like [url http://www.userful.com ]Userful[/url] that sell a linux-based desktop for public computing. The [url http://www.koha.org ]koha[/url] folks maintain [url http://www.koha.org/installation/support.html ]a list of people and organizations that can provide contract-based customization work as well as maintenance and support[/url]. In other words, OSS can offer many of the same reassurances that commercial software does.

There is a growing body of OSS library systems. All of these applications are a testimony to the quantity and quality of modern computing environments, particularly those fueled by the web. The acronym [url http://www.onlamp.com ]LAMP[/url], which first became popular in Germany, is often used to define how MySQL is used in conjunction with Linux, Apache, and Perl/Python/PHP. These OSS development tools are well-documented and offer a low-barrier entry point to scratching that itch in your organization. To these, I would also add Zope, AxKit and Cocoon. Zope was the world's first application server, and is a wide-ranging, object-based platform for powering web sites, applications, and almost anything you can throw at it. AxKit and Cocoon are two OSS systems from the Apache Foundation, and represent an exciting intersection of component-based computing and XML technology. There are many other tools as well, but these are all mature, well-defined building blocks for putting together powerful applications. Revolutions have been built with less.

Art

I have heard those arguements. I always counter that if the ILS vendor does not support open standards and cross platform support then libraries should find one that does.

Let's be clear, for a majority of users business or otherwise, the functionality of Office 95 or WordPerfect Suite 6.1 was all that is needed. Anything else is smoke and mirrors

What we are seeing is a lot of time and money spent (and mindshare bought) in proprietary software that benefits the software companies. Then expensive training and certifications are "sanctioned" by these software companies to further lock-in users and providers of technical support to the whims of the software companies. Can you say "Software Assurance?"

What OSS can do for libraries is to get us off the software upgrade treadmill that supports large software companies and spend that money, time and effort toward services that will benefit our patrons and communites. Because I will not understand the need to pay for Windows/Office when Linux/Open Office can do the same thing.

Ed, I have a feeling you're preaching to the choir.

I'd like to hear from folks who are interested in moving in this direction, but have encountered obstacles. What are those obstacles? Were they overcome? How?

Are there current struggles for which we can brainstorm and troubleshoot?

A totally basic question, then:

Let's say that all my colleagues use MS Office, we're all on a MSFT exchange network, and we use Outlook to coordinate schedules and make meeting requests.

If I switch to OpenOffice, can it coordinate with my colleagues, who are using MS Office?

I feel nervous about posting this; perhaps the answer is obvious?

Laura Solomon, the coordinator of Automation and Technology Division of Ohio Library Council, has recently released her IT Asset System. You can find out more information about the system here: [url http://www.heightslibrary.org/webmaster/olc_showcase/default.php ]http://www.heightslibrary.org/webmaster/olc_showcase/default.php[/url]

Here?s a summary of what the software does:
Every public library has to deal with tracking assets, and the IT Department at our public library spends a lot of time doing this. Previous to this system, IT asset tracking was done in our library using paper and cumbersome notebooks and records were handwritten. This system does more than simply track asset tag numbers and what equipment they are attached to. A brief summary of features follows:
· Ability to track what software is installed on a particular machine
· Ability to dynamically generate asset record forms for print out
· Ability to track which OS and service packs are on a particular machine
· Abililty to track user information and physical location for a machine
· Ability to track peripherals in the same manner as computers
· Ability to track vendor information
· Ability to track how many licenses of a particular software have been installed and are or are not available
· Ability to distinguish between purchased and leased machines
· Ability to duplicate a given item so that it can be slightly edited and saved as a completely different item

Josh,

Thanks for posting the IT Asset tool. It looks really cool, and from your post, similar to Tech Atlas. We need to post a permanent link to that tool here on WebJunction so that folks know about it.

We had a great discussion yesterday, and that can continue if folks still have things to add re: getting started with OSS.

Art, I spoke with someone on the phone last week who was really excited about your program - and the potential to partner with other orgs to get OSS initiatives off the ground. He was so excited that he sent a message about your article (and our event) to a list serv that reaches a lib system of 107. Can you give us some tips on how to create and foster these types of partnerships?

Art, are you suggesting that there's a PR opportunity for an organization who could take that initiative forward - like ALA or OCLC, as suggested by djsarge yesterday?

Do you think these are some of the best positioned org's to do that? Who else?

These are great points, and I have been far more active in regional associations than anything at the national or international level, so I am not sure I am qualified to give a good answer on the association connection to OSS. My guess would be that some combination of IFLA and OCLC would have some potential, maybe the work on [url http://www.ifla.org/VII/s13/pubs/wgfrbr.htm ]FRBR[/url] is a bit of a model for this if libraries were to go the blueprint route. Although blueprints tend to have a Java connotation, that is more a consequence of Sun's early work in this area. However, if you peer through the Java haze, their approach has the enormous benefit of being engineered close to reality. What I wouldn't want to see is an internationally-authored RFP or RFI. I don't think we have a shortage of this kind of documentation in libraries.

Blueprints typically have a reference implementation, are architected to be used in numerous environments, and describe a pattern that gives a sort of virtual DNA for bringing in contributions from other systems. If you think of the links between, say, serial check-in, and the kind of predication mechanisms used for weather systems, or circulation calendars and iCalendar-logic containers, it hints at the potential for putting together applications that leverage the commonality between library and non-library systems. Maybe there are better models out there than the Pet Store, but it could be a starting point.

The secure desktop is something almost all libraries need, and I am sure we have gone all over the map to create. For better or for worse, we need to have windows solutions here as well as linux-based ones. Not that I don't think linux is better suited for public desktops, it's just that many organizations don't have a choice in what they run. I also have grown fond of hardware solutions here as well, things like the Sheriff card and the Magic Card. Where does hardware fit into OSS? I would like to build a USB device for entering passwords where we don't want to give student assistants access to some systems. What would be open sourced for something like this? Maybe the answer would be a blueprint for this as well.

Has anyone tried [url 'http: ]User Mode Linux[/url] for public access systems?

As I understand it, UML allows you to set up a complete Linux system running within a normal Linux process. Kind of like Linux within Linux. The advantage to this setup is that if a user damages something in their system, they damage only their copy, not the enclosing system. When they finish using their copy, the user copy of the system disappears, taking the damage with it.