Vendor to Open Source ILS in 1 Month #4 - Mon Jun 02

"Planning" an ILS Migration

We had our 9:00 AM Evergreen migration meeting this morning, courtesy of Grant Johnson, who is leading this wild and crazy ILS migration. This has been our main local planning session for the last 2 weeks, providing staff with the latest updates and also setting the tasks and priorities for the day. There were 12 people present at today's meeting and, true to form, it started with about 5 minutes of laughter and Grant being the target of a few well-intentioned puns. Thanks Grant - a natural leader!

I said it recently in a session I did at APLA on staff collaboration: humour will get you 60% of the way to a successful collaboration - as long as people are respectful and you know when to stop :-)

Some of the issues we dealt with this today:

• The circ status wasn't showing up in the OPAC as the institutional ID code was different in the setup tables when the data was imported. We spent some time trying to determine why this behavior was not as expected. After 1/2 hour we think we figured it out, so it was forwarded to Equinox for review. They promptly set our minds at ease and helped us understand even more about how the system works re circulation and workstation used. The nice thing about this exercise is the 1/2 hour led to a much better understanding of how the system works.
• Another issue was how we label Robertson Library in the hierarchy in Evergreen and therefore the output in the OPAC and likely systems such as SingleSearch. It was decided that the label should be UPEI-Robertson Library to be clear when coming from an external interface. But we can also leave that for another day...
• There was also a discussion of the fact that records for journal entries without copies do not display. One possible workaround would be to create a "dummy" item record for these items from the record ID so they show up for the short term. Another option we discussed was to move holdings into our Resolver and place a link in the record to this as an interim step. The most recent discussion revolved around creating an interface for staff to edit the local holdings XML file that Dan created from Sirsi and then write a program to convert that on the fly and have a "mash-up" in the OPAC to display the data. The nice thing about this is we wouldn't have to modify the Evergreen bib record but would have a detailed local holdings record set for loading into Evergreen when the Serials module is available. Not sure what the final approach will be, but we have options, including the somewhat mystical Evergreen transcendence bit :-) Sounds a bit like some magic Quark thingy...
• We reviewed the permissions for Staff and the specifics of how it is setup, as it is different than Sirsi (what isn't). The discussion started around what permissions should be given to which staff, but that quickly became "Why don't we just let all staff have the same permissions?" I suspect we will simplify and do something along these lines, Systems Manager personalities notwithstanding...
• After our meeting Paul added the Circulation rules to the live system and performed a few test transactions. With that the Circulation staff are now able to test circulation and patron combinations in the live system to their heart's content.
• After lunch we started thinking about the impact of the change on other operations, such as ILL. We use the RELAIS system (also going open source soon) and need to modify a few scripts, which currently do a combination LDAP (to a local DB) and NCIP (to Sirsi) to validate patrons and pull down information to the forms. Since we will not have Sirsi later this week we need to change how these work. We are thinking that an OpenURL or NCIPish query/result string from Evergreen will allow us to give RELAIS what it needs without too much effort. Failing a solution to this by Wednesday, we will redirect ILL links and request functions to a page providing information on doing manual requests until we get it working. We also know that we are lacking a Z Server for the Evergreen instance, so are looking at options here as well. Sounds like a post-migration requirement to me!
• I finally remembered to book training with Equinox for this Thursday. We will start with cataloguing training, since staff in that area have had little time to work with Evergreen.

There are a few interesting bits about this discussion that may not be obvious from my brief descriptions:

• This isn't so much "planning", as it is "doing with the thought that a plan would be good" ;-)
• This is a skunkworks planning process: as we progress we figure out how stuff will work and then we make it so. If we can't make it so, we move back the appropriate number of steps until we have something that works and give us a reasonable path to a better outcopme. Much different than the normal process - the big advantage is we get used to solving problems and making accommodation, rather than striving for perfection at every turn. In this Web 2.0 world this is the way to grow. The disadvantage is that we could create configurations that will need to be changed down the road and could lead to some additional work/headaches down the road. We hope not, but working in this mode always creates that risk.
• With the obvious caveat that stress can be high with such a project, this is fun! Kind of like a never-ending Hackfest :-)
• Apparently this is similar to Open Air Showering :-)
• The all important Serials and Acquisitions modules do not yet exist for Evergreen, so we have been working hard on solutions for these as well - Evergreen Migration post #6...