Software is something I've not really thought much about until recently. As Jurkowski notes in the text, "The assumption is that any new computer will come with Microsoft Windows" (p. 19). Having this presumption myself, I don't really think about what will be in front of me when I sit down at a computer, and since most operating systems and basic productivity software comes fairly standard with a computer purchase, I haven't given much thought to what else is available.
I did have to recently give thought to some children's gaming software for the library where I work. An issue with the inability to monitor one computer's usage due to its location has led to the board's decision to have internet access removed, thus the need for some entertaining gaming sources aside from the net. I found myself relying heavily on reviews, both professional as well as user-generated. The two biggest considerations I dealt with were system requirements and entertainment value - not extremely technical, but necessary to give thought to. Thankfully the system requirements were fairly easy to figure out and match with what we have here at the library and from there it was a matter of "collection development" mentality to be sure to select software of interest and use to the kids. That's the extent of my personal experience thus far.
I found it very helpful to consider rubrics/comparison charts. I guess I had a mental one, but to actually physically create one is of greater assistance. It allows you to consistently compare the software so you don't miss looking at certain criteria for one and not another. It also aids the justification process!
I enjoyed reading Chapter 9 in Jurkowski, mostly because it's highly relevant to my work situation right now. There isn't much discussed in regard to AT, but the consideration of evaluating software, especially library automation software, is something that will be of great import in the near future. I think I'd mentioned at some point during a post for our cataloging class that Pennsylvania's subsidized ASP/ILS (Millennium) will no longer be subsidized or supported by mid-2010. That leaves those libraries - including mine - that have used Millennium for the last several years scrambling for new options. Our hope is that the state will present us with another system which they will be willing to subsidize, likely an Open Source system. Otherwise, many libraries will be sorely struggling with decisions that don't leave too many feasible options. Regardless, in the near future we will likely be dealing with everything a transition and conversion entails. If anyone has any info, comments, etc. regarding Open Source automation systems, I'd love to hear it!