[Web4lib] An Analysis Of Open Source ILS Market Penetration

[Bob Molyneux]

Rick Mason smoked me out. Yep, I worked for SirsiDynix as its Chief 
Statistician--up until May 9, 2007 when I was laid off. But before being 
evil vendor-scum, I had a life in the library world he has somehow 
completely missed in his expert searching. For 20+ years, I have compiled 
and analyzed library data as well as having a life in IT. I suggest my 
book, Internet Under the Hood as a good (albeit a bit dated) introduction 
to networking and the Internet for LIS students. A bit more checking might 
find that many of the library data series I have worked on are available on 
the Internet for folks to use. There was a reason Sirsi hired me.

My library data work has largely involved compiling longitudinal data and 
analyzing them. It is something I do. A sensible person might exclaim in 
horror: "Why would anyone do THAT?" To look at trends. Trends start 
somewhere and it is a fact that we often don't know when things start in 
the library world and that makes subsequent analysis untidy. I submit we 
now have a begin data on objectively studying the impact of open source 
ILSs--although a better one will follow in a bit when the NCES updates the 
public and academic library data. The LISNews post is the beta. I hope to 
make this report a regular...say every six months. In five years the barge 
may have picked up some steam.

When I was still at SirsiDynix, I talked to Marshall Breeding, whose great 
work at lib-web-cats I have relied on for this analysis, about doing a 
similar analysis tracing all ILS vendors using not just counts but some 
other measures such as circulations, population served (which public 
librarians seem to prefer), or maybe expenditures--like I did here. This 
kind of analysis would give a different view than counts because of the 
skewed nature of library distributions, simple counts don't give the whole 
picture. I thought it would make a helpful addition to his work. In any 
case, I have asked Marshall again for data from lib-web-cats that would 
enable me to do that relatively rapidly and on a regular basis. He measures 
which library has which ILS differently from the way I did, of course, but 
I think that is addressable.

I have written to Dan Scott separately to apologize for including 
Laurentian as having a "relationship" with Equinox. I misinterpreted it and 
I regret it.

"This isn't to suggest that any of his data is suspect." Well, thank you. 
They are, I believe, correct but errare humanum est. I have citations for 
you to check and if you can't do the math, I can send spreadsheets. All 
sources are available on the Web. If there are errors, I want to know about 
them so I can correct them.


Bob Molyneux
drdata at molyneux.com
XyWrite forever!
0x 09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0