I am concerned by the fact that the IP Registry appears to have gone 
around figuring out the IP ranges for schools based on public records 
from the IANA and a bunch of vendor records.  I'm sure that was 
difficult, and their site says it took four years.  When it was done, 
they announced that 58% of IP ranges were wrong, and began selling the 
service to vendors and telling them what our IP addresses are based on 
their analysis.

I claimed the account for my institution and discovered that there were 
26 vendors already pulling my university's IP ranges from the IP 
Registry.  Unfortunately, the IP ranges were wrong.  To name a few 
problems:

1) They conflated us with another school in the same university system.

2) They could not know that there are a couple of IP ranges that we 
prefer to be treated as "off campus" even though they belong to the 
University.

3) They had no way to know that one particular range of our IPs is 
assigned to a library consortium in our state, and used for proxy 
servers that serve the other institutions in the university system plus 
several dozen public libraries.

The third point is critical.  By distributing these erroneous IP ranges 
on my school's behalf, without permission, the IP registry has 
effectively granted access to 26 of our subscriptions to basically 
everyone in my state.  We are thus in violation of our license 
agreements and will be at risk of legal action by the publishers until I 
can sort this mess out.

Because this involves multiple institutions -- my own, the broader 
university system, the aforementioned library consortium -- I am going 
to have to contact and explain the situation to a lot of people, and 
spend a lot of time checking and re-checking IP ranges, all in service 
of updating the IP Registry's records.

Then they get to turn around and charge the publishers for my work.

But frankly, their business model feels like extortion to me. We have to 
verify their records, or there's a chance that our resources will be 
accessible to people who should not have access because their analysis 
was incorrect.  They appear to have engineered a situation that puts my 
institution in potential legal jeopardy, which we can only get out of by 
improving the data that the IP Registry is selling for a profit.

I am not happy with them.  The basic idea -- a centralized repository of 
IP ranges for bulk updating publisher records -- is both sound and 
useful.  But their business model leaves a bad taste in my mouth.  If I 
could, I would opt out of the system.  But they do not appear to have 
made a mechanism available to do so.

Will Martin

Head of Digital Initiatives, Systems and Services
Chester Fritz Library
University of North Dakota