Owning a pet is a privilege not a right, one that you are free to exercise as long as you don't harm/impact your fellow citizens. And we don't like stepping in nor picking up your poop.

1. Increase the pet license fee to help defray the cost of "waste removal" from public property. Yes this puts more burden on the respectable pet owners, but they can help themselves by self-policing the bad pet owners in their community. Stop being polite and ignoring bad behavior when you see it - chew the offenders out.

2. Also use part of the increased fee revenue to add more animal control officers or police. Give them the authority and responsibility to catch more offenders and issue civil infractions with substantial penalties.. $200 USD first occurrence, 4-5x that for the second occurrence. Unpaid bills automatically promote the offender to strike 3..

3. Just like baseball, you get 3 strikes. On third offense, revoke the pet license and take their pet away.. permanently. Furthermore, put a ban on issuing the individuals or their immediate family members another pet license for 5 years or more. Its not the animal's fault surely, but some folks will never change behavior without threat of severe consequences.

Lack of fundamental software development skills and knowledge of the underlying environment is the true issue. Perhaps I'm a bit crusty, but I think that all new programmers should spend the first 2 years of their careers coding from the command line.. a shell of their choice (bash/ksh/etc), vim or emacs, a build tool (ant/make), using scripts to start, stop, and deploy your own services, knowing how to configure and actually use a remote debugger instead of splattering code with unnecessary print statements, learning how to trigger stack traces and interpret them, merging code by hand, etc. After this "apprenticeship", they would graduate on to "journeyman" status and allowed to use an IDE if they want to, but they would always have the strong foundation to fall back on in times of need. More importantly they could train new generations of programmers.

Its pretty easy to roll your own basic geo-location searches. We implemented similar functionality for a hotel travel site several years ago offering basic point and radius as well as bounding box searches on a major commercial RDBMS, bypassing the optional packages and avoiding several hundred $$$ of cost. Just use the Great Circle algorithm mentioned above -- any decent programmer should have no problem getting this done. If the functionality is used heavily, you should expect to do some tuning and performance optimization - perhaps even some caching of data or pages.

The biggest concern will most likely be finding a good source of geo-location data - points of interest, city centers with meaningful locations, neighborhoods, and weird shapes such as islands, boundaries such as states and countries. The original poster mentioned Canada.. how good is the GIS provider's information for Nunavut? You may be surprised at the data quality issues for even some well populated areas. You should also think carefully about how this data will be maintained, how often it will be refreshed, and if you need to manage your own edits or overlays to the existing information.

At any rate, unless your needs are complex, don't switch databases or run off to sign expensive contracts to solve a simple problem such as this.