1. BigPicture.

In 1994 I was working on ways to deliver a simple mapping interface, with no cost per seat to a few hundred *low powered* (486) PC's.

I came up with a system that used ArcInfo to produce huge arrays of map images, updating them overnight in areas where updates.

By inserting each image into a database, I could support hundreds of thousands of images -enough to fake a vector driven solution.

I wrote a little windows application at home called BigPicture that did just this.

We were even able to have CD's burned and ship a complete free package.

We also had "hot spot" points in the database (eg a point inside every land parcel and on every manhole) so we could fake "ID".

It worked a treat. The IT department hated it. It was never deployed on the Manukau LAN.

2. PizzaMap.

In 1995 I wanted to put property mapping on the internet for Manukau City.

There was no money for an ArcInfo license, and besides ESRI had probably realised by then that this sort of use of ArcInfo licenses shouldnt be encouraged.

A web version of BigPicture was an obvious way to go.

As a refinement on Big Picture users were able to type an address, and see their property highlighted.

This was achieved by storing parcel coordinates in INFO (ie the Info in ArcInfo) offline, and using InfoLib in a CGI process to return them. These were added to a precooked postscript tile (made by the same ArcInfo process as BigPicture). The postscript was then ripped "live" with Ghostscript using parameters that produced an image at the required extent or scale.

Web users could query their address and it could return a map in about 9 seconds. Low scale navigation was via an image array.

We were beaten to first web mapping system in New Zealand by a Genamap powered solution at Landcare. Damm!

The name Pizzamap came from the look of the maps we were serving, and the way they were sliced up. It wasn't "Pete's-map"

Pizzamap ran on a retired Sun Sparc, and was a remarkable piece of skunkworks because we managed to get it passed the IT people. The beauty was, no one knew what that box did, and they were too scared to touch it! Someone told me years later that they used to show it to their GIS students and "marvel at its crude but effective operation". I liked that.

Eventually the marketing people found it and freaked out, not because of the wildy out of date data etc -but because it didn't implement the official branding standards! Goodbye PizzaMap.

3. Map-TV.

At some point we did the obvious thing, and wired ArcInfo up to the net via the IAC (RPC) interface.
Then ESRI released ArcView IMS, and I remember being at the local demo. Never have I seen the users so excited (well maybe at the Arcview 2 demo).
What was really cool about AV-IMS for me, was how simple it was to get going. Point and click. What was not cool was the price. I have always been cheap, and never had a department etc budget so I have no perspective on pricing. But it seemed pricey.

I had managed to avoid learning Avenue to that point, but this got me working on my own "AV-Web" solution. It avoided what I considered the big downfall of AV-IMS, which was that (then) fat Java client. This made Av-IMS awful to start on dialup connections. Without DHTML we had to drop rubber band zooms and drag pans BUT it was super fast on the internet.

I cant recall what if anything happened to AV-NET, but no doubt I would have hit license problems if I had got serious. I was happily naive about these things!

A bit later, I quoted on my first job as an indpendent contrator. The client wanted an Intranet mapping system and (I later found) had a budget of about $60K. Knowing nothing, I pitched for a $5K solution. My plan was to use MO, but not MO-IMS. I didn't need the "web stuff" that MO-IMS brought to the table because I had worked out other strategies. Luckily for all concerned, someone (Andre Post) at the client site made me check that this was allowed under the MO license. To my amazement it wasn't. These days, I give away software for desktop use that I sell for web use. In those days, I was astonished because I could only see the problem from a technical point of view. "How can that web connector and some example apps be worth all that money?".