Once my contract arrangements with Motive Power came to an end and when it appeared that CarComms would not be getting any traction, I needed to find something else to do for money.  Luckily, the valley was only just at the beginning of what would be a firestorm of job loss and company shutdowns, so it was not impossible a task.  I don't remember how exactly, but at some point I got in touch with a recruiter at LeapFrog.

I went down and met with the VP of Software Engineering, Dave Conroy.  Dave had to be just about the nicest, calmest guy I had come across in my silicon valley travels, with the possible exception of Jeff Winner at Netscape.  I also met with the two main software developers there.  The interview was reasonably quick, and the arragenments would be contractual to see if we all liked each other.  I would start right toward the end of November.

A few interesting facts about LeapFrog at that time.  They were split across two locations, one in Los Gatos and the other in Emeryville (both in California).  They were looking to go public.  They were hiring new people.  The project I was to work on was starting late-- the toy business has a rather strict scedule and the project should have started in September, I was starting from zero in late November.

Shortly after I started work on the project (which was going through spec revisions, name changes, etc), a product management team was assembled for it in Emeryville.  Emeryville was only about 40 miles to the north of Los Gatos where the engineering group was, but during a conference call that distance starts to grow rather quickly.  Day-to-day, I was actually the "old-timer" on the project by a couple of weeks, with the exception of input from Dave or others up there now and then.  This made for some interesting times, to say the least.

The toy business does follow a rather tight schedule, the precise details of which escape me presently.  It goes something like this (give or take):

• September - start project
• December - toy fair in Asia
• February - toy fair in New York
• April - project should be feature-complete
• June - release to manufacturing in China

Since the toy business has extremely tight margins, and since LeapFrog was looking to go public, manufacturing toys in China and shipping them on the slowest boat possible were essential to the bottom line.  Getting the product manufactured in June to arrive for the September school buying rush (and then into the winter holidays) is a fact of life in the business planning for the company.

Having started this project late and with a completely new crew split across two company locations trying to learn tools and processes, and with the name and specs for the project one which I was working changing even into April, the chances that I would complete this project in May were slim at best.  Add to this the fact that LeapFrog's accounting department would routinely turn a Net-30 invoice into Net-30++ (one time, Net-51) and I am surprised tot his day that it actually got done.

On April 1, a Chinese fighter jet collided with an American spy plane off the coast of China.  This put quite the strain on the Sino-American relationship, which in turn had LeapFrog executives doing backflips.  Had China decided to block exports (including toys... something like 95% of the world's toys are made there), LeapFrog would have been toast.

There were other issues along the way.  The product I was working on and another product that another contractor was working on both required some on-screen text formatting, but there was no code library to format text on their embedded hardware.  We had to come up with a spec that would work for the two products and I wrote that code to deploy... that is, there was almost no time to properly design and test it... it had to work, and it had to ship.  That was true of the product in general, which was just a tad frustrating.

In the mean time, the toy fairs are used to tune and tweak a product, and in the case of this one the tweaking went to extremes.  The product was to include an embedded dictionary, but as late as April the details of the dictionary, the UI, and the source of the definition data was still undetermined.  Buttons and formfactors were changing as well, which in some cases required some new engineering tweaks in the hardware, and there was also a new monochome LCD screen, along with development tools that were being developed during the product develpment cycle.

I had a feeling we were in trouble when the person I was sharing a bullpen cube with, Sean O'Brien, was passing some time with me by going through some of the questions that were destined for the unit.  One was regarding the first 12 amendments to the Constitution.  There were others.  I think the unit was to have a pool of 6,000 questions or so.

Well, as a fresh newbie starting a project about two months later than it should have been staretd, with a product management crew that was newer than I was, with a spec that would not get nailed down, and with a second contractor dropped from consideration, we got to June and I wasn't done.  This lead to what you could call a crisis in toyland, and Dave and three of the other programmers jumped in to help clean up and finish off the code pieces (there were several different blocks of code in this device, so that part wasn't difficult to break up).  In the mean time, the VP of hardware was giving me dirty looks all the time.

In the end, we signed off on it on Auguest 13, just over two months late.  They had to manufacture the units without software and the finish them, and then fly the first batch over at a rather tremendous marginal cost.  In the end, I'm sure I received full blame for how the project finished

On the other hand, I did receive a rather impressive offer for full-time employment with LeapFrog, which I turned down at the time.  I had lost interest in the silicon valley work environment, and had already committed to moving to Phoenix.  LeapFrog eventually went public and did pretty well (NYSE:LF), though I don't track it myself.  Sean is still working there as of February 2005, so he is 100% vested with his options... good for him.  Good for all of the people there.

The toy business-- the technology-based, educational toy business, anyway-- is not unlike any of the other businesses I was a part of in silly valley during my time there.  In the big picture, I'm happy with my decision, though I must say the cash woud be nice nowadays...