Beginning of 1997, you could feel the first deep internet bubble effects in the Bay Area. Prices were increasing at an incredibly fast rate, companies were popping out like young mushrooms in the sunlight, salaries were exploding. But all in all, the good roles were already distributed. The gurus were already hired since 95-96, working for one of the largest software companies around or another. The 1998-1999 hires had missed the first train of opportunities and were somehow delighted to find the internet bubble on their way.
Bubble ? You said bubble ? Oh, oops, yeah a bubble can explode. Most of the late hires went away, and all the jobs not deeply related to technology went away too.
Seven years ago, Opera was a young company still far from making money ; Netscape was on a bad slope with AOL standing in its back ; Microsoft almost stopped working on Internet Explorer after the end of BW1 (First Browser War).
2005, Opera is growing and growing, and its visibility in the mobile market is a fact. Mozilla is growing and growing, its market share in the desktop browsers' segment (and I don't forget Minimo...) increasing every day. Microsoft is about to ship IE7, and still hires. Google releases incredibly cool and well done software. Everywhere around us are popping out technology-based companies, not quite the 1998 dot-coms. These companies have nothing in common with pet.com, they don't show a burn-rate as their sole company fact, they have brains, and use them to write code. The product can fail, the value of the company is the value of its brains. Technologies that people expected to appear in dad's browser in 2000 become only mature - or widely accepted - now : markup-based vector graphics (SVG), gauges and other complex form elements (XForms), graphic widgets (canvas), per-site stylesheets and scripts (CSS and GreaseMonkey), scriptable plugin APIs, and so on.
The last time the Web needed so many implementors was seven years ago. Definitely a nice time for geeks....