I've lived through the previous tech bubble, but was only an employee. It didn't really require a lot of cash back then to start up, depending on what you are doing. A shared server at an ISP or university and some perl scripts could get you going. Racks of modems in your garage to start a small ISP. What really drove the bubble was that larger companies would buy these small operations, often including shares in the bigger company, and package them together into a group which then in turns gets sold to even bigger companies like Cisco or do an IPO. This is not very different from what's happening at the moment, except the action on the IPO front is not as strong.
Things also got crazy because a
company's turnover could mostly be based on paying Cisco, Microsoft and
Compaq money for equipment and software. It was all based on growth of revenues, not profits, so moving stuff in expensive chunks while spending all your investments cash gave you that growth.
I know some people who didn't
sell, they remained small in comparison but are still in business today
and doing very well. My advice is that if
you sell, make sure you get some cash for it to park somewhere to use
during the bad times when raising cash will become impossible and skills
will be cheap when you actually do have cash. Parking might mean an
asset you can borrow against when it's hard to get funding based on a
promise. Don't sell with a deal made up of mostly shares in the acquiring company, based on an
assumption that tries to time the market. Most experts can't even time
the market right.
Apart from that, keep making stuff people outside of
the technology industry use and give them value that is worth paying
money for, and try not to make it depend on things you get during a
bubble: easy funding and clients with easy funding. Trust me those
things are hard to determine, you only realise how much of your client
base also depended on the bubble after it's over. Most importantly the
effects of bursting bubbles take several years to unravel, so don't
think that if you're still fine a year after the crash that
you are in the clear.
That said, if you currently have a business that at least breaks even and doesn't make all your money off technology customers, then you're already at an advantage. During a bubble salaries are
very good and it takes some guts not to get lured in by an easy salary
and rather take the initial cut in income to do a startup instead. When the music stops and you are still
turning some profit you will be glad. You will be cutting expenses, but you won't be without a job.