Sales at software companies, [1] big and small, are falling as a result of the recession.  Such facts raise the question of the business model some of these companies adopt.  Many regard their development activity as a research and development item, and recognise that new development, essential for them remaining credible, must continue in bad times and good.  Others have a business model where the main revenue comes not from sales but from the predictable revenue stream of on-going maintenance.  One of the newer players manages its business entirely by metrics that disregard the R&D element of software development altogether.  Their software development is almost shut down, as they weather the storm in the same way as the bear survives the winter - in hibernation.  Public sector software purchasers rely on that sort of vendor at their peril.  The winter may be longer and harder than imagined - and the bear may never wake up, but die.