Where have all the Texans gone?

A further 28 percent will "modernize gradually" and only one in five admit "it's rip and replace time".

When people talk about "legacy" systems, what do they really mean? Is the term "legacy" used as a convenient way to hide the organization's past inaction?

We all know that IT is a dynamic industry, so how on earth do IT shops end up with systems they can no longer get tape for and software that's long past its expiry date running central business systems?

In most other industries, managements accept the fact that operational infrastructure needs to be constantly refreshed to avoid the risk and expense inherent with forklift upgrades.

Transport companies update trucks, manufacturers update machinery, telcos update network infrastructure, and hospitals update medical equipment, but, heaven forbid, how dare we proactively refresh core information management systems to avoid being stranded with a paperweight that just happens to house all our business data?

It just doesn't make sense. To all those organizations with an archaic black box that pre-dates human-readable code: seriously consider doing a pilot project to make the move.

There is clear evidence that modern computer systems are up to the task of automating mission-critical processes, and what's more encouraging, is today's systems seem to be much less likely to leave users in the "legacy" quagmire generated from yesteryear's approach of 'one system per application'.

Don't be afraid to take a chainsaw to your legacy systems, because sooner or later you'll have to, and it's better for you to start the engine than to be told to start cutting with a dead-cold machine.

With that approach, whether your modern IT systems are in-house, hosted, commercial, or open source becomes academic.

Fire up the chainsaw every few years, that's my motto. And clear away the "legacy" problem for good.

Can't get the board to ditch the dead wood? Tell me about it at Rodney_Gedda@idg.com.au