It's a shame, isn't it. The term 'legacy' could mean such good things, but in IT we use it in the same condemning way we might refer to burnt out old warehouses, or rusting farm equipment. No system worth its salt wants to be considered as part of the legacy environment, and indeed, it's a CV decision as to whether people want to work with them, in development or support roles.

I remember my first job interview, with ICL. "What do you think of COBOL?" they asked. "It's a dinosaur," I replied. "Can't imagine it's got that much life left in it; only a fool would want to work in that field." It was on hearing the words, "We're a COBOL shop," that I knew the interview hadn't gone all that well. But here's the real punch line: that interview was back in 1987.

Today, we hear a range of conflicting views. We know for a fact that Mainframe isn't dead, or even dying - it has both adapted to the latest trends and business needs, all the while sticking with some of the stress-tested architectural principles that made it such a good platform in the first place.

There are undoubtedly applications still delivering the goods that were first commissioned decades ago, but meanwhile, we know of organisations that are held back by slow-running, inflexible applications that nobody dares touch.

It's not just big systems either - legacy can arise from making the wrong technological bets, like all those who invested their efforts in MySpace profiles, only to discover that Facebook was the de facto standard. Software and hardware, from email systems to network management tools, all can find themselves in the legacy dustbin.

So, where's the truth? Are we forever chasing technological rainbows, or is there plenty more to be had from our older systems and applications?

Is it a simple case of ignoring the marketing and choosing the right tool for the job using it until it stops being of value? What are your experiences - do you have any horrors lurking at the back of your data centre? What's in the pipe today that will become legacy after it gets deployed (or worse, before)? Is your organisation battling on with clunky kit, or coping fine and wondering what all the fuss is about?

As always, we'd love to hear your views.