Isn’t it interesting how a single word can have such different interpretations?
When a charity talks about a legacy, especially a large one, it is usually something that they are delighted to receive. But when an IT magazine publishes an article called “Living with Legacy”, you can be pretty sure that they are talking about something they see as a problem. In this case, old (or very old) computer systems that are often business-critical, but expensive to look after and to adapt to the changing needs of the business. Given half a chance, many IT managers would be delighted to be rid of their legacies…
Both types of legacy are something that has been “left” to a future generation. If not well planned, both can create all sorts of unforeseen problems.
For a charity, a legacy may be tied up in restrictions that can tie the hands of the charity in using it. If too many legacies are restricted (e.g. to preserve a particular building), there may be insufficient general funds to pay for a fundraiser for other needs. An extreme example, but it illustrates the point that it can be helpful to discuss and plan a legacy with the chosen charity.
With IT systems, one problem is the technology – systems can be written in obsolete languages, to run on ageing platforms. Over time, the number of people able to understand the programs, let alone change them without causing problems, gradually reduces. While, of course, the cost of maintaining them steadily rises.
Another big problem (not just for IT systems) is foreseeing the future! When a system is built, all sorts of assumptions can be made that later turn out to be false or unsustainable. When BT built its customer service systems (CSS), everyone just knew that telephone lines only had one end, the customer end – all lines went from BT to a customer end. So that was all that had to be recorded. Sadly, the introduction of new technology (leased lines) meant that lines now had two ends… When computer systems were built in the 1960s & 70s, no-one thought they still be around in the year 2000, so information about the year was often recorded using just 2 digits rather than 4. Which was why there were all the dire predictions about what might happen due to the “millennium bug”. It cost billions to update these long-standing legacy systems.
When you’re thinking about a new computer system to support your business (or charity, for that matter), it pays to think forward a wee bit when writing the requirements. It’s still guesswork but it’s worth considering at least a few questions about what might be. What do you expect the lifetime of the system to be? How might people want to use it in future? What assumptions might completely derail your business model or the design of your system? If BT had considered the question “what if… technology changes and lines have 2 ends in future?”, that would have saved them millions in building a completely new set of systems for private circuits. What are your “what if” questions?