I find that the work I get is more interesting, in that I don't usually have to do much maintenance - obviously, the corrollary of this is that I try to ensure that my code is easily maintainable for those who come after.
So if, to take an actual example, I end up working round one of the limitations of XSLT 1.0 by an exceptionally exotic technique that owes a lot to FORTH programming techniques (and which took me a couple of days to dream up), I document it in depth, in a comment, right there in the code, and explain how it could be done more efficiently using XSLT 2.0, in case they ever upgrade.
Actually I've often found that, although I get the interesting work, I get less of it per unit of time - it's not unusual for me to be taken on, and for the project to then suffer delays from other parts of the organisation. I'm usually kept on with repeated extensions, because they don't want to lose me. Although I generally try to do something of value to the client during these phases, there are times when I get good rates for doing very little, as my post count in TPD demonstrates
So if, to take an actual example, I end up working round one of the limitations of XSLT 1.0 by an exceptionally exotic technique that owes a lot to FORTH programming techniques (and which took me a couple of days to dream up), I document it in depth, in a comment, right there in the code, and explain how it could be done more efficiently using XSLT 2.0, in case they ever upgrade.
Actually I've often found that, although I get the interesting work, I get less of it per unit of time - it's not unusual for me to be taken on, and for the project to then suffer delays from other parts of the organisation. I'm usually kept on with repeated extensions, because they don't want to lose me. Although I generally try to do something of value to the client during these phases, there are times when I get good rates for doing very little, as my post count in TPD demonstrates
Comment