4 ms·
It's hard to automate e.g. fixing bugs in an ancient C codebase. And some things, such as various elements in a build tree that need to stay in sync, such that
by barrkel 17y ago
It's hard to automate e.g. fixing bugs in an ancient C codebase. And some things, such as various elements in a build tree that need to stay in sync, such that when you modify one bit you have to modify other bits, cannot be easily automated because of the brittleness such automation can introduce. Instead of simply making sure that files are in the right place and following (human) scripts annotated directly in the source code, you end up down in a maze of twisty auto-generating scripts, all not quite alike, trying to figure out what wrote what, where when and why, and route all the right configuration elements from one place to the next, adding complexity and coupling, etc.
So yes, ostensibly everything can be automated, but many things are boring and repetitive yet resist automation, while for others, introducing automation is either too much work to be worth it in the first place, or just introduces more code you need to maintain.
- jerf 17y agoIn that situation, you have two choices: Aggressively refactor until it is actually manageable, which I find fun as long as it's not the only thing I'm doing all the time, or find a new job. I am serious about both choices and am not surreptitiously claiming one or the other is the only choice.
- barrkel 17y agoRe C codebases: time spent fixing bugs and refactoring is time not spent putting in new features and selling new releases. It's a trade-off, and perfectly manageable code is at one end, unfortunately. Re builds: such things as I described aren't really refactorable without introducing the coupling / brittleness thing I talked about, as linkages need to be maintained between source-code level constructs in a handful of different languages. But these kinds of jobs can have their upsides...