Mohtalim

This is weighing heavily on my mind since I've been trying to root cause something since November. It seems like I always need just one more week to figure it out.

I can write a new feature in an area I know well in a week, though I won't have tested it completely before the week runs out.

It sounds like you need a way to track your progress and break down what you think you still need to do in order to solve the problem. This is exactly what was taught to the metrics students in the technical track. Unfortunately for you, I wasn't in that class However, from talking to my classmates who were in that class, it was similar to the metrics I did learn about to track project progress in my metrics class. The difference is mostly that you're measuring yourself rather than the collective progress of the team. If you want an extreme view of how to do this, look at PSP (Personal Software Practice) that came out of some research from the SEI. It's a LOT of overhead, but the long term benefits could be high. There are adaptations of agile approaches that would probably work better for you, but they're quite a bit less documented than this.

http://www.sei.cmu.edu/tsp/psp.html

No, I need the bloody logic analyzers to run at 1333 MHz. Wouldn't hurt if the emulation team got their act together, either.

That's part of your problem for sure that the tools are slow, but the other part is that your estimates are too optimistic and you have no idea by how much.

I had a string of several estimates over a contiguous period end up being under by almost exactly a factor of two. So, I decided that must mean that I was good at estimates, but needed to double all my guesses. However, I've noticed that over the last several months, every task is running much longer than even my "corrected" estimates. Either I'm getting dumber or I'm working on harder stuff.

Were you tracking the estimates or your project manager?

I was. A few of the estimates were requested by managers, but I don't recall them ever checking status after that. I have had them track status on tasks assigned to me but estimated by others. I typically beat those estimates, but those are usually longer time periods then anything I estimate for myself.

What motivated you to start tracking your estimates?

Nothing. It was just something I noticed. I'd say something like, "I'll be done with that in a week". But then at the end of the week, I was still working. I wouldn't finish until the end of the next week. I was annoyed enough at being so far off that it stuck in my memory long enough to notice when it happened again.