Make progress every day.

[dougo]

I hate debugging. It's tedious, painstaking, and frustrating. It's like when you try to untie your shoelaces but you end up with a knot instead, and you have to sit and carefully untangle the knot when you just really want to get your damn shoe off.

What's worse is that once you finally fix a bug, no matter how sure you were that it was going to be the last one, there'll usually be another one hiding behind it. It's pretty much impossible to estimate just how many bugs are in your way—it could be one or a hundred, but there's no way to tell until you're done.

But it sure is a great feeling when you find out that you just fixed the last one! After finally getting this particular feature to work after three weeks (when I thought it would take one), now I can start actually using the feature and taking it for a spin. Wheeee!

(Minor bright side: now I know a lot more about online partial evaluation than I did before. Maybe someday it will come in handy again.)

Comments

[jtemperance.livejournal.com]

The absolute worst part of all of it is when managers at software development organizations try to pretend the process is predictable and hold you to firm deadlines.

[dougo.livejournal.com]

Yeah, or graduate advisors.

[jfb.livejournal.com]

I love debugging. But congratulations on getting to stop.

[mshonle.livejournal.com]

I don't like debugging when I'm building on unstable but released open-source tools (like Eclipse M9). You have the problem that you don't know if a bug is really yours or theirs. Bugs in programs 100% all my own are far easier to deal with. Fixing other people's bugs takes decades of experience to do really well.

[dougo.livejournal.com]

I actually did find a bug in MzScheme a couple months ago, so yeah, that's always at the back of my mind. It was just a performance bug, though, which isn't so bad. But "unstable" is always a relative term. And of course open-source is much better than closed-source, because if you suspect it's not your bug you can actually follow up on your suspicions.

[mshonle.livejournal.com]

I've rarely had problems with proprietary software I didn't have access to the source to. Generally, there's an outstanding QA team to hammer out the bugs. And, if I get problems I could always make a small test case and submit the bug.

But in my experience, open-source is the less reliable, not the more. But maybe I only pick good "closed source" software.

[dougo.livejournal.com]

I think jtemperance put his finger on it—I'd like debugging if I didn't have any time pressures. It's a challenging puzzle, but the solving process can be tedious (especially without a good debugger). I will say that I like fixing bugs, it's just the finding them that's a pain. Of course usually it's a trivial fix (a typo or something) which makes it even more annoying that it took two whole days to find.