Thanks! That's exactly what I was trying to say. Scrolling by exactly one line wouldn't be optimal, but if you scrolled by, say, 2-1/2 lines (just an example), it would be better. The problem isn't so much a matter of how far the button/key/gesture/whatever scrolls it by, but the fact that until now it's been (as far as I can tell) arbitrary. If we assume the user writes more-or-less within the lines on the lined paper, then a reasonable step size based on the lines would be a lot more useful. Unless someone comes up with something brilliant, it's probably only practical to ignore the case of someone that disregards the lines completely, however.

Good thought, but actually no. I just have messy writing, and thus when taking written notes at the speed of the lecture, I spend almost more time panning (due to the zoom level) than I do writing. The current state kind of ruins my use-case. :P
...That does bring up another point, though. Xournal's capability for annotating PDFs is one of its best features, in my opinion - especially for filling out forms or the skeleton notes some profs distribute in advance. In these cases, there are often no visible line markings, but the concept of a "line" is actually more important than when writing freely. I don't know if it would make sense to allow adjustment of line spacing in general, but that would make a big difference for this sort of use. I'm mentioning this just so you can keep it in mind as you design this and then next-gen UI, but I'm not too concerned about it.
Anyhow, great work so far.