KB wrote:One of the biggest factors, as has been mentioned upthread, is that it is no simple matter to train someone to understand the ins-and-outs of Scrivener so that they know it well enough to code it. It's a large app with a big codebase. [snip] For a developer to do a good job, they need to know not just what the different features are, but how they are integrated. They have to know all of the things that most users will never notice.
Could you tell us a little about your test suite? I've seen these good test coverage go a long to helping devs A) wrap their heads around what the heck the software is supposed to do, and B) provide regression tests to make sure they didn't accidentally break an obscure feature.
Not that throwing money or new devs at a codebase just as it's trying to wrap up a release is a good idea! But maybe after v3, more devs could pitch in to document the system better via more tests?
OTOH, maybe you have 100% test coverage and that still doesn't solve the problem.
PS: don't actually strive for 100% test coverage, that way lies madness & diminishing returns.