The bargain is a little different with other programmers, but for non-programmers who use open source, here’s the deal we make:
- The programmer writes something and releases it.
- The users use it, and report bugs.
Pretty simple, eh? By that I don’t mean “report bugs, and then ceaselessly complain about them and insist that they be fixed or the program will die and the programmer will go straight to hell,” but, say I release a plugin for previous and next in category links, which I only look at while I’m testing and constantly rebuilding. If you actually use it, and look at the results, then you are responsible for noticing that if you post entries in categories A, B, B, B, A, then of course the first A entry doesn’t get rebuilt when you save the second one, so the first one doesn’t get a “next entry” link built until the next time you rebuild either that entry, or the whole blog, and you are responsible for telling me about that stupid bug.
Now, if only I could find someone to make responsible for figuring out how to fix it. Bleah, it’s going to need an MT3.0-only hook on MT::Placement->save(), to dig out the previous-in-category entry and rebuild it. Bleah.
Quick straw-poll of the zero users of the plugin who are likely to read this: do you actually want the current weird behavior, where an entry that has the category as a secondary category gets included, so you change from navigating through the A entries to navigating through the B entries when you hit one with A as primary and B as secondary? Right now I have a new version that defaults to primary only, and does secondary only if you ask for
<MTEntryPreviousInCategory secondary="1">, but if nobody actually wants secondary, rebuilding would be easier if I leave it out entirely.