Question for you Addon devs. I recently introduced a bug in my addon that completely broke it for any new installs. Not my finest moment. But I missed it because I haven't been doing full end-to-end testing every time I make a change since that's a big pain. It seems it's something I should be doing, though, since I don't have coded tests and I can't really figure out how one would do that for an addon.
I'm wondering what you all do for testing. Do you have in-code tests, if so, how? Do you just use it a bunch to try to catch bugs? Perhaps a list of things to check before a release?
For me that depends on the changes made and the situation.
For BfA I tested my addons with no other addons running to make sure loading process is correct and tested the "main" functionality. Still a few errors made it to Prepatch and I had to push a few updates.
Normally I dont touch the core of the loading process of my addons. I've created a "common" part that all my addons share. With BfA I touched it for the first time in a long time so I ran way more tests than usual.
That means for small changes and additions I dont run a full test nor many tests. For example I've added "connected realms" to my money broker and pushed the update without deeper tests other than checking on a few realms if the sum seems right.
"Luckily" my addons are not that popular so that most errors are never reported before I run into them myself
The cataclysm broke the world ... and the pandas could not fix it!
Like Rilgamon said, it totally depends on what (and how much) was changed. I've pushed bugs plenty of times. It's hard to catch all bugs that could happen under all manner of circumstances. That is why large projects (such as WoW) have whole teams for QC and play-testing.
"You'd be surprised how many people violate this simple principle every day of their lives and try to fit square pegs into round holes, ignoring the clear reality that Things Are As They Are." -Benjamin Hoff, The Tao of Pooh