Closed eazar001 closed 9 years ago
A thought just now precipitated: without even inspecting the code, I can speculate with a very high degree (but not complete) of certitude that the bug resides in priv_msg_paragraph
. This is most likely being caused by an apathetic counter that injects newlines independently of newlines that already exist. If this is the case (in my estimation, most likely), then this is probably a 1 to 2 line fix. You would probably just have to reset the counter on every newline encountered during list deconstruction.
However, you shouldn't do this now ... as if there are latent interacting bugs that can surface because of this "fix", then you would just end up breaking a (still) functional yesbot. Find the time to properly test this for a few minutes after fixing it.
Which leads me to another point, manual testing of certain facilities is obviously becoming tedious. You may want to set up some testing suites to automate some of the drudgery. An obvious and most likely ample candidate for the majority of situations you face would probably be Hendricks' quickcheck
. Food for thought.
Just now (without even inspecting the code) I can already envision carriage returns causing problems. So be wary of this when you finally decide to debug.
Oh, lowly Pissant of a Bug! Behold, the Power and Weight of my Boot crushing the Life from thy worthless Breath with the simple flex of Commit e833777a3ee5f353ec7e95b31d975d5f98808350.
I'm not sure if I'm observing anomalies, or if the observances actually have merit. Inspect
auto_nl
variants when you have time; make sure no superfluous newlines are being generated. This is a low priority bug.