Quoll Work Thread, 2024-11-18 - 2024-11-18 12:29
I was busy playing soccer on Sunday, and I actually was intending to try to take my newly found less sad brains and do some cleaning up around the house, but I also want to resolve the issues I am facing. And the way to do that is to do another test run…
To summarize, my change to have Ada not nag about continuing the conversation after checking topics caused issues with
some of my Farewell Responses where I don’t print anything if the conversed-this-turn variable is set to true. That
is, my fix to the first was basically to set conversed-this-turn to true and to my surprise, the setting of that
variable leaked to the next turn. I should be able to fix that by limiting my fix to only “explicit” requests to list
suggested topics, but I found myself questioning my original pattern of blocking off Farewell Responses behind
conversed-this-turn being false.
Despite doing some code archeology, I could not figure out why I chose to do that. My guess is that there are cases where I would have an interlocutor leave as part of a conversation node, but then have that cause the Farewell Response to fire. So let’s say Ada was talking to Glitch, and Glitch would say something like “And that’s this part of the tutorial; see you at the next one, Ada dearie!” and Ada would describe Glitch leaving to then be followed up with the Farewell Response where Glitch says goodbye to Ada.
But that’s just speculation and I tend to prefer verification over speculation. So with that in mind, I am actually
going to not do my fix to listing suggested topics yet, but remove the check of conversed-this-turn on one of my
Farewell Responses and see what happens to hopefully confirm my speculation.
It’s annoying that the tests take a while to run (maybe about an hour and a half), but this is the exact reason I spend so much time maintaining them…