Maybe, but it may well be not. That is the main reason we are discussing this.
Hmm, this point merits some consideration. I do play long(ish) sessions, but I tend to exit the game fairly regularly too. The glitch did happen to me twice though and the 2nd time it happened it was one of the shorter sessions. How long is Quake-long anyway? What's the average session time till the stability issues begin popping up? That I don't really have much of an answer for. Usually I start experiencing them when you start doing multiple Episodes in a single play session. If it is one Episode, it usually isn't a problem, but once you get to Episode 4 you are bound to get some spawning issue in E4M3 or E4M5 or wherever else it may decide to occur. Usually these showed up in Cooperative sessions which took several hours and a simple replay was all it took for said issue to be fixed, so it wouldn't have held such consistency.
You *are* thorough in your approach, just like me when I'm hunting for bugs. I appreciate that. I have been in your shoes before, trying to bring something up and having it disputed, with the main difference being that said individual hadn't played the game for three years and was entirely relying on hearsay. I believe entirely in giving everything the benefit of the doubt, a known glitch like this would be quite intriguing, and it wouldn't be the first time an odd thing happened in Quake that the rest of the internet seemed to have glanced over.
No quicksaves or quickloads. I don't use them. Good to know. I know they are issues in certain games, not sure if there are any issues that come about from using them in Quake. I generally play without saving, in DOS Quake there are a lot of anomalies that appear when you load saves. GLQuake fixed all of them from what I know, I killed myself in that particular run after counting the fish and reloaded to see what would happen in the middle of a level. Still had nine fish, plus can confirm I hit 78 Enemies.
Yep, I did go there. They might have been alerted. Good to know, time to test with that difference.
So I have redone the level in GLQuake 0.98 (1.09) on Windows 8.1. Recreated my earlier starting test procedure, with the variance of adding a right click restart on death and a kill command on success.
I proceeded to quickly make my way over to the Rotfish to alert them as soon as I could. I alerted all the Rotfish around the 4th minute, then went back to doing the rest of the level. Do note this is a level where enemies tend to wander if alerted.
I can tell you that the Moving Platforms around the Water can injure the Rotfish. In fact, I managed to experience one getting stuck inside of one and dying. That explains your dead Rotfish.
I can also mention that, by alerting the enemy they also have a tendency to have more issues. It is best to kill what you alert. I think I have managed to recreate your situation. I have a save with eight Rotfish (and can confirm that is where the issue is coming from), one of which might get stuck and killed, thereby giving you your seven. Funnily enough I have noticed that somewhat in the past and had forgotten about it until now, there have been a few cooperative games where E4M4 tends to be the level where things just disappear.
So, through testing, I can confirm that you indeed have found a recreatable bug and I will restore the edit to the page, albeit edited now that we understand somewhat the factors that cause this issue. As said, I have seen issues involving 77 Enemies in the past myself, which suggests this is not related to saving.
I also can say that I have noticed one particular Rotfish in a tight corner that seems momentarily stuck until it can escape. That might be your culprit, though it is hard to tell because sticking around causes them to work properly.
Congrats, you were quite right, there is a bug involving the Rotfish. Your edit will be restored, thank you for this conversation and for being willing to discuss things.