While auto-turns is certainly MUCH better than not having it, it still leaves much to be desired.
As a programmer myself, I understand that how things work under the covers is not nearly as simple as it may sound, but if auto-turns already keeps going until an interrupt that affects a player race, I would strongly suggest hiding and "auto-turn-able" turns and simply continuing to "sub-pulse" until the original turn-length the player asks for.
Example: I hit 30 days. Unless there is an interrupt that affects ME, the next thing I want to see is 30 days. I don't care that 2 NPRs in a system I haven't discovered yet are fighting, and I don't need to see "last increment" entries for their behaviour.
In fact, giving me those intervening turns can be a source of intelligence, actually, especially if I started the game with no or very few NPRs. Suddenly, as I enter a system, I get some unexpectedly short "turns"? Must be an inhabited system...
I can understand that "under the covers", the NPRs have to make decisions, and it works best to handle this, internally, as "turn" events, but this should most certainly be hidden from the player, and with your existing "sub-pulse" mechanism, I don't think it would be very hard.
Based on my experience, this is one of those ideas that sounds great in principle and is not-so-great in practice
.
The reason is that Aurora is still not perfect
. In particular, you can still get into situations where you end up taking waaaaaaaay too big a timestep, and end up jumping well into some detection range. Whenever I see a funny timestep, I do three things:
1) I
do not back up my current DB. The reason is that if the funny timestep was too big, then I want to have the opportunity to back up to the previous save (typically no more than 4-5 clicks of the advance time button) and go forward with the same orders but a smaller timestep. If I stomp my backup DB, I've lost that opportunity.
2) Set the sub-pulse length to 5s, and run a 20 minute update. If the timestep was too big, then I should see an event right away, and it will be obvious. If the timestep was ok, then I still want to be running with a small timestep for a while in case the event that caused the interrupt had to do with one of my ships, and something is about to happen in a minute or two. If I immediately clicked a (typical for me) 1-day increment, then the first timestep would be 30 minutes, which might be too big....
3) Backup my DB. Now that I'm (fairly) sure that I'm not suffering from a weirdness associated with 30-minute timesteps, I want to make sure that I can get back to this point.
So yes, seeing a weird timestep tells me that something's probably going on. But I actually already have a pretty good idea if I've jumped into a system with an NPR -
the race generation code takes noticibly longer than the system generation code alone, and mostly I just role-play around the knowledge.
Another problem is needing a "progress meter". In the old days, there were frequent cases where Aurora would hang during an update. It's important to know that the reason it's taking 3 hours to advance by a day is that it's churning away 5s timesteps.
John
PS - Please read the "where should I post" FAQ. In particular, Steve has requested that people put their suggestions and bug reports into the official threads, rather than starting new threads. The reason is that he uses the official threads as "filing cabinets" - the problem is that once he's read an idea in a non-official thread, it's very difficult for him to find it a few months later when he wants to implement it (since it will be marked as read).
