26-01-2003 05:34									#1
	Scenarios stability and demo numbers Support Apolyton, buy GURPS/ Alpha Centauri

Richard's comments about Dawn in the D7 comments thread are very interesting in that they show a stability "problem".
Minor, or unrelated changes, in the code, may turn a winnable scenario into a very hard one.
Here is an example I just experienced:
I implemented wall removal: Breached walls will now disappear, thus cities should fall even faster after walls crumbled.
Just before that, I had implemented the new tech model based on knowledge instead of levels.
I tested the siegewars scenario just to see if the walls disappeared correctly when I would breach them.

I couldn't breach the walls. I tried 4 times and lost every time. I thought either I was playing badly or something else. I played the D7.2 jar I had instead of my current development code, and I won, piece of cake.

So I checked the tech code. The new RP fed haven't been adjusted to give the same results as the previous version. I divided the growthrate by 10, and I won. The goblins got too much tough out of tech growth.


Conclusions:
The tech model rewrite has deep impact on gameplay even in scenarios where it wouldn't be expected to.
I am quite sure every other change (like making the ai be more aggressive for Jericho) can have impacts on all scenarios.
Thus I suggest that we do heavy testing of all the scenarios before we release a next version (be it 7.2.3 or 7.3 or whatever).

	27-01-2003 03:37									#2
	Re: Scenarios stability and demo numbers Tired of ads?

quote: Originally posted by LDiCesare Richard's comments about Dawn in the D7 comments thread are very interesting in that they show a stability "problem". Minor, or unrelated changes, in the code, may turn a winnable scenario into a very hard one. (snip) I am quite sure every other change (like making the ai be more aggressive for Jericho) can have impacts on all scenarios. Thus I suggest that we do heavy testing of all the scenarios before we release a next version (be it 7.2.3 or 7.3 or whatever).

I agree in general . We just have to be more careful in the future. That was my fault for the last demo .

One thing I would like ideally to have is one or two dedicated scenario playtesters who aren't coders. Once the program-crash bugs from any changes are mostly eliminated, we could turn the package over to the scenario testers. They could play each scenarios looking for things that have been broken. This would be a valuable service that would essentially free up probably ten hours of extra coding time for new demos since that is about the amount of time to test all the scenarios carefully, and right now programmers are doing it. Given the current D7.3 plan Richard would probably end up doing some of this anyway in presenting and tuning the tech system. How does this sound, if we can find the volunteers?

Volunteers for scenario playtesters anxiously sought! Any takers?

	27-01-2003 13:56									#3
	Support Apolyton, buy GURPS/ Alpha Centauri

I volunteer for all kinds of playtesting. These stability bugs are to be expected, since so many things are changing, and IMO it is a very good sign that we can get rid of them fairly easily, just by changing a few numbers. Since I am familiar with the guts of several models, I should be fairly good at this.

I am curious about that tech growth. What was the growthrate number that worked right, and what was the one that made them too good? Were they getting their RP´s from battles, government input, or the economy? Could you tell me about it in the tech thread?

	27-01-2003 17:47									#4
	Support Apolyton buy from Amazon

quote: Originally posted by Richard Bruns I volunteer for all kinds of playtesting. These stability bugs are to be expected, since so many things are changing, and IMO it is a very good sign that we can get rid of them fairly easily, just by changing a few numbers. Since I am familiar with the guts of several models, I should be fairly good at this.

Wonderful news Richard, that will indeed help accelerate our progress, and you're an excellent person to be doing it! Martin has also volunteered to try to fix the Dawn tutorial stuff, so between the two of you (provided he's interested beyond the Dawn ) we will have a strong scenario and advanced testing team!

	28-01-2003 01:42									#5
	Support Apolyton buy from Amazon

Well, yes I'm interested beyound Dawn, but I'll not commit myself befor I have had a shot at Dawn. After all I have no idea how much work is in it.