meade95
06 Jun 05, 22:40
The below is an email from Scott -
___________________-
Hi guys,
I just posted the latest patch on the web site.
http://www.hpssims.com/Pages/updates/up_POA2/Updates_POA2.html
The entire list of changes, and there are many, are in the Readme file. But a couple of the big things in this build are:
- “sanded off” some rough edges of scenario creation/adjustment
- added patrol movement paths
- completely reworked AI aircraft paths
- added a facing command (which will not be overruled by the AI except for targeting)
- updated AI support routines.
As I said in regards to the last patch file, in general things are taking longer to work on now than before. That’s true in terms of finding them in the first place, and then figuring out how to fix them. The reason is that the program is more complicated than ever, and we’ve already picked off the easy ones. There are still two bugs in this code that I know of, but haven’t been able to squash. One is an access error, and the other is a floating-point error. I believe the latter is in the sighting routines, but I’m not sure. They each happen very infrequently, even with the large and complicated test game I’m running (once every 500 turns or so), and it seems when I try to use any sort of debugging technique... they disappear. They also don’t reappear from running the autosave file, so I’ve concluded that we are probably using an un-initialized floating-point variable someplace, probably involving some sort of object like a bridge or IP.
Eventually, though, we’ll get them - especially now that the batch routine is available to us (put in for the military JFCOM project) so we can run games uninterrupted for days on end.
I also still want to work on the loading/unloading interface and issuing orders to dismounts after an unload action. That process is complicated right now, and you sometimes don’t get the results you want or expect.
That all aside for a moment, I want to say something to you about where this project is and where we’re going. I know some rumors have been floating around, but with the effort you’re all putting in I want to make sure you know the real deal. First, while I do have plans to do POA-3, the serious development of that will not start until I’m happy with POA-2. In fact, it can’t since POA-3 will *be* POA-2, but more “AI intensive” in the interface and ease of use. I don’t want to go into a lot of details since I haven’t really done much concrete on it yet, but one aspect I’m aiming for will be that players can assign complete missions at a higher level. However, obviously that is meaningless until I get POA-2 to work and act the way I want - kinda like putting a V-8 engine on a go-kart frame.
I also want to say that POA-2 is essentially all that Jeff and I have been working on since it was released. In the brief periods where we didn’t have any reports or issues to work on (the eye of the hurricane, if you will), Jeff worked on ADC-3, and I tried to catch up on other business related stuff... but those periods passed quickly and POA-2 took over again; POA-2 is our overall priority.
I don’t really know when POA-2 will be done. To begin with, even though I said I wouldn’t... I keep adding things I think are necessary or make it easier to use (as the latest build shows in spades). Then, on top of that, the code is so complicated and there are so many possible missions and activities and permutations of things that every change we make has the possibility of being the “butterfly wings in China causing snow in LA”. All I can say is that we still have a few things I’d like to look at, but I’m hopeful that we’re at least getting close. We just ran a series of scenarios for the military (as part of a JFCOM project), and POA-2 performed very well. In over 100 runs there were 3-4 crashes, but we went through literally hundreds of hour’s worth of intense game play on many different computers without a single problem. And the scenario was pretty complicated and diverse in terms of weapons systems and capabilities, and it used a very large map (~30,000 hexes).
I’d like to add one final thing, which unrelated to the program, and ask that if you hear of people complaining about not getting responses to emails they send to us, please either let us know or make sure they are sending to support@hpssims.com. It is my personal policy to reply to every email we get in some way (unless it is part of an ongoing play test issue or something like that). But in the real world sometimes email doesn’t get through. Spam filters may delete real emails (either their original to us, or our reply to them), mail servers go down at inopportune moments, or any number of things can happen. But we will never intentionally blow anyone off. So please ask them to give us the benefit of the doubt and try again. We can’t help them if we don’t know they have a problem or if they don’t get our reply.
Anyway, I apologize for the long-winded post. But with everything that has gone into POA-2 so far, from us and from you, at the end of the day I’m going to make sure it was something worth doing.
Thanks again!
-Scott
___________________-
Hi guys,
I just posted the latest patch on the web site.
http://www.hpssims.com/Pages/updates/up_POA2/Updates_POA2.html
The entire list of changes, and there are many, are in the Readme file. But a couple of the big things in this build are:
- “sanded off” some rough edges of scenario creation/adjustment
- added patrol movement paths
- completely reworked AI aircraft paths
- added a facing command (which will not be overruled by the AI except for targeting)
- updated AI support routines.
As I said in regards to the last patch file, in general things are taking longer to work on now than before. That’s true in terms of finding them in the first place, and then figuring out how to fix them. The reason is that the program is more complicated than ever, and we’ve already picked off the easy ones. There are still two bugs in this code that I know of, but haven’t been able to squash. One is an access error, and the other is a floating-point error. I believe the latter is in the sighting routines, but I’m not sure. They each happen very infrequently, even with the large and complicated test game I’m running (once every 500 turns or so), and it seems when I try to use any sort of debugging technique... they disappear. They also don’t reappear from running the autosave file, so I’ve concluded that we are probably using an un-initialized floating-point variable someplace, probably involving some sort of object like a bridge or IP.
Eventually, though, we’ll get them - especially now that the batch routine is available to us (put in for the military JFCOM project) so we can run games uninterrupted for days on end.
I also still want to work on the loading/unloading interface and issuing orders to dismounts after an unload action. That process is complicated right now, and you sometimes don’t get the results you want or expect.
That all aside for a moment, I want to say something to you about where this project is and where we’re going. I know some rumors have been floating around, but with the effort you’re all putting in I want to make sure you know the real deal. First, while I do have plans to do POA-3, the serious development of that will not start until I’m happy with POA-2. In fact, it can’t since POA-3 will *be* POA-2, but more “AI intensive” in the interface and ease of use. I don’t want to go into a lot of details since I haven’t really done much concrete on it yet, but one aspect I’m aiming for will be that players can assign complete missions at a higher level. However, obviously that is meaningless until I get POA-2 to work and act the way I want - kinda like putting a V-8 engine on a go-kart frame.
I also want to say that POA-2 is essentially all that Jeff and I have been working on since it was released. In the brief periods where we didn’t have any reports or issues to work on (the eye of the hurricane, if you will), Jeff worked on ADC-3, and I tried to catch up on other business related stuff... but those periods passed quickly and POA-2 took over again; POA-2 is our overall priority.
I don’t really know when POA-2 will be done. To begin with, even though I said I wouldn’t... I keep adding things I think are necessary or make it easier to use (as the latest build shows in spades). Then, on top of that, the code is so complicated and there are so many possible missions and activities and permutations of things that every change we make has the possibility of being the “butterfly wings in China causing snow in LA”. All I can say is that we still have a few things I’d like to look at, but I’m hopeful that we’re at least getting close. We just ran a series of scenarios for the military (as part of a JFCOM project), and POA-2 performed very well. In over 100 runs there were 3-4 crashes, but we went through literally hundreds of hour’s worth of intense game play on many different computers without a single problem. And the scenario was pretty complicated and diverse in terms of weapons systems and capabilities, and it used a very large map (~30,000 hexes).
I’d like to add one final thing, which unrelated to the program, and ask that if you hear of people complaining about not getting responses to emails they send to us, please either let us know or make sure they are sending to support@hpssims.com. It is my personal policy to reply to every email we get in some way (unless it is part of an ongoing play test issue or something like that). But in the real world sometimes email doesn’t get through. Spam filters may delete real emails (either their original to us, or our reply to them), mail servers go down at inopportune moments, or any number of things can happen. But we will never intentionally blow anyone off. So please ask them to give us the benefit of the doubt and try again. We can’t help them if we don’t know they have a problem or if they don’t get our reply.
Anyway, I apologize for the long-winded post. But with everything that has gone into POA-2 so far, from us and from you, at the end of the day I’m going to make sure it was something worth doing.
Thanks again!
-Scott