Build 4.4.258 Feedback

[navwarcol]

Phoenix,
Are you saying the scenario crashes every time you run it(battle of the roadblocks)? Or are you saying this patch crashes every time?

[RockinHarry]

Beeing asked for showstoppers, here´s some remaining through all the previous builds:

The dashed line draws the course of retreat for a german para unit, that was previously ejected out of foxholes by a single UK Inf Coy of about 80% strength of the german unit. Took just one casualty for the germans to say good bye, although they have veteran stats equal to the single UK unit. The UK AI also didn´t bother to use Arty until the germans were out in the open.

In revenge I battererd the single UK unit with all available Arty and after purpose done the now seldomly seen rout draws the UK unit about 1 mile into the german "lines".

So the main showstoppers still are, dug in units get on the flight without noticable pressure way too often and retreat/rout paths as well remain FUBAR. AI Arty usage appears to be a game on its own and remains a riddle to me.

From my playtesting these are not single occasions and happen as much with stock maps/missions, as well as with my own stuff. So this is the main reasons (beside a number of minor ones) that I´m actually done with the Veritable project.

[RockinHarry]

And the continuation. The UK routers received bits of fire from rearward german units and decide to make more sightseeing in german Hinterland. As soon as they get into recovery, they´ll make some nice FO and advance detachement, giving the AI Arty some nice yet unseen targets. I marked what I consider "frontlines" with red and blue lines, to show that the AI obviously has no awareness about such things.

[Remmes]

ORIGINAL: phoenix

I have an easier repeateable one, Ramses. I don't have to do anything at all - just run it - it crashes every single time. I've just mailed you it. Wondering if you could check - make sure it's not my machine before I pester anyone with it.

Peter

Confirmed, exactly as you described. You have mail Phoenix,

Ramses

[Arjuna]

It's usual for units to end up with no current active order. It happens for a variety of reasons.

Example 1. A unit may be performing a peripheral task within the boss's plan. Eg Say the main force is attacking but at the time the plan was developed the unit was unable to reach the FUP. In such a case it may be given an independent attack or told to just to stay put. It gets given its own order with a time frame. Now if the main force slips its attack, we don't slip the peripheral tasks. So these may actually end. In which case the unit goes on hold at its current location. That is perfectly fine.

Example 2. A unit is given an order such as Move to X but in the course of moving it gets to a point where it becaomes stalls and replans. This fails too so it abandons and waits till it gets new orders. If we obviate the stall and say go for it anyway then you will be complaining that HQs and bases are leading thr charge. So I believe this is the lesser of thew two eveils. You as the player need to monityopr the messgaes and the unit logs, so you are aware of these cases.

[Phoenix100]

Navwarcol - it's a crash in a customised HTTR add-on scenario I'm playing. It crashes without fail when I run it from the last save, same time (game time) each and every time. I've sent it to Dave, anyway.

[wodin]

Dave I think you should just sort out obvious bugs rather than tweaking gameplay or features...look at that another time..otherwise this is going to be never ending.

The game has it's quirks but no worse or more than any other and I for one think it's better now than it was on release and better each patch.

Get any obvious bugs fixed like CTD's then the COTA exp out then start looking at gameplay and feature teaks with the EF game where you have more time to experiment and check code and outputs etc. I do feel Harry needs to be on the testing team for the EF game as you wont get a more methodical tester.

[Arjuna]

ORIGINAL: phoenix

I have an easier repeateable one, Ramses. I don't have to do anything at all - just run it - it crashes every single time. I've just mailed you it. Wondering if you could check - make sure it's not my machine before I pester anyone with it.

Peter

Peter,

Thanks for the save. Yes it does repeat. The error occurs in the new code I wrote to ensure static arty get accessed for onCall fire support requests. This searched the onMapBoss's opPlan for any current plans that have a static arty unit as their subject. Alas I hadn't catered for the fact that inside the onMapBoss's opPlan you hvae the default plans for each side task. These remain current even when no unit is currently assigned. So a simple test to see if it has a subject first before testing to see if it's a static arty fixes this crash.

BTW this had nothing to do with Ramses crash. I'll comment on that seperately.

[Arjuna]

Ramses,

Your save is tad more challenging shall we say. So far I have been able to confirm that the crash does occur in the save you gave me when you run the game fast and issue waypoints to KG Holtmeyer. The error appears at first blush to be inside MapVisitTable""GeneratePath() because it reckons it cannot find the previous loc of the route. The thing is that it also occurs if you select some other force as well and issue waypoints while running fast. However, this crash occured inside MapSearchParams::GetTravelCost() claiming that a loc previously deemed acceptable now has a cost of zero.

It does not occur if the game is paused and you issue the waypoints. I have tried it on a vanilla run of the roadblocks scenario and cannot get it to occur - ditto on the tutorial.

Somehow the map search params are being corrupted. It may be that the scenario data is currupted. I am asking Miquel to look into this. Very interesting.

[navwarcol]

I do see a lot of this behavior as well(retreats into the enemy/away from ones own LOC). I am assuming that since it seems to have been present in the prior patch as well, that it is probably more complex than I am getting ready to cavalierly suggest[:D] but here goes anyway... is there any way , since we can bring a colored line to its supply source, and immediate commander, to just make it a guarantee that when it retreats it automatically heads down the direction of one of those colored lines, either toward its LOC or towards its immediate superior in the OOB? Routing..I have less issue with as that basically implies just a panicked run...which should at least be away from the unit forcing the rout, but could easily still be towards other enemy units.

ORIGINAL: RockinHarry

And the continuation. The UK routers received bits of fire from rearward german units and decide to make more sightseeing in german Hinterland. As soon as they get into recovery, they´ll make some nice FO and advance detachement, giving the AI Arty some nice yet unseen targets. I marked what I consider "frontlines" with red and blue lines, to show that the AI obviously has no awareness about such things.

[RockinHarry]

ORIGINAL: wodin

Dave I think you should just sort out obvious bugs rather than tweaking gameplay or features...look at that another time..otherwise this is going to be never ending.

The game has it's quirks but no worse or more than any other and I for one think it's better now than it was on release and better each patch.

Get any obvious bugs fixed like CTD's then the COTA exp out then start looking at gameplay and feature teaks with the EF game where you have more time to experiment and check code and outputs etc. I do feel Harry needs to be on the testing team for the EF game as you wont get a more methodical tester.

I´m aware that my feedback is little productive ATM, but on the reverse the game is currently not that productive on me, with regard on Veritable project, that so far I spent many man hours with. Doesn´t mean the game does not work within confines of design purpose. I´m sure Dave and team know about game engine capabilities and limits themselves without my input, so I see no real benefit to the excellent team already at work.

I think the current public beta testing is bits of a mess, since everybody has a different focus and different issues. Maybe Dave should decide on a particular mission (for us) to use for bug hunting, thus easing the process a bit.

[Arjuna]

ORIGINAL: navwarcol

I do see a lot of this behavior as well(retreats into the enemy/away from ones own LOC). I am assuming that since it seems to have been present in the prior patch as well, that it is probably more complex than I am getting ready to cavalierly suggest[:D] but here goes anyway... is there any way , since we can bring a colored line to its supply source, and immediate commander, to just make it a guarantee that when it retreats it automatically heads down the direction of one of those colored lines, either toward its LOC or towards its immediate superior in the OOB? Routing..I have less issue with as that basically implies just a panicked run...which should at least be away from the unit forcing the rout, but could easily still be towards other enemy units.

ORIGINAL: RockinHarry

And the continuation. The UK routers received bits of fire from rearward german units and decide to make more sightseeing in german Hinterland. As soon as they get into recovery, they´ll make some nice FO and advance detachement, giving the AI Arty some nice yet unseen targets. I marked what I consider "frontlines" with red and blue lines, to show that the AI obviously has no awareness about such things.

There is code in the retreat/route routines that attempt to go back along any route it currently has or towards its currSuperForce. However, this can be overridden if in doing so would expose it to high enemy fire. I am sure if I hard coded it so that it always did so then I'd receive complaints that it was being stupid in not seeing that it was effectively cut off.

For now I'm leaving the retreat/rout routines as is.

[dazkaz15]

ORIGINAL: wodin

Dave I think you should just sort out obvious bugs rather than tweaking gameplay or features...look at that another time..otherwise this is going to be never ending.

The game has it's quirks but no worse or more than any other and I for one think it's better now than it was on release and better each patch.

Get any obvious bugs fixed like CTD's then the COTA exp out then start looking at gameplay and feature teaks with the EF game where you have more time to experiment and check code and outputs etc. I do feel Harry needs to be on the testing team for the EF game as you wont get a more methodical tester.

I agree wodin

I'll keep my game balance feedback to a later time.

The CTD would be 1st on my list of things to fix, although for me it is a lot more stable than the last build, but still have these CTD when manipulating order waypoints especially with the Del key.

[BletchleyGeek]

ORIGINAL: RockinHarry

And the continuation. The UK routers received bits of fire from rearward german units and decide to make more sightseeing in german Hinterland. As soon as they get into recovery, they´ll make some nice FO and advance detachement, giving the AI Arty some nice yet unseen targets. I marked what I consider "frontlines" with red and blue lines, to show that the AI obviously has no awareness about such things.

Indeed, retreat lines favour keeping the unit out of harms' way (which is realistic), giving to this to much more weight than to fall back towards one's forces. Introducing the latter preference, is doable and will happen sooner or later in the code. Just keep reminding Dave of this.

On the other hand, routing units - whose morale has been destroyed - shouldn't be able to request on call support as you report these guys are getting.

Yet, all other things being equal, this happening is quite reasonable. Why? There are no German units detached to corner and destroy that British unit forcing a surrender. The British unit is moving quite freely along what is - for all intents and purposes - space empty of enemy forces and APer fire envelopes (as I understood, Harry has tweaked the ranges of firearms so they are shorter). Talking about "front lines" on a system where ZOC's are painfully realistic, doesn't make much sense. Expecting to annihilate a force dispersed over several hundred meters with indirect fires is an unrealistic expectation.

So yes, retreat lines should give priority to rejoin the side forces rather than following the safest route (which might indeed lead troops behind enemy lines). The fact that these safe routes exist, on the other hand, point to a somewhat faulty in planning your defence positions. Flak units or AT units, deployed in the rear on commanding positions with Max Aggro settins will engage with fire any enemy unit coming within their effective APer/AArm range. That would reduce the possibility of having such "gaping holes" in defensive deployments. Keeping a mobile reserve to deal with infiltrating enemy units isn't something unreasonable, either.

Regarding your woes regarding the "arty game". If you have stated these concerns in a number of posts, I'd appreciate very, very much if you could sum them up in one single post, for discussion and focus.

And about the beta testing process: this is public beta testing and it becoming a big chaotic is quite normal and reasonable. If anything, I'd suggest that CTD's and other, rather than being reported in this general thread, would be reported in the Tech support forum (and in these General discussion like threads pointers made to those threads).

By the way, I think Dave invited you to become part of the internal beta testing team. Did you follow up his invitation?

[jimcarravall]

Harry,

I'd like to see some more focus beyond complaining about what's not working to someone's personal preference vice the latest immediate game release.

I see all the pieces in place to dynamically model any conflict (terrain, organization, weapons effects, command / control mechanisms, soldier capabilities (morale / cohesion etc.), supply effects, command effects) but the feedback tends to focus on what players expect at the macro level instead of what caused the effect at the micro level.

Perhaps there's a thread needed to focus on the "real" aspects of combat to the exclusion of the users' "desired" reaction to their orders.

If you want to model the 82nd's Airborne seizure of the Rhine Crossing in Nijmegen (without the capability to "cross" using alternative means to secure bridges), fine; post to your peers.

If one is interested in (spending Dave O'Conner's) development dollars for a more realistic simulation, fine; add to Dave's ulcers .

Of course, I'm being facietious. But, there's a difference between seeking out the "flash bang" and supporting a "bang" that fits into what was available to "flash" in the first place.

Hope this helps.

[navwarcol]

ORIGINAL: Arjuna

ORIGINAL: navwarcol

I do see a lot of this behavior as well(retreats into the enemy/away from ones own LOC). I am assuming that since it seems to have been present in the prior patch as well, that it is probably more complex than I am getting ready to cavalierly suggest[:D] but here goes anyway... is there any way , since we can bring a colored line to its supply source, and immediate commander, to just make it a guarantee that when it retreats it automatically heads down the direction of one of those colored lines, either toward its LOC or towards its immediate superior in the OOB? Routing..I have less issue with as that basically implies just a panicked run...which should at least be away from the unit forcing the rout, but could easily still be towards other enemy units.

ORIGINAL: RockinHarry

And the continuation. The UK routers received bits of fire from rearward german units and decide to make more sightseeing in german Hinterland. As soon as they get into recovery, they´ll make some nice FO and advance detachement, giving the AI Arty some nice yet unseen targets. I marked what I consider "frontlines" with red and blue lines, to show that the AI obviously has no awareness about such things.

There is code in the retreat/route routines that attempt to go back along any route it currently has or towards its currSuperForce. However, this can be overridden if in doing so would expose it to high enemy fire. I am sure if I hard coded it so that it always did so then I'd receive complaints that it was being stupid in not seeing that it was effectively cut off.

For now I'm leaving the retreat/rout routines as is.

Ah.. OK, this makes sense to me then, if you are saying that the retreating unit basically made a conscious decision to go in the "direction of least resistance"..as it is a move I would also consider on occasion as commander. I will accept it gladly as long as it is not just a sign that on retreat it wanders aimlessly

[Arjuna]

Re OnCall Support. Units that are routing or rout recovering cannot call for OnCall Spt. End of story.

[RockinHarry]

ORIGINAL: Arjuna

Re OnCall Support. Units that are routing or rout recovering cannot call for OnCall Spt. End of story.

Not retreating/routing, but sooner or later they recover to normal state (as happened in the same game as posted above and screeny below). Would be interesting to know, where the formerly routed UK unit is now moving to and who calls Arty on the german IG unit further down. As said, I also observed something like that in original St vith (german side played) and has nothing to do with my own files.

[RockinHarry]

ORIGINAL: jimcarravallah

Harry,

I'd like to see some more focus beyond complaining about what's not working to someone's personal preference vice the latest immediate game release.

I see all the pieces in place to dynamically model any conflict (terrain, organization, weapons effects, command / control mechanisms, soldier capabilities (morale / cohesion etc.), supply effects, command effects) but the feedback tends to focus on what players expect at the macro level instead of what caused the effect at the micro level.

Perhaps there's a thread needed to focus on the "real" aspects of combat to the exclusion of the users' "desired" reaction to their orders.

If you want to model the 82nd's Airborne seizure of the Rhine Crossing in Nijmegen (without the capability to "cross" using alternative means to secure bridges), fine; post to your peers.

If one is interested in (spending Dave O'Conner's) development dollars for a more realistic simulation, fine; add to Dave's ulcers .

Of course, I'm being facietious. But, there's a difference between seeking out the "flash bang" and supporting a "bang" that fits into what was available to "flash" in the first place.

Hope this helps.

Well, bear in mind that my actual drive is the creative side of things, with Operation Veritable aimed at beeing a major contribution to the game. If I would be just a player or consumer I´d probably be less nit picky about things. Making maps and creating custom ESTABs too is a major and very time consuming effort and I don´t want to see "fantasy" battles unfold on realistic maps and with deeply researched ORBATs. Maybe my bad choice to attempt on that particular operation, but I rather scratch now on the limits, than after spending many more work hours on that.

Thus I drop out here and try not to be PITA for Dave anymore. There´s already enough to work on for the final patch without my meager contributions.

Hope this helps too

[Arjuna]

Chill out everyone. While I appreciate that everyone is trying, in their own way, to make this the best game we can my preference is to receive feedback from whatever bent. I'll make the call as to what we filter out and what we focus on in terms of development. I have no problems in saying we won't change this or that.