Careless Fences Cost Lives

[Other Means]

We can put men on the moon, but we can't keep a distant fence hidden in a video game?

BFC is going to be given a fair proportion of the US GDP*? Woohoo! Wargames for everybody!

*about three fiddy.

[ChrisND]

We can put men on the moon, but we can't keep a distant fence hidden in a video game?

Thanks, now I know I don't need to take you seriously.

[eltorrente]

Thanks, now I know I don't need to take you seriously.

No need to get upset over a silly comment.. take it easy there..

BTW, I DO understand something about making video games, as I've been in the industry for many years, with many shipped titles under my belt. This isn't an insurmountable problem - it's a matter of putting some man-hours on it if it's deemed important enough.

Battlefront obviously has a small budget and few man hours to go around if minor things such as this have been lingering for 10+ years.

[Steiner14]

We can put men on the moon, but we can't keep a distant fence hidden in a video game?

Are you sure about the former?

[JonS]

Given the programme can calculate lines of sight from square to square if the fence were "semi-human" then surely the game engine could simply not show physical changes if they were out of sight?

Think about what you're asking for a minute.

Constant LOS checks. From everywhere. To everywhere. All the time. Yeah, that sounds easy

OK there is a processing overhead but surely not huge.

Surely

Possibly it is accomplished through the engine - it must already be calculating which squares are in sight to do a check for troops ...

Must it?

[eltorrente]

Think about what you're asking for a minute.

Constant LOS checks. From everywhere. To everywhere. All the time. Yeah, that sounds easy

It wouldn't have to be like that, not at all.

Anytime an object is destroyed, it would do an LOS check from that square to enemy units. If the LOS check fails, then it isn't updated on the opponents map, and the opponent thinks the fence or whatever is still there intact, until he rounds the corner and puts eyes on it.

The destroyed objects become enemies units, in essence, waiting to be seen.

This isn't rocket science, but it would take programmer time, and for that it would need to be deemed important by Battlefront. How does it affect game play currently, and would the improvement be worth the time to implement.

Personally, I'd rather see them use the time to make vehicle AI work better..

[Sitting Duck]

Think about what you're asking for a minute.

Constant LOS checks. From everywhere. To everywhere. All the time...

I think I would rephrase this...

If I understand correctly, the game already handles constant LOS checks from every unit to every other unit.

I think what's is being discussed is a LOS check from each unit to every terrain tile. This could probably be handled in the way you suggest, but rather than use 100% real-time calculations, it might be faster to figure out terrain LOS and store it in a table. When "compiling" a map, a table is produced indicating LOS from each terrain tile to every other terrain tile. As play ensues, spotting rules can be applied that leverage the baseline terrain LOS data in the table to determine if a given unit can spot terrain/changes.

If I understand the point correctly, currently "terrain spotting" is instantly known to all units and players. We all know every detail of the maps we play on - including changes due to unit activity - before our units have LOS. We might call the current situation Omniscient Borg Terrain Spotting.

One possible implementation of less omniscient terrain spotting would be to implement 2 additional maps, for a total of 3 maps for each battle: the "real" terrain map with full detail and then one terrain map for each player that is changed as the player develops LOS to individual terrain tiles. Note that the respective player terrain maps would need to updated as a player's units develop LOS to individual terrain tiles that have been changed since they were last observed.

[bTW - Since there is only one terrain map for each side, I guess this would really be like Borg Terrain Spotting, but it does get away from omniscient spotting since there must be some unit with LOS to view the terrain/change.]

Another possible implementation of relativistic terrain spotting would be to implement an additional map for each unit. I'd guess fully relativistic terrain spotting would be too taxing. Furthermore, I don't think fully relativistic terrain spotting is really too different from the previous idea.

[Guest]

Anytime an object is destroyed, it would do an LOS check from that square to enemy units. If the LOS check fails, then it isn't updated on the opponents map, and the opponent thinks the fence or whatever is still there intact, until he rounds the corner and puts eyes on it.

The destroyed objects become enemies units, in essence, waiting to be seen.

In the meantime doing LOS checks to all enemy units? Depending on map, scenario, and number of units, you could be talking about hundreds or thousands of additional LOS checks every frame. Recall that terrain damage isn't just created by vehicles driving through things.

A big arty barrage at the back of the map might smash a few dozen wall / fence / hedge sections. If you've got a battalion advancing that doesn't have LOS to that area for half an hour you're generating a very large number of LOS checks on a regular basis for little return.

Programmer time isn't the only limiting factor on things like this. Feasibility and performance hits also have to be measured against utility.

[sburke]

or you could just buy a few jeeps/kubelwagen and go a little fence crazy with your opponent wondering are those tanks? Maskirova as the Russians would call it.

[eltorrente]

In the meantime doing LOS checks to all enemy units? Depending on map, scenario, and number of units, you could be talking about hundreds or thousands of additional LOS checks every frame. Recall that terrain damage isn't just created by vehicles driving through things.

A big arty barrage at the back of the map might smash a few dozen wall / fence / hedge sections. If you've got a battalion advancing that doesn't have LOS to that area for half an hour you're generating a very large number of LOS checks on a regular basis for little return.

Programmer time isn't the only limiting factor on things like this. Feasibility and performance hits also have to be measured against utility.

Agreed - but there could be a few compromises in how it's implemented. For instance, objects which are destroyed by artillery or large guns will just show up as normal (since you'd see the explosions in that area in the distance and it's obviously not something sneaky being done). Maybe the ONLY destroyed objects which get LOS checks are those which are destroyed by vehicles driving over them, or infantry blowing holes in them, so there wouldn't be too many.