Terrain engine update fantasy list

[artik]

I'm sure they will get around to making a better terrain editor... I do agree that the map editor should be more.. user friendly

I'm not asking new features, I'm not asking anything... I'm asking that a program would work as it expected to do.

I'm a programmer.

I had written lots of software of different kind, from UI to low level hard core computing and physical simulation. There is one important principle: Fix bugs before writing new code (see: http://www.joelonsoftware.com/articles/fog0000000043.html)

Having multiple bugs in TE causes major problems to the terrain developers. I had just lost an entire work day I spend on a terrain due to corruption of a file. And I lost about 2-3 work hours on various bugs so I needed to fix things or find workarounds. And I had only started working on a new terrain!

So how many work hours the community members had lost due to bugs?

Of course I'm not trying to tell HTC how to develop software. I just ask for something that works as expected. AH2 is very stable and high quality software. But the tools around aren't. And not because they miss a feature or so, it is because they just don't do what they expected to do (a.k.a. have bugs... many of them)

[Tilt]

Tree foliage that is not impervious to ordinance........ would be  my prime wish even if the tree trunk etc was its old uber hard self I think the game play tactic of hiding in plain sight behind a couple of trees using the canopy of same trees as a revetment is old old now......


Still not up to speed with Skuzzy's logic re the effects of lag during a player destroying an object for a better view..........  I see the initial argument but cannot see how a graphically enhanced delay would not even stuff out........  if player 1 is destroying the object (tree) for a better view then he is not sighted until the object is destroyed.... hence upon firing at the object he would have to...... 1)wait for dust to settle 2) whilst re arming 3) sight & range onto target 4) fire  

this whilst the target may have been given notice of Player 1's presence by the wanton destruction of habitat.

That this would all use up resource to sap either the server or the FE or both I can understand

actually its the same for town buildings now......  (kill one and the object remnant is soft and can be seen thru)

Bridges are Gud. I don't mind if the are destructible or not. I think the problem with destructible bridges is linked to the local terrain design as a whole. If the bridge is the only route then its poor. If its the shortest of several routes then it's "interesting". Also the distance from spawn points (defending and attacking) make a difference IMO.

I can see scenario's where the bridge enables a spawn closer to the target field. If the bridge is destroyed then this spawn is lost but there must still be a an alternative. That's if the new terrain engine still keeps fixed spawn points!

[Tinkles]

Tree foliage that is not impervious to ordinance........ would be  my prime wish even if the tree trunk etc was its old uber hard self I think the game play tactic of hiding in plain sight behind a couple of trees using the canopy of same trees as a revetment is old old now......


Still not up to speed with Skuzzy's logic re the effects of lag during a player destroying an object for a better view..........  I see the initial argument but cannot see how a graphically enhanced delay would not even stuff out........  if player 1 is destroying the object (tree) for a better view then he is not sighted until the object is destroyed.... hence upon firing at the object he would have to...... 1)wait for dust to settle 2) whilst re arming 3) sight & range onto target 4) fire  

this whilst the target may have been given notice of Player 1's presence by the wanton destruction of habitat.

That this would all use up resource to sap either the server or the FE or both I can understand

actually its the same for town buildings now......  (kill one and the object remnant is soft and can be seen thru)

Bridges are Gud. I don't mind if the are destructible or not. I think the problem with destructible bridges is linked to the local terrain design as a whole. If the bridge is the only route then its poor. If its the shortest of several routes then it's "interesting". Also the distance from spawn points (defending and attacking) make a difference IMO.

I can see scenario's where the bridge enables a spawn closer to the target field. If the bridge is destroyed then this spawn is lost but there must still be a an alternative. That's if the new terrain engine still keeps fixed spawn points!

If one or two bridges were the only 'passage' across a river (was thinking of mini-rivers in towns  ) then the way to repair said bridge would be cargo via m3/251.   Bridge would take a few rounds at least, or only take damage from.. say SAP/AP bombs (making them used more).

Just a thought.

[Kazaa]

How about some better lighting, HDR, bloom?

Please respond.

[Easyscor]

I'd like bridges added to the MA as well, it would give an added dimension to map design. There is a problem with putting them on an MA map however. Someone playing with detailed water turned off would see the bridge but no river. A narrow strip of water like a river just vanishes with the game in this state, but driving into the invisible river will still stop the GV.

One solution would be to add river bank objects either side of the bridge or turn it into a bridge V base tile with river banks and a river coded in as objects. The other would be to make coastlines a non-removable graphics option in the game, so turning off "detailed water" would turn off the pretty wave effects but leave a flat blue water area with sharp coastlines. I'm not sure how much of an FPS penalty this would impose on a very slow system however.

I'm expecting that the option to disable Detailed Terrain will be removed. If I'm right, you'll only be able to remove Detailed Water, and its associated Water Reflections, but you'll still see all the shorelines as you can now. Detailed Terrain will show the riverbanks, and the removal of this option has been a needed change for a long time now.

[Easyscor]

Artik, the current Terrain Editor works very well when used exclusively with its' GUI. That said, its' core documentation and functions are very dated. Had you accepted my PM'd suggestion for a Skype meeting back in October, it would have saved you most of this grief. 

The Skype offer still stands and I think there's still a short window of time to put your terrain into play before the new game engine puts a halt to uploading new work.

[Drane]

How about an under water environment with terrain and objects that can be fixed or move under water?

Would be great for expansion later if not already included.

[artik]

Artik, the current Terrain Editor works very well when used exclusively with its' GUI. That said, its' core documentation and functions are very dated. Had you accepted my PM'd suggestion for a Skype meeting back in October, it would have saved you most of this grief. 

The Skype offer still stands and I think there's still a short window of time to put your terrain into play before the new game engine puts a halt to uploading new work.

Thanks for the offer. Really... I hand't forget it and I really appreciate this/

Finally working with wiki gives me good results. I have no major/unresolved problems with the functionality itself. Maybe it sometimes not intuitive or user friendly, but that is not the thing I'm complaining about. Also I'm not talking about small annoying bugs/unexpected behavior (that I can live with them)

The problem is that TE crashes, freezes, and what is even worth corrupts my data: sample 1, sample 2.

In the second case, I lost lots of work because it corrupted the file. Learned from the experience, now I put all the data under source control (git) and commit it every few steps, so if something bad happens I can trace back at what exact point it started and recover for future freezes and loose at most 10-20 minutes of work rather than many hours between the backups.

How would you feel if a Photoshop or GIMP was crashing every hour and was unexpectedly corrupting your files?

[Easyscor]

How would you feel if a Photoshop or GIMP was crashing every hour and was unexpectedly corrupting your files?

That's exactly what will happen with GIMP 2.6.x if you don't have enough memory when working with an intense waterd file. It happened to me about every 20 to 30 minutes. There are situations when building terrains require better hardware then the actual game! I'm not saying your hardware is at fault. I'll look at your examples and see if I spot anything obvious but this isn't the forum for an answer so I'll reply in a PM.

[SirNuke]

The player that destroyed the tree could have been doing it to fire at the remote player and would not wait for anything to clear.  To artificially restrict firing until everyone got the update would be frustrating as well.

I can't see how that is more frustrating than a bush stopping/flipping a king tiger at 30Mph, or tanks using trees as walls. Destroyable trees is a killer feature that would really enhance the game.

A few milliseconds of display lag and sometimes weird situations can be dealt with (we deal with warps all the time), we are not pro gaming in here.

[Tinkles]

Since a new terrain engine is in development. Perhaps an idea I mentioned a few months ago would be beneficial.

Instead of making all the trees destructible. Why not make certain areas of trees destructible?   Since the trees we have now are 'coaded' as indestructible (part of the ground terrain), why not make a few tree models that are made to be destroyed?  These would be in populated-only areas. Not in the boonies were even b29s wouldn't have a need to be at.

3k around GV spawns and on the way to towns/around towns and bases (up to 4k around bases and 2k around towns including inside the town).



If you think it is possible to do this, then do it. If not, I think most of us understand why. 


As an example.



Red dot is my wirble.

Area in the blue lines is the 'spawn point' all the trees in this area would be destructible, all trees outside of this area, wouldn't be.

All trees in the green area would be destructible also and again, only in the green/blue areas. So the trees on the mountains in the background.. would be the titanium trees.   

Would it be possible to copy the tree (or make a few models of trees) and make them destructible? Then put em everywhere near spawns and bases/towns (perhaps some ON the bases (if they had them, because I don't know) ), if it's possible.

[Easyscor]

 

Skuzzy has got to be wishing he'd used the company line on this one.

427* destroyable trees and man made objects per square mile x 512 x 512 miles equals 111,935,488 destroyable objects. That's almost 112 MILLION objects to track on your FE in real time. I can imagine what would happen if everyone in the arena decided to simultaneously start killing trees and buildings. Oh, the humanity!

*The count is take from ntt0005.til

Tinkles, imho this would need to be all or nothing because the terrain builders would need to manually do what you're asking, and you don't even have Detailed Terrain active now! At least, that's how that jpg looks to me.

[Wiley]

Skuzzy has got to be wishing he'd used the company line on this one.

427* destroyable trees and man made objects per square mile x 512 x 512 miles equals 111,935,488 destroyable objects. That's almost 112 MILLION objects to track on your FE in real time. I can imagine what would happen if everyone in the arena decided to simultaneously start killing trees and buildings. Oh, the humanity!

*The count is take from ntt0005.til

Tinkles, imho this would need to be all or nothing because the terrain builders would need to manually do what you're asking, and you don't even have Detailed Terrain active now! At least, that's how that jpg looks to me.

Well, you'd only need to keep track of the ones you can see.  What's ground crud render distance? 4 miles?  Just for simplicity's sake, make it square, so 16 square miles of constantly changing terrain under your aircraft.  16*427= 6832 objects that would need to update their status of whether they're up or down at all times when you're close to the ground.

They'd each need a unique identifier which would need to be able to handle the 112 million or more count, so an integer, positive means up negative means down...  That's still 2 bytes per object, 13,664 bytes that would need to be updated several times a second to be playable.

It's a significant amount of processing/data transfer.

Wiley.

[SirNuke]

Well, you'd only need to keep track of the ones you can see.  What's ground crud render distance? 4 miles?  Just for simplicity's sake, make it square, so 16 square miles of constantly changing terrain under your aircraft.  16*427= 6832 objects that would need to update their status of whether they're up or down at all times when you're close to the ground.

They'd each need a unique identifier which would need to be able to handle the 112 million or more count, so an integer, positive means up negative means down...  That's still 2 bytes per object, 13,664 bytes that would need to be updated several times a second to be playable.

It's a significant amount of processing/data transfer.

Wiley.

or the server could send info only about the trees that are down, as the server and client both have the map file.

[Wiley]

or the server could send info only about the trees that are down, as the server and client both have the map file.

Doh... good point.  Based on the way people act in most online games I've seen the terrain would be pretty much leveled most of the time anyplace people are busy though.

Wiley.