scary_pigeon

yes. i forgot to reply to your pm phlerp. we are planning on having AG radar since important aircraft we're modelling had them. what are your ideas?

a small thing, but for us hopeful types - being mentioned in the press - however briefly is a positive feeling. its almost like getting paid. http://www.pcpilot.net/(yzvfe1unuzyopvauj2...image=t6&H=NEWS

it was because i was too lazy to add them: other stuff to do and all that. main reason to do it in my opinion is so that splashes from vertically above can be seen.

I continue to develop boring ungraphical code things. we took a decision to move to script approach for ai and physics and generally end all the hard coding for things like placement of cannons, calliber of wepons and things like that. So i've scripted cannons that can have their own texture, their own caliber, fire rate, etc, etc. and be selected individually for firing from the AI script. So in test now we have a harrier and a skyhawk that take off, turn on each other, fire missiles then use cannons. Recently a new guy has become involved in helping us. another wanabe games programmer like me. He has been making some interesting changes to the fm lua script. So thats how things are now. all this ai combat and such is occuring on the developing server. I will next be writing a client to view this.

it is likely. i kind of think the particle effect to water splashes kind of works though.

http://www.thunder-works.com/stanley-airfield-JT.rar collection of test shots. an example of one EDIT: notice that the C-130 Hercules (argentine air force) appears in game for the first time. :)

might be in some ways a different sort of game? our idea would be in comparison to falcon4.0 the game experience is sort of 1970's-1980's level tech for us, in falcon 4 - very modern stuff. but the avionics we model - this ought to be done top notch. as said, we're a small team, small budget so for us, the simpler tech level of the equipment we model makes its a tiny bit easier to do right. hopefully the fm will prove top notch too. falcon4.0 had a good dynamic campaign which many people appreciated. For us this is an important goal to. Being honest, this is a rather a taxing task. its notable that not that many flightsims have dynamic campaigns apart from psuedorandom mission generators. We will do what we can - and already, the essential stuffs been programmed - the player bubble concept: this hopefully lets us model extensive battles without crushing system performance. I think now the hardest problem is becoming gameplay programming. the mission/campaign system is all scripted and it is this script were the cleverness of having a decent campaign experience emerges or doesnt. The script will be open for anybody to enhance or replace. I think we'll do a good job of this, others may improve it.

we've got that book. it was good, interesting and kind of useful.

its amazing how simple niggely things can take so long to fathom. there was an issue with missiles being fired - then when it comes to remove them from the system, they were mysteriously hanging in mid air. after many hours I found out why and all is well. in the simple test, the AI fighter plane turns on the other fighter and launches its missile. the missile then flies to the target - doesnt cause any damage and flys off at some unguided tangent. we've not had guided missiles before. so it was fun to see it happen. what will happen next is cannons. in previous movies we've made you will have seen cannon firing. This was done without creating specfic places for the cannon. now we'll have scripts to tell us where the cannons on the aircraft will be placed, and what sort they are.

now they are guided on target - but some bugs have been reveiled. for some reason the number of objects in the game keeps going up in a particular link list -all copies of the original test objects. the end result is that deleting objects from the scene takes ages. have to figure out the cause and solution to this later after work.

at the moment the missile are guided - but they fly the wrong way. i have to improve script code.

missiles. i read somewhere that they have a high power phase to start with then a sort of cruise phase where thrust is less but constant - or is that just for SAM's is it true at all? i'm going to be modelling some missile FM's shortly.

right now, I'm trying to add in missiles! only just now are we adding in guided missiles. I'm pretty excited though about some improvements to AI - simple stuff that lets us quite precisely fly the aircraft without resorting to fake AI. still working towards this demo which will be out soon :P

"tally ho"

development has not stopped, dont worry about quietness. sometimes it pauses - i got a new PC and well, part of the problem is like this: I am writing on a game because i love playing games.. ..and the new PC works - it proved impossible to resist spending my holiday playing games. dante continues to do work on the objects, I am working on the AI mostly.

recently all my effort is to add this scripted AI to the behaviour of objects. Here we see a sea harrier, with a bomb 'attached' somewhat, above the aircraft. Scripting gives the position of where the missile type object is attached. script can then release that object. Its quite generic - I attached a sea harrier as a 'missile' to the sea harrier too. Its incredibly boring this part of the project. to get the AI to fly the plane in a non fake physics way - just to take of and follow waypoints seemed to take me days and days - thousands of failed take offs with different parameters. I just tried that now actually, attached a sea harrier as a bomb and it worked nicely as expected (of course the mass of the 'bomb' isnt being added to the physics model at moment) the harrier sort of glides into the ground. and does that nice bouncing stuff.

this is been a sort of unpublished goal for some time.

yes, local host will default address thing. 127.0.0.1 a nice thing about putting it on another PC is that the AI and all that is done server side. I would imagine considerably improved performance on the client side which just concentrates on rendering the locality.

how many of us? there is just me and dante 'full time' well really its part time since we both have jobs. Theres a few helpers mostly on the art side - and possibly somebody is going to replace our terrain with something better - otherwise i just update what we've got when I get round to it. at the moment: continuing the scripting approach, doing the AI in lua. I'm looking forward to thursday - a day off from work - so I get a bit of time to get more of this AI done. whats done so far is fairly basic still. testing the lua approach - I got the aircraft to take off conventionally. I must say, im not much looking forward to doing the vertical landing AI. it seems quite fiddly. conventional take off was kind of easy - just put throttle to max, pull up, retract gear, fly at particular height. having said that - unless cheating approaches are used - conventional landing in a precise area might be a bit harder than vertical landing.

that was a useful post mbot. I had peceived ground effect to be a stabalising effect. i thought that when it rolled it would recenter as well as give exta lift cushion. :) and its nice to know people are reading this.

hello :) quite a lot has been done. but little to show. graphically the difference between using a particle mesh data structure to hold objects and a link list are nill. Its just in frame rates and expnsion. dante has been busy created a pucara cockpit. He's now gone on an expedition to argentina taking some copies of what we currently have availble to show some project friends what we've got so far. Its a case use your imagination, Imagine if this cockpit demo was working with the carrier take off and landing demo, plus that bit with the combat in.. ..and imagine it all as a game. thats the state the project is in. a state were we must bring all elements together. Also dante apparantly has organised with curator at museum to sit in take lots of photos of cockpits. Pucara's and Canberra's are available. this will let him get exactly what he needs to complete more cockpits to his usual splendid standard. I did something minor and new today: before we'd had bomb and bullet splashed, thought we'd add splashes for the aircraft blowing through the water. - code wise, a decision has been made with regards to missions and campaigns and internet. This is one the same thing now project wise. Before we had the idea of the single player mode would be run in a very separate way to the way online play would be done. It turns out technically for me to do something that I think is kind of neat and tidy. The server is the mission/campaign manager. This is the program that will do all the funky particle mesh - player bubble goodness. A client app will then interface with this and receive appropriate telemetry. In this way, we will not be making any mistakes about tagging on network code at end, or tagging on campaign at the end and coming up lacking. instead, it will be all one thing and that there will be options for people to run their own little persistant falklands wars. I think this will be excellent for cooperative team play as well as straightforward dogfight mode. I am thinking that I am looking forward to this game. I play wwii online, flying the stuka quite often at the moment. And when I play it i am always thinking, I cant wait till I'm doing this in JT. Flying a pucara probably, napalming a load of my virtual countrymen. In IL2, in hyperlobby, for while my username was Vulching Hero. I used to fly an IL2 practically exclusively and alls I did was vulch. I never got the satisfaction of a dogfight kill. Yet with vulching I always found great satisifaction. The tatical forcefulness of systematically destroying the enemy airbase defences. then the orgy of destruction begins... ...In jet thunder- vulching hero will be in the air again. flying harriers hovering over a spawn. perhaps a pucara loitering over a temporary harrier runway. Its going to be so much fun.

just received from natural point, a track ir vector expansion. this is interesting. I rather like the idea of leaning left and right and stretching up and down and stuff like that from our virtual cockpit. So I think after the usual boring programming im doing now (no physics or graphics, just data handling) I'll take another look at this TrackIR thing. Probably I will send some code to their programmer and hopefully he might be able to apply trackir to that. but in the end, we'll get it working somehow.

haha yeah. it has a serious side though. it is a wasteland of inbred englanders who farm sheep and such.. ..but they guard for britain a notion: Antartica. By having a presence in the south atlantic we gain a fact on the ground argument for a stake in any future spoils with regards to antartica. the demonstration of british projection may remind other south american nations that might be interesting in future antartic claims to be aware of britains capability and interest in the region.

just an update. entering a murky boring graphicless time now. I've decide to kill all birds with one stone now. the campaign system is going to be network based. In one player mode, the system will just link to the localhost address 127.0.0.1. I like this as it creates a uniformity. its a bit complicated but when its done, its all done (Apart from content and FM tweaking) The campaign server will know everything and talk to clients on a 'need to know' basis. Coupled with some pretty cutting edge cunning lifted from my MSc project hopefully the netcode will work quite well when talking across the net, not just to localhost. Of course, there might be a catch. At the moment my network coding skills is limited to TCP/IP which is laggy and slow mostly - so there will be some fun getting the thing to work as fast as possible using precedence filtering and what not. Though in tests last year with what is now inferior code between me in UK and dante in Brazil were were able to have some fun flying around. Lag wasnt completely aweful despite the distance.

that was intereting. I'd not thought of the technical problem of one object being at the edge of your bubble due to movement and engaged with an object outside of the bubble. perhaps I'll just ignore the problem and nobody will notice? :) players would just assume that they'd dissengaged or one party had backed out - and as you approach, the bubble expands towards the disengaging aircraft and the guy doing a 'runner' regains his courage. its funny. I've done combat demos of the simplest AI and had my friends playing them. They were forever attributing all sort of notions to very simple but emergent behaviour. Oh its angry now, oh look its running away. even though the routine was just a simple lock on to nearest object and follow.