The Man Without Qualities

yay! so maybe OFPS-BECTI + GcCTI hosted at same servers after all?

That is something that BIS as entire company struggles to make it reliable. Hence, Jezuro alone won't fix it. It is better to address AI issues in the threads related to this issue:

https://de.wikipedia.org/wiki/Multiple_Launch_Rocket_System Here you see, some rockets can travel 80km+ so we would need maps of 100-200km lenght each side so 10.000 -40.000 km2 size to be able to maintain a strategic distance to stay out of range. https://en.wikipedia.org/wiki/Mortar_(weapon) Men-portable mortar do also 5-12km already, so that would be acceptable for the largest BIS islands.

I guess I understood Jezuro in a way that he want to bring a CTI which is more "back to the roots". At least I hope that is that way. Having drones flying around marking the targets and firing with hitech arty hitech shells across the entire range from 1 base into another would jeopardize that "back to the roots" idea. Jets are bad for game idea too, I hope I see not more then transport copters in his version, everyzhing else could be done in custom versions.

Hi Jezuro Picking up dvdbrewster Qs about HC interworking...is you mission optimized/tested to work togeter with HC? HC as such is a workaround, but for most CTI it helped to deal with more AI at server side. Do you have a test environment with at capable server with multicore-CPU and some HC clients supporting the server? Cheers

MP support! Ich will ein Kind von Dir!

Oops! That comment confused me, thought the is also a MP version......

Any chance that someone is hosting latest GcCTI as MP for public testing access permanently? Especially with proper setup HCs? Otherwise performance testing and tweaking would be difficult. I was trying to host server&client on a 4 core locally, but I failed to get it running with proper performance.

For pure function testing Zargabad is always welcome. Limited amount of towns, resource-friendly desert (but I hate desert!) vehicle-friendly, CTI like missions could be consumed like a snack rather then a 15 hours warfare.

Did some 1st playing, your mission has more or less the same issues like BECTI based CTIs when it comes to MP performance: - sudden server&client FPS drop when town defences are spawned - to many unnecessary spawned AI, see sugestions here ... Following the suggestions I made previously it would... ...on game start (without any faction spawning town defences) the faction set to act as "resistance" as in other CTI would start with a larger "cash" pot and with a premade base able to fill up the squads faster then to normal factions. The normal factions could grab towns pretty fast since the would need to just overwhelm the few AI of town guards and AA soldiers to get those towns. But soon they would meet the squads of the "Resistance" faction aiming to get back towns. Another option would be that the "resistance" faction is getting at game start fully manned and equipped squads spawned with some delay between each spawn and positioned at some spawn locations close to towns close to the normal and hostile factions. That way they could abort normal patrol duty and enter reinforcement pattern to help the few town guards of the towns attacked 1st. That way would the server load would rise pretty fast at game start, but would be more predictable over the game duration. And no extra town defense groups that might add up when the game becomes busy.

I guess he is keen on bullet-accurate equipment status when you leave. Saving gear-loadout templates works at almost every CTI mission I am aware of...

:-) So you fight with that too... But keep in mind that you eliminate RL simulation. In RL artillery teams far behind frontline (1-40km) shelling around for attack preparation etc. Won't work if AI not active. Especially @ OFPS this is also used as an exploit, even more if commander can activate towns manually. They order attack teams into towns and let them hide before they activate a town. More promising is IMHO to limit the amount of permanent AI to some patroling guards and some AI for air defense. The rest can be managed by AI teams under full control of the (AI or human) commander. Traditionally CTIs were designed that 2 factions have an equal amount of playable slots that could be played by humans. In the CoC, those slots were below the commander, any of those slots could become commander. In addition all conquerable towns on map could spawn defense squads - which was an issue when the total amount of groups was limited. Best approach was 3-side CTI where the "Resistance" was exactly like the 2 other factions except some AI routines and weaponary to be more like "local tribal with low-tech weapons". Driving that to the edge would mean to design it that as many factions could be in a CTI as supported by BIS, 1 is set as "default" owner of all towns (except those towns that are set to be owned by factions at game start). 2nd to not using full temporary town defense groups which become larger with every defense level upgrade and could kill the server if a low flying plane activates all towns @ island BUT provide larger amount of teams per side under the control of the commander. That approach plus setting per game setting the amount of slots open for humans would keep some squads exclusive for the (human or AI) commander who could assign them to defend towns or patrol between. That is already available in many recent CTI versions and should be sufficient. If commander does not assign nuff squads to protect rear towns or do humans not following orders---then it is ok if they loose. Calc example: 3 factions, 25 squad slots each, 10 reserved for rear duty, 1 side owns all towns at start. 75 groups with max. 20 units each should be the upper limit for all factions together here. Base service must be done by commander when assigning base service they act as guard except some structure construction is in queue - then they act as worker and perform funny annimations. An island with 40 towns would have 80 groups, 40 for guards and 40 for AA - each group not more than 4 units at any time. Guards are just there to provide the commander/faction itelligence about the enemy activity based on sensors as designed by BIS. AA should be just there to prevent/disable careless movement of enemy air over hostile area and rapid town grabbing by just flying to enemy towns. Intellygence by both group types could be used by AI commander routines to send AI squads meant for rear area security efficiently around, if nothing special just in a pattern on patrol between towns, if town is under attack then as reinforcement. Of cource that won't work with vehicles because they would get lost over time due to BIS pathfinding and AI driving issues. But infanry should be sufficient if nuff squads wandering around. Saving large town defender groups should eliminate the performance issues and gaining more emphasis on commander skills and communication. Ooops, sorry for 2 much text...

Ohhhh, urgently required! :-) May you drop some hints about the intended design of the game logic? I loved the general concept since MFCTI to jump into AI slots in the CoC and take over AI duty and leave at any time and AI takes back duty. That was dropped/did not work in A3 pretty good, even Benny had to fight with tons of bugs. Well, before repeating a lot: https://forums.bohemia.net/forums/topic/204731-favouriteleast-favourite-aspects-of-warfarecti/?page=2 Basically a (feature-) bastard of crCTI and BECTI, but bug free or with workarounds for BIS bugs....would be a dream. What I understand is that almost all MP missions involving massive utilization of AI suffer from buggy headless client implementation. Is your approach targeting large volumes of AI/player? BR

Is it a fresh CTI code or is it recycling of code made by Benny/Cleanrock/Melvin?

So I concur that in any case the vehicle would loose some speed, most old Russky-Tanks a lot, Leo-2 less. Why is this working for wheeled simulation? When I turn there I loose speed.

OK, then move it to the todo for A4...

Sounds weird to me. Guess it needs a tiny fix, tanks in RL tend to slow down whenever steering actions triggered, 1st because most common method is to use breaks at the track of the side one wanna turn to. 2nd because turning causes higher mechanical resistance everywhere at tracks and wheels. You consider scotg's finding as worth to fix?

In RL, the radiation of such systems (Panzir, Tungushka) are not along one single and narrow beam. Even best engineering cannot avoid that totally. Hence, head on direction of dish to be able to detect is not necessary in RL. Even to other directions, even rear, there is radiation, just less. Even old HARM AGM-88 were able to home on selected patterns of AA Radar. Later versions were able to "remember" the location from where a contact was detected. Pattern to home on were AA radar as well as jamming patterns. So basically almost any kind of radiation from 500MHz to 2GHz. At the moment it seems that such a behavior cannot be simulated in A3 - which is supposed to simulate systems in 2035. AGM-88 is 1980s design, taken into use before 1990 and further developed in later years.

"Unload" works with automatic utilization of parachute if height >=xxx ? That is odd. What if I want to bomb by using fuel trucks dropped from a plane? Isn't it better to redesign the function in a way that the user can select the method (just drop or chute or any other method (rockets)).

if you take the time to read the previous 5-10 Posts you will find out that there are technical (and hence performance) limitations to simulate weapons of 2035 properly. Despite that you want to have high tech weapons? The permanent issue is visial range and radar range. @ POLPOX: They did/do quite good for simple ballistics etc.

If you cannot simulate weapons as they would work in RL, turn back time until you can. Back to 1985 for A4! Or better 1960! Seriously!

A simple statement by BIS could close this: Yes, we confirm the AI status as described in lex's videos - no, we won't fix it in A3.

Ummm, one purpose of a petition IS to find out if someone cares...and then show the result to the addressed party of the petition. Relevant decision makers should be aware about the condition of A3. If not, this company has bigger problems then AI issues.

Ask AngryInsects if you seek for some cool CTI-MP experience. Or maybe OFPS guys. Just few servers up since it is hard to keep the mission logic itself functioning over time.

Ummm, this was mainly about AI. Sorry that I did not made this clear. Of course there were many other bugs, but one could always find a workaround. Main feature that made OFP and ArmA unique in comparision to other competitors was a quite useful AI. Now AI becomes rather a problem. Even the sensors upgrade feeds the AI knowledge of about the environment. I am fine with the units BIS was adding, or better said - I do not really care. My issue is now that during a game I have do remember (and deal) with so many issues and their workarounds that playing is only fun (IMHO) with as few as possible AI.