IL2DCE: A dynamic campaign engine for IL-2 Sturmovik: Cliffs of Dover

[csThor]

Quite honestly the idea of stuffing most things into the template was proven insufficient by DGen years ago. It was not flexible enough to depict changes of aircraft types (i.e. the switch from Bf 109 to Fw 190 within a sub-campaign, not between two of them) or sudden transfers to other locations for reasons you as a player had no control over, a thing which later happened so frequently (i.e. transfer of parts of Fliegerkorps VIII from Kharkov to Orel during Citadel).

I, personally, believe that only certain things should be defined on the template:

- airbase locations (i.e. deactivating air bases which weren't in use at a certain time)
- major railway lines (= primary supply lines)
- major railyards
- if applicable shipping lines (= major supply lines)
- major roads (= secondary supply lines)
- major harbour installations
- minor harbour installations
- bridges (weak spots of supply lines)
[ - Army Group or Army Level Supply Dump location(s) ]

... basically all things that are mostly static or cannot be moved somewhere else. Anything that is mobile or even semi-mobile should be defined in another file. I mean I am a rivet counter when it comes to historical details and I would invest considerable time into adding such details. But not everyone would do the same and I understand that. But if you cram all such things into the template and don't use external files you're automatically limiting the amount of historical details, accuracy and flexibility your tool can generate.

[41Sqn_Banks]

Originally Posted by csThor

Quite honestly the idea of stuffing most things into the template was proven insufficient by DGen years ago. It was not flexible enough to depict changes of aircraft types (i.e. the switch from Bf 109 to Fw 190 within a sub-campaign, not between two of them) or sudden transfers to other locations for reasons you as a player had no control over, a thing which later happened so frequently (i.e. transfer of parts of Fliegerkorps VIII from Kharkov to Orel during Citadel).

I, personally, believe that only certain things should be defined on the template:

- airbase locations (i.e. deactivating air bases which weren't in use at a certain time)
- major railway lines (= primary supply lines)
- major railyards
- if applicable shipping lines (= major supply lines)
- major roads (= secondary supply lines)
- major harbour installations
- minor harbour installations
- bridges (weak spots of supply lines)
[ - Army Group or Army Level Supply Dump location(s) ]

... basically all things that are mostly static or cannot be moved somewhere else. Anything that is mobile or even semi-mobile should be defined in another file. I mean I am a rivet counter when it comes to historical details and I would invest considerable time into adding such details. But not everyone would do the same and I understand that. But if you cram all such things into the template and don't use external files you're automatically limiting the amount of historical details, accuracy and flexibility your tool can generate.

The good thing is that the new engine allows to calculate the waypoints for ground units on the fly, so there is no need to define roads, railroads and bridges. (IIRC destruction of bridges can't be logged at the moment). Airfields are also directly available, of course I can't disable a airfield.

I agree it would be practicable to define the locations of cities, rail yards and harbors for every map (the default definition can be overwritten by every campaign). But this would be a "map template" (location of map characteristics) opposed to a "campaign template" (location of units).

I still have some problems with adding "scripted events", of course I can define that unit x was transferred to airfield y at date z. However, what happens if airfield y is already in enemy territory? We see that "events" won't work in combination with a dynamic frontline. Of course a dynamic frontline is not desired if you want to create a historical campaign. So basically if you want to create a historical campaign you want to define the frontline and location of air units and maybe even for ground units. I don't think you want to define that for every day but let's say for every week. Between these dates the frontline may develop dynamically. But what happens if a unit is destroyed? Will it reappear with the next define situation?

I think it would be very possible to add "reinforcement events" to a dynamic campaign that define which air or ground units become available at a certain time. It may be even possible to define a "preferred location" for them that is used if it is in friendly territory.
It would be cool to add some "special events" where a scripted mission is loaded, e.g. to stage a scripted paratrooper operation or a key assault. But then again this would produced problems with the dynamic front line.

One solution for the front line dilemma would be to link the events to the front situation, i.e. they only happen if the corresponding area is still in own hands.

[csThor]

Originally Posted by 41Sqn_Banks

The good thing is that the new engine allows to calculate the waypoints for ground units on the fly, so there is no need to define roads, railroads and bridges. (IIRC destruction of bridges can't be logged at the moment). Airfields are also directly available, of course I can't disable a airfield.

You misunderstand the purpose, Banks. By defining certain roads and railways I force the engine it use it as primary channels for supply traffic. Even today this vital part of military operations is dependent on good railway connections and good roads. Even with our modern offroad vehicles the need for roads is still there as offroad driving considerably increases wear and tear and therefor ups the need for maintenance. These supply lines are primary targets for the medium bombers and ground-attack aircraft of the two sides. Don't forget that!

Originally Posted by 41Sqn_Banks

I agree it would be practicable to define the locations of cities, rail yards and harbors for every map (the default definition can be overwritten by every campaign). But this would be a "map template" (location of map characteristics) opposed to a "campaign template" (location of units).

Guess why I am such an opponent of a "campaign template". It's not able to depict the constant shift of forces that was so common in later years.

Originally Posted by 41Sqn_Banks

I still have some problems with adding "scripted events", of course I can define that unit x was transferred to airfield y at date z. However, what happens if airfield y is already in enemy territory? We see that "events" won't work in combination with a dynamic frontline. Of course a dynamic frontline is not desired if you want to create a historical campaign. So basically if you want to create a historical campaign you want to define the frontline and location of air units and maybe even for ground units. I don't think you want to define that for every day but let's say for every week. Between these dates the frontline may develop dynamically. But what happens if a unit is destroyed? Will it reappear with the next define situation?

Remember when we discussed the issue over at sturmovik.de? I told you back then that I don't think a totally "dynamic" campaign is realistic. I am of the opinion that a single pilot can't influence the course of the war that much because too many factors are totally outside the scope of a flight sim. I do, however, agree that a transfer list is more sensible in a static environment as the BoB. Some transfers (i.e. from one area to another, from one supreme HQ to another) should be pre-defined (i.e. as part of the concentration of forces for a historical operation) or done as reaction to enemy actions.

[41Sqn_Banks]

Originally Posted by csThor

Remember when we discussed the issue over at sturmovik.de? I told you back then that I don't think a totally "dynamic" campaign is realistic. I am of the opinion that a single pilot can't influence the course of the war that much because too many factors are totally outside the scope of a flight sim. I do, however, agree that a transfer list is more sensible in a static environment as the BoB. Some transfers (i.e. from one area to another, from one supreme HQ to another) should be pre-defined (i.e. as part of the concentration of forces for a historical operation) or done as reaction to enemy actions.

I'm totally with you a dynamic campaign is not realistic. But how to create a semi-dynamic campaign? Which are the dynamic variables, which are static?

If looses are dynamic but front line is static there could be a situation that one side has lost all ground units but still the front line doesn't move. Or a vital air unit can't operate anymore because it lost all aircraft in a unlucky operation. How to have static locations/transfers of air units of the front line moves dynamic?
In addition it's not possible to simulate the whole ground war in detail for performance reasons.

As far as I can see at the moment the only practicable way would be to have a scripted definition of the exact locations of air and ground units for certain dates, e.g. 1 August 1940, 7 August 1940 and 14 August 1940.
Between those dates everything happens dynamic, but when one of the dates is reached the scripted definition is applied and basically resets the dynamic changes to the ones defined for the date.
The only exception could be the squadron of the player where looses and experience of the squadron mates remains throughout the campaign.
This would result in a basically static campaign progress with randomized missions and dynamic development of the own unit.

[csThor]

I took some time to really think about this but quite honestly all I can offer is an opinion and some snippets.

Originally Posted by 41Sqn_Banks

I'm totally with you a dynamic campaign is not realistic. But how to create a semi-dynamic campaign? Which are the dynamic variables, which are static?

Quite honestly I think only the missions can be dynamic, not the setting and the "strategic environment". At first at least I think you should concentrate on making the missions dynamic enough to be believable (use of submissions, triggers, events etc) and logging damage states to static area targets once that is possible (i.e. airfields, bridges, radar stations, railway installations etc). The second step would be to give these things a meaning: damaged railway means impact on supply situation in an area and - if possible - rerouting of supply trains to another line (with all the congestion this overload of traffic causes), damaged airfield means lower maintenance states for your squadron if the hangars are down, etc etc pp ...

Originally Posted by 41Sqn_Banks

If looses are dynamic but front line is static there could be a situation that one side has lost all ground units but still the front line doesn't move. Or a vital air unit can't operate anymore because it lost all aircraft in a unlucky operation. How to have static locations/transfers of air units of the front line moves dynamic?
In addition it's not possible to simulate the whole ground war in detail for performance reasons.

Influencing frontlines is a very difficult matter and I think this shouldn't be attempted too much, if at all, since too many factors and variables are outside the scope of a flight sim. As you said it's not possible to simulate the whole ground war with its ten-thousands of vehicles, guns and tanks so a situation like "one side has lost all its ground forces" just should not happen.

Historically ground warfare was heavily dependant on different sets of factors such as training, doctrine, Command&Control, Communication and whether one side was on the strategic offense or defense. The Red Army of 1941 was numerically powerful but totally unprepared for the Blitzkrieg style of war practiced by the Wehrmacht. The germans won a lot of battles not through firepower but through coordination, movement and training. We can't simulate those advantages in a flight sim ... Which in turn means the impact of a player should be confined to the location he is in and the timetable for the advance/retreat could be shifted only minimally.

Originally Posted by 41Sqn_Banks

As far as I can see at the moment the only practicable way would be to have a scripted definition of the exact locations of air and ground units for certain dates, e.g. 1 August 1940, 7 August 1940 and 14 August 1940.
Between those dates everything happens dynamic, but when one of the dates is reached the scripted definition is applied and basically resets the dynamic changes to the ones defined for the date.
The only exception could be the squadron of the player where looses and experience of the squadron mates remains throughout the campaign.
This would result in a basically static campaign progress with randomized missions and dynamic development of the own unit.

This may be a way. We do, however, have to take into consideration that there were two reasons for unit relocation during the BoB:

1.) A unit may have lost too many aircraft and/or its crew is suffering from fatigue so RAF FC may decide to move it to a secure location to recuperate and bring a fresh squadron to the battle area. Same goes for the LW.

2.) The LW may want to concentrate certain assets in one area for increased impact on operations, just as it happened to the fighter units which were concentrated in the Pas-de-Calais area in August/September 1940.

[41Sqn_Banks]

Dev update. I will host a test server that runs the latest build in a few days. The server will not run 24/7 but for a couple of hours on maybe some days a week.
I have not decided yet if I will include spawn points or instead run a coop script that allows selection of AI aircraft.

[Osprey]

Guys, I have been watching this thread with interest and I'm delighted to see work resume. If I may offer my opinion from a different perspective, one which developers such as yourselves may think is nonsense but this comes from the users point of view, and indeed it may not even be possible, but all the same.

In 1946 i've never flown DCE type stuff and had only done the odd co-op, we always built our own maps versus AI based on history for 're-enactment', immersion and fun only. Competition for me was match based in the USL - my competition days are behind me for now. The point is that I don't care about winning or losing wars, I care about my crew being able to fly together at once in a war like scenario, getting into a fight and returning in 2's and 3's back home, perhaps to prepare for another attack. The whole thing need not run for more than 2 hours. To do that in 1946 I had to have known numbers and spend a long time in the FMB - I don't want this for COD.

My crew is RAF, we are growing a Luftwaffe arm (or trying to).
I would like something which enables players to join the server, warm up their aircraft at our own base and await some sort of instruction, from the game and the flight leader. What the mission needs to do is generate a German bomber formation and a target and let everybody know appropriately. For German flyers they need to be told a rendezvous point, and for RAF they need to be notified of a Luftwaffe buildup over France, where, approximate numbers, height and heading. From that point flight leaders take over, that is perhaps an in game menu available (I've seen code for this) for flight leads only which gives a co-ordinate, height and heading (or vector and height) to the closest bomber formation. The German flyers probably should be told the target/route (maybe this is a briefing option).
The mission can be padded with AI, perhaps by setting a range, when one side has too few humans. We would also want a refly option.

In terms of history there could perhaps be 4 phases (different colour front markers defined on the map?), 1. Kanalkampf, 2. Outer airfield and RDF station attack, 3. Inner airfield attack, 4. Blitz.

I do not mind doing FMB work as long as it is either basic each time or a one off.

Ok, you may think at this stage that I'm talking rubbish but what this does give a squadron is the ability to re-enact the kind of BoB events which happened even though they aren't the actual events and we don't know what is going to happen. The Germans are given escort or the leader can sweep ahead supported by AI fighters (single and twin), the RAF have to scramble and intercept, perhaps with some AI fighters spawned on patrols over the south east.

There is no competition in this at all, I don't care who wins - in fact it's better without competition imho because it's more relaxed and fun that way. Anyway, please tell me what you think, I hope I've provided some info on what quite a few crews would like to see. If you have any questions please ask me.