Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

321 Excellent

1 Follower

About sniperton

  • Rank

Profile Information

  • Gender
  • Location

Recent Profile Visitors

1239 profile views
  1. As @=621=Samikatz said, it seems to be a mission/career design issue. Recent QMB or career missions are always saved as _gen.Mission files in the mission folder, so you can always resave them and investigate them in the ME for any oddities experienced. I don't say it's easy, I only say it's possible. At least you can submit them in addition to bugreports.
  2. I think it would be more precise to say that the AI always knows where you are, but is programmed to disregard this info if and when the sight is blocked. If the AI is programmed to avoid situations like that at all cost, it's quite natural what's you're describing as a result.
  3. ^This. Just proceed as normal and on the pre-mission briefing map you'll see the correct waypoints and the mission will play fine. BTW it seems to happen each time when I return to an interrupted career after playing another one. I have never experienced it when playing career missions in sequence.
  4. Tests in the QMB are not reliable in this respect because the AI is already aware of you. Tests with unaware enemy confirm that the AI actually has blind spots which reflect the geometry of the plane type. E.g. a Spit pilot is blind if you come from low 6 or low 9 to 11, but does see you if you come from low 7 to 9. In contrast, a P-39 pilot can't see you if you come from low 7 to 9, but does see you at low 9 to 11. The main problem I guess is that once the AI acquired visual contact with you, he doesn't lose it as easily in maneuvers as we disoriented humans do.
  5. Yes, it's possible, you have to modify the alpha layer in the objects.dds file of the mod you use. (Normally it's a white rectangle in the top row. Fill it with black.) Or ask for it here if you can't do it yourself:
  6. English is my 3rd language, so please forgive me for not having been clear enough. But believe me, access to the game code is really not required here (as long as the game code interprets your mission commands in the same way). You're very good at creating individual missions, Gambit, but a truly "dynamic" campaign rests in Pat's hands. If I were skilled enough in Java, I would have volunteered happily. Anyway, Tailspin's old guide to DCG is worth reading for anyone looking for inspiration.
  7. Not at all. One creates a mission either manually by scripting via the editor (as you do), or by generating the mission script from code (as Patrick does). Once the mission is created, the mission script is interpreted by the game engine which re-creates the mission according to the parameters you have set. Here we have nothing to do with how the game engine ("the actual game code") interprets the script. We can take the source code "as is"; it just reads and executes your commands with some randomness. On the other hand, if you create a mission script from code (as Patrick does), you can use templates and variables instead of static values to create any mission purely by changing the actual value of those variables from code. Now if you change those variables from mission to mission according to previous mission results and a "background strategic plan", you can create a sequence of missions where each mission is built upon the previous one. The code that creates the next mission (I called it "strategic AI") can follow a user-defined plan with some randomness while also factoring in previous mission results. I don't say it would be a milk run, but with PWCG Patrick has already created a programming environment that could be enhanced to that effect.
  8. What @zdog0331outlined here resembles in several ways the background logic and working of Lowengrin's DCG for Il-2 1946. One important difference is that Lowengrin, instead of dice-rolls, could rely on actual combat results of AI units all over the map, because the AI was not that resource-hungry in the old title, and consequently a large portion of the map could be populated with AI units fighting all the time during the mission (planes strafing supply convoys, bombers attacking supply depots, artillery shelling troops concentrations, tanks occupying positions, etc, etc). All losses could be simply read out from the game logs and could be factored in before the next mission was generated. This is not possible in Il-2 GB. Anyway, PWCG already has some dynamic elements, so I'm sure it already has a basic type of strategic AI working in the background. And if PWCG can generate a variety of missions according to some set preferences/parameters, I don't see why it could not dynamically change those parameters over time according to a strategic plan with randomized probability factors, all based on a user-customizable template map as Lowengrin's DCG did. We can only hope that @PatrickAWlsonshows interest in some of these ideas and finds the time to implement a few. For anyone interested, I attach Tailspin's old guide to DCG; it's rather technical, but gives you an idea how the illusion of a dynamically evolving strategic war can be created on mission-level. dcg_campaign_info_v321.zip
  9. Even if you remove the physical detent (it's easily possible), the large deadzone around the 50% position will remain. I used this HOTAS for years, and liked it for its good ergonomy, but @cardboard_killeris right, it's better to get rid of it.
  10. I don't dispute that. My point is that two-engine aircraft are more vulnerable to such attacks than one-engine aircraft, and the Stuka with its wing tanks is more vulnerable than the 109.
  11. I guess the problem here in-game is not so much with the .303cals as with the DM of 109s. I find the .303cals pretty effective against Stukas and Heinkels, but 109s seem to be almost invulnerable other than by shooting at them from a high deflection angle -- "if you're Annie effing Oakley in the sky".
  12. If you're still in Moscow, you can only get .303cals; no Russian equipment so far. With .303cals your lethality will suffer a lot compared to .50cals. You'll have more opportunity to hit, but less chance to kill. Expect situations like "I peppered him, I peppered him, but he flies away as nothing happened!".
  13. I get your point, but do we positively know if HE shells affected flight performance IRL as much as they do in-game? For me the question is not so much if this or that DM is historically correct, but rather if the difference between this or that damage is justified and proportionate. I have no doubts that .50cals are bloody effective in-game. Under ideal circumstances, I can down a G14 with 6 to 23 rounds, depending on luck and whatever, and E7s are much easier targets, 7 to 11 hits bring them down. Here we must also factor in that .50cals are mainly wing-mounted. My personal stats show that I'm 3 times less effective with wing-mounted guns. I shoot more and I hit less, even from convergence. Again I ask if the difference in hit ratios is proportionate and justified historically. These two problems merged together (huge differences between flight penalties and hit ratios) can easily lead to the perception that .50cals are off. I don't say they are off, I just say we have two culprits that might contribute to the perception of many that .50cals are off.
  14. Really interesting. Took the P-40 with 4x .50 cals and fired at a Bf-109-E7 from the convergence ranges of 100 / 200 / 300 / 200 / 100 m, respectivelly. Target was friendly, not maneuvering, I was aiming a bit above the center line, straight into the top of the fuselage. Two pilot injuries out of five test cases, but nonetheless I was shooting until the plane went down either burning or trailing thick black smoke. Rounds expended / hitting / (hit ratio) according to TacView are as follows: 100m: 180 / 9 (5%) 200m: 152 / 3 (2%) 300m: 260 / 7 (2.7%) 200m: 244 / 11 (4,5%) 100m: 100 / 8 (8%) My conclusions, made hastily as always: 1. .50cals AP rounds are indeed very effective if they hit something vital. 2. It's much harder to hit anything vital with wing-mounted guns. They are much less precise than other guns, even if used at the correct convergence range. With wing-mounted guns you have a much lower hit probability. My skills are limited, but under the same test conditions I could score a 10 to 25% hit ratio with the Macchi (in contrast to the 2 to 8% achieved here). How much the difference is realistic, I really can't judge. 3. This thread started with the perception that .50cals don't cause enough aerodynamic penalty. That's the point. As I see now, the problem is not that .50cals cannot kill off targets, but that they cannot help you to get into a situation where you can kill off your target. So back to square zero, aerodynamic penalty caused by .50cals.
  • Create New...