Pre-M v07 feedback and bugs

changelog forthcoming.

short version:

  • paratroopers! (thanks zerg & spiked & floz)
  • sortie queuing (wooo zerg)
  • a series of balance stuff (mostly zerg, a few by me)
  • lots of occlusion baking by spiked + floz

Recon plane disappeared from USSR barracks. Is that intentional?
(note: I had no radar, so maybe it only appears when the radar is built?)

I must have missed it when I moved the planes over to the stockpile system. Fixed in SVN.

At the end of v07 game, when teams were getting killed, that’s what was in the console:

[ 56112] <prereqs>: Counting error, gerradar, nil [ 56112] <prereqs>: Counting error, gervehicleyard, nil [ 56112] <prereqs>: Counting error, gbrvehicleyard, nil [ 56112] <prereqs>: Counting error, gbrtankyard, nil [ 56112] <prereqs>: Counting error, gbrradar, nil [ 56112] <prereqs>: Counting error, gertankyard, nil [ 56112] <prereqs>: Counting error, gbrgunyard, nil [ 56112] <prereqs>: Counting error, rusradar, nil [ 56112] <prereqs>: Counting error, ustankyard, nil [ 56112] <prereqs>: Counting error, usradar, nil [ 56112] <prereqs>: Counting error, rusvehicleyard, nil [ 56112] <prereqs>: Counting error, gergunyard, nil [ 56112] <prereqs>: Counting error, usgunyard, nil [ 56112] <prereqs>: Counting error, rustankyard, nil [ 56112] <prereqs>: Counting error, usvehicleyard, nil
IMO that’s not good. And it repeated for every dying team.

I’m not sure why that’s happening. It’s a error message I put in that appears when the gadget’s count of a number of prerequisite units owned drops below zero. I’ll check the logic again, but if that doesn’t work I might change it to perform a count from scratch on finished/taken/given/destroyed, which will use a bit more performance but will probably be more reliable.

Does it only happen when teams die?

I fixed one bug–it was not checking whether a building was under construction when it was destroyed.

In my experience that only happened end-game, when teams were dying (but it did not happen for every killed team, at least some passed error-free). That was in 2 3v3 games.

The known bug that I fixed is that when a building is destroyed while under construction it erroneously deducts one from the prerequisite count. (You only get credit for a barracks when it finishes construction.) When the count drops below zero the error is triggered. It does make sense that it tends to happen when players die, since that is when they tend to have barracks destroyed while under construction and all their barracks destroyed. I hope this is the case because otherwise I’ll have to rewrite the gadget :stuck_out_tongue:

I’m probably going to turn the pinning level back down a bit; turning it up so high has depowered suppression tools a bit more than I would have liked. Notably, human wave attacks against MG nests are working when they shouldn’t really be.

Also thinking of buffing BARs a bit and removing the bazooka from the US rifle squad, perhaps to be replaced with another BAR/swapping the zooka into the assault squad.

Commissars in USSR squads are also up for debate; they render the traditional anti-USSR spread tactic pretty much infeasible, and that was one of the cool mechanics playing against them (deciding to spread and hold flags against them, or cluster and hold their forces back).

Also I’m thinking about the flanking damage against inf; I’ll test more of this later.

Not exactly pre-M v07 bug, that one was there long before, but it’s still there in v07. PTRD inf, if once forced to go prone, never gets up again and spends the rest of his life crawling. I suspect the same bug affects DP light machinegun inf, never built much of them to be 100% sure.

Its the same problem with MG42, bren, .30cal…all the units that go prone to fire. Longstanding bug, which I don’t much know how to fix - I figured I’d do serious hunting and debugging once we ported the scripts to lua and/or I rewrote them for that purpose.

Substantial work time begins tomorrow night; I’ll work out an appropriate value for the inf prone damage and pinning level, as well as flanking damage, and then hopefully start in on ship handling/balance.

Something feels wrong about T-70. I tried to kill some M8 Scotts with it. However, it did like 10% max HP dmg to M8, and 1 hit from M8 75mm gun did ~80% dmg to the T-70.

Contributing factors:

  • Perhaps I made ARMOR_POWER too high? Journier also mentioned that the effect of armor seemed strong.
  • According to tarrif.net, 20K’s penetration drops off very rapidly with range–between 100m and 1000m it loses more than half of its penetration.
  • Damage is moderately low since I based it on shell weight (although 20K shell weighs a good deal for its caliber).
  • On the other side, at current the armor gadget assigns the M8’s HE shell explosion an effective penetration of ~63mm. Is this too much?

63mm for 75mm HE shell (not HEAT) seems excessive. As a rule, first-gen HEAT rounds (like those used during WW2) had penetration approx. equal to their caliber (so 70-90mm seems reasonable for 75mm round), but those were HEAT, not HE. HE had much less.
Front armor on T-70M (we have the M, not the regular T-70) should be sloped 45mm plate, just like on T-34. That kind of protection shouldn’t be heavily vulnerable to medium-caliber HE (75mm).

Nemo mentioned 105mm HE being used against Tigers–how effective was that? (That is not much thicker than the frontal armor… or is it the roof armor that matters?) Just trying to get a sense of the overall scale for HE vs. armor.

I didn’t know that about the T-70M, I’ll increase the armor accordingly.

That’s second-hand info from spiked, but I know 105mm howitzers and the 105mm armed shermans were used as last-ditch weapons against tigers. The damage there wasn’t so much the penetration as the sheer amount of energy/shaking that the impact created (and the tiger’s rather finicky mechanical systems). Such a hit would almost certainly result in a mobility kill, but probably not that much more (maybe a concussed crew).

Now, how to render that in terms of damage, I can’t really tell you. I’d probably base the damage on the weight of the vehicle, but its hard to say.

You probably realise this, but the Shermans have a slight tendency to drive towards enemies backwards… (well they do so less then forwards, but I still thought it was quite often). This is just from one AI game (CRAIG) I watched (so maybe it has to do with the way the AI gives orders and it works in principle?!).
Hmm, just watched the replay, and it wasn’t just the Shermans actually - Though I thought they had plenty of space to maneuver, the grey AI (bottom right, American) had quite a few “reverse drive” tanks in his first armour push towards the middle flag (about 3-5 minutes in if you speed up the replay to x6) - hmm, I wanted to provide the replay, but there’s already a new version of Spring on #buildserv, so I reckon the replay won’t work anymore…

Oh, and Vehicles being able to drive backwards looks really good in principle! Nice to see you guys on the cutting edge of things :slight_smile:.

Edit: Uploaded the replay after all - even if I reckon you won’t be able to view… : http://replays.adune.nl/?1474

Reverse movement is still a very new feature, not yet tested enough (the fact it isn’t in stable spring doesn’t help much too). Probably reverse speeds would need to be reduced, so the units are less inclined to backpedal all the way. We still have plenty of time till the next engine release, it’s not like we will package a test build to the installer :slight_smile:

Bug: towed 17-pdr gun uvmap is really messed up (in towed mode that is). Build it and see. (note: I used the gmtoolbox unit to produce it).