Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Frankiek]

Let me explain the situation from a code point of view.

DB3 top gunner pits exchange the fuselage material in the "old fashioned" way, in the "reflectWorldToInstruments" method:

Code: [Select]

  public void reflectWorldToInstruments(float f) {
    HierMesh hierMesh = aircraft().hierMesh();
    Mat mat = hierMesh.material(hierMesh.materialFind("Gloss1D0o"));
    this.mesh.materialReplace("Gloss1D0o", mat);
  }
IL-4 top gunner pit has the modern version of this code.
In that newer code version, the material change happens in a "reflectPlaneMats" method.
But since this method never gets called directly by the base cockpit class, a variable "bNeedSetUp" got introduced, which makes sure that on first update of the cockpit, the material replacement method gets called.
This is the way used for many aircraft, it always works.
To make matters visible, I've added a few log output codes to the IL-4 top gunner pit, see source below:

Code: [Select]

    public void reflectWorldToInstruments(float f)
    {
        if(bNeedSetUp)
        {
            System.out.println("### CockpitIL4_TGunner reflectWorldToInstruments(" + f + "), bNeedSetUp=true");
            prevTime = Time.current() - 1L;
            bNeedSetUp = false;
            System.out.println("### CockpitIL4_TGunner reflectWorldToInstruments(" + f + "), calling reflectPlaneMats(), bNeedSetUp=" + bNeedSetUp);
            reflectPlaneMats();
        }
    }

    protected void reflectPlaneMats()
    {
        HierMesh hiermesh = aircraft().hierMesh();
        Mat mat = hiermesh.material(hiermesh.materialFind("Gloss1D0o"));
        System.out.println("### CockpitIL4_TGunner reflectPlaneMats(), mat=" + (mat==null?"null":""+mat.hashCode()));
        mesh.materialReplace("Gloss1D0o", mat);
    }

    public CockpitIL4_TGunner()
    {
        super("3DO/Cockpit/DB3-TGun/TGunnerIL4.him", "he111_gunner");
        bNeedSetUp = true;
        prevTime = -1L;
        prevA0 = 0.0F;
        hook1 = null;
        System.out.println("### CockpitIL4_TGunner <init>, bNeedSetUp=" + bNeedSetUp);
    }
And this is the result:

Code: [Select]

[2023-02-24 07:36:33.699 UTC +1] dT:    0 ### CockpitIL4_TGunner <init>, bNeedSetUp=true
[2023-02-24 07:36:37.000 UTC +1] dT:   51 ### CockpitIL4_TGunner reflectWorldToInstruments(0.4), bNeedSetUp=true
[2023-02-24 07:36:37.000 UTC +1] dT:    0 ### CockpitIL4_TGunner reflectWorldToInstruments(0.4), calling reflectPlaneMats(), bNeedSetUp=false
[2023-02-24 07:36:37.001 UTC +1] dT:    0 ### CockpitIL4_TGunner reflectPlaneMats(), mat=4734828As expected, all code works fine and the material gets replaced successfully.
No clue why this doesn't work for those who witness the green fuselage, and even more, why it only fails to work on the IL-4 but not on the hundreds of other aircraft that use the same coding scheme.


Mike

I don't want to bother too much but  as this bug is still active in my system out of curiosity I check it out in UP3RC4 (Panos edition) and wonder what? The bug was already there but not in UP2

[SAS~Storebror]

Here is a Full Logs
https://www.mediafire.com/file/da5q18qk8bk9qsn/log.zip/file

Thanks, KT.
I identified a net skin replication error triggered by a user joining the server with an unpatched Ultrapack 3 RC4 version.
In former times this would have caused the player to become invisible and weird things would happen.
Now we get this log entry instead and the base issue is covered.
It's essentially not a bug, but merely a sign of a working mitigation.

FMB, F-84G.
I just noticed the wrong years for F-84G. FMB: 1955 and 1956.
If I'm not mistaken, the correct years are:
1951 - G-1, G-2
1952 - G-3.

Agreed, will fix this in the next update/hotfix.

as this bug is still active in my system

That's the interesting part: How come the bug still appears when the code doesn't leave any room for it?
I'd really be keen to get a class dump of your UP3 instance.
Just run the Selector, enable dump mode, run the game until the issue appears, exit the game, and provide a zip file of your "dump/com" folder.
Thanks.


Mike

[Frankiek]

I finally managed to discover what triggers the bug it is the viewSet=36 in the conf.ini if I change it to 56 or 57 the bug is not showing. You could try to replicate in your system with this setting.

Just in case I am attaching the dump of my Game UP3.4 hotfix17. The only mod I have is WxTech new cockpits for the P-36

https://www.mediafire.com/file/lvk4v3ruxb357uh/dump.7z/file

[SAS~Storebror]

Ah - that makes sense.
So it turns out to be an "auto Autopilot" thing.
Any viewSet setting where "auto Autopilot" is inactive works with modern code style, whereas active "auto Autopilot" triggers the issue.
Thanks, that way we can cover it successfully.


Mike

[Frankiek]

I confirm that is definitely an "auto autopilot " issue. In UP2 the bug is not there while in UP3RC4 when the new version of the IL-4 mod was introduced is present

[SAS~Storebror]

Current Hotfix 18 Status:

P-35 propeller animation not working anymore

Fixed.

Some Buildings are Missing.

Pending.

FMB, F-84G.
I just noticed the wrong years for F-84G. FMB: 1955 and 1956.
If I'm not mistaken, the correct years are:
1951 - G-1, G-2
1952 - G-3.

Fixed.

IL-4 auto autopilot bug: Fixed.

"Static Spawn" Trigger net replication issue: Fixed.

New Message Trigger Option added: Message Triggers can optionally apply only to the player who "triggered" the trigger now, i.e. all players get individual messages from such a message trigger.
This comes in handy e.g. when you want to warn players not to exceed certain altitudes, stay out of certain areas, give hints about waypoints etc.
An "activated by current player only" trigger can optionally be triggered multiple times (default: Just once). There is, however, no choice like "can be trigger <n> times" with "n>1" - either the trigger can be triggered once, or any time repeatedly.

Last but not least, a bug causing trigger messages to be shown off to the left in case of multi-line messages has been fixed, too; alongside with this, the logic of displaying texts from multiple activated triggered has been changed too, they now show up all at once, each with their own centered position and timeout, and each separated by a blank line.


Mike

[Shermy]

Happy Thursday! This Saturday I'm hosting a 100ish person event with Ultrapack but unfortunately I can't seem to increase the number of players for the server beyond 32? Is there anyway around this?

[SAS~Storebror]

I'm afraid that's not feasible.
Client hosted games are limited as they have to power both the client side of the game (for yourself) and the server part (for all others)simultaneously.
You will need a dedicated server for that event.


Mike

[Shermy]

I'm afraid that's not feasible.
Client hosted games are limited as they have to power both the client side of the game (for yourself) and the server part (for all others)simultaneously.
You will need a dedicated server for that event.


Mike

Would you happen to have documentation or links to how we could go about getting a dedi operational?

[SAS~Storebror]

https://www.sas1946.com/main/index.php/topic,69834.0.html

[Shermy]

https://www.sas1946.com/main/index.php/topic,69834.0.html

Thanks Mike. We've got it all sorted out. I know this next question doesn't pertain to bug reporting and I apologize if I'm being annoying but is there anywhere we might be able to advertise our event on these forums? I would enjoy giving the people of SAS a chance to fly some pvpve again. 

[SAS~Storebror]

Sure thing.
You can always post your event in the Lounge: https://www.sas1946.com/main/index.php/board,1.0.html

Alternatively, the "Ultrapack Discussion & Support" board would fit, but I think it won't reach as many people there: https://www.sas1946.com/main/index.php/board,186.0.html

The more the merrier


Mike