[Paul B]

Hi Guys,

When you take a switch button and make it one sided and invisible (i + '), even though it does not have the hitable bit set and the sprite button is NOT set as blockable it still prevents Duke from shooting through the invisible switch from the side the button is facing. However, Duke can successfully shoot through the switch from the opposite direction just fine.

This was tested using EDuke version 4065. I'm not sure how long this has been like this, so I just thought i'd bring it to your attention as it is a very subtle but an annoying problem.

[Micky C]

Pretty sure that's standard behavior.

[Paul B]

How can it be standard behavior when standard behavior never had the option to hide a sprite with (' & i) hidden property. When you hide something it should not be hittable if the hittable flag is off. Otherwise whats the point of having a hittable flag bit. Is that only for walls and not sprites? Seems wrong somehow.

[Kyanos]

I think it's more standard behavior of the switch itself. It's a coded tile that can be switched when shot...

[Paul B]

Even if I'm not using a switch tile that cant be changed when shot? Maybe its just a default for all switches regardless of the switch. Seems a bit odd to me. But I guess it would make sense for those switches that can be triggered by a shot. It would be nice if the hitable tag was applied only to the hittable switches. Thanks for weighing in on this.

Anyway, i've given this more thought and I think I can work around this problem, it's not a show stopper just changes my opinion on the way I thought the sprite hitable tab was to be used. Thanks for the help guys!

[Kyanos]

Now that we're talking about it I noticed that tile 1109 acts odd too. It's a fence with transparency and I can't seem to make it blockable without hitscan. Should be cstat 17 I believe, but it's not letting me shot through. If I toggle the blocking bit then I can shot through it, is that normal too.

[Helixhorned]

You're hitting hardcoded game behavior. When spawning from premap, switches become both blocking and hitscan sensitive. Same goes for every wall or floor aligned sprite with a non-zero hitag, excluding certain tiles.
See game.c:A_Spawn(), searching for "A_CheckSwitchTile". Here's the relevant code:

        if ((CS&48) && PN != SPEAKER && PN != LETTER && PN != DUCK
                && PN != TARGET && PN != TRIPBOMB && PN != VIEWSCREEN && PN != VIEWSCREEN2)
            if (!(PN >= CRACK1 && PN <= CRACK4))
            {
                if (SS == 127)
                    goto SPAWN_END;

                if (A_CheckSwitchTile(i) && (CS&16))
                {
                    if (sprite[i].pal && PN != ACCESSSWITCH && PN != ACCESSSWITCH2)
                    {
// (Remove colored switches in multiplayer...)
                    }

                    CS |= 257;

                    if (sprite[i].pal && PN != ACCESSSWITCH && PN != ACCESSSWITCH2)
                        sprite[i].pal = 0;
                    goto SPAWN_END;
                }

                if (SHT)
                {
                    changespritestat(i, STAT_FALLER);
                    CS |=  257;
                    SH = g_impactDamage;
                    goto SPAWN_END;
                }
            }

        if (CS&1)
            CS |= 256;

The two-letter "convenience macros" expand to some member of sprite[i]:
CS: cstat
PN: picnum
SS: shade
SHT: hitag (edit)
SH: extra
The SPAWN_END label is just before running the EVENT_SPAWN event.

So, as a workaround, you could code a mutator containing only an EVENT_SPAWN definition that reverses setting the hitscan bit for the particular tile of interest.
For the future, it might be interesting to expose A_CheckSwitchTile() to scripting. (More on switches maybe another time.)

[Paul B]

Hi guys,

I didn't want to start a new thread for this, but i've come across a bug. I'm testing my map using the eduke32 -map <mymap.map> -server switch for multi-player purposes. I know multi-player is not functional right now but I use to be able to test my maps using this switch to make sure the multi-player aspect is sound for Duke Megaton players.

Here's the problem: When I start the user map with -server switch the Duke Start Location gets moved from the default start location squishing Duke in non valid player space or just killing Duke on the spot by squishing him.

Any suggestions on why this might be happening?

This problem only surfaced after the release of: 20131228-4230
The build version: 20131226-4223 works just fine.

Thanks for looking into this.

[LeoD]

Paul B, on 10 March 2014 - 05:32 PM, said:

Here's the problem: When I start the user map with -server switch the Duke Start Location gets moved from the default start location squishing Duke in non valid player space or just killing Duke on the spot by squishing him.

This problem only surfaced after the release of: 20131228-4230
The build version: 20131226-4223 works just fine.

That's a known issue, unlikely to be fixed any time soon, I suppose.

[Paul B]

LeoD, on 11 March 2014 - 11:25 AM, said:

That's a known issue, unlikely to be fixed any time soon, I suppose.

Just curious on how you know that? Where are you getting your facts? I hope you're wrong otherwise i'll just have to remain on that version of Eduke as it is critical I can test my Multi-player maps. All maps I make are for the future with co-op and vs in mind. This would be a big pain in the process of testing if I cannot test my user maps for multi-player functionality such as switches or explosions which happen because of sprite pal 1. Plus all user maps I make going forward will be 100% compatible with Megaton since that's where the majority of the user community resides.

[LeoD]

Paul B, on 11 March 2014 - 11:31 AM, said:

Just curious on how you know that? Where are you getting your facts?

"I suppose" needs no facts, just experience of seeing how things are going.

Paul B, on 11 March 2014 - 11:31 AM, said:

All maps I make are for the future with co-op and vs in mind. This would be a big pain in the process of testing if I cannot test my user maps for multi-player functionality such as switches or explosions which happen because of sprite pal 1.

Your workaround for testing could be DNKROZ - DNCLIP to get going in your map.

Paul B, on 11 March 2014 - 11:31 AM, said:

Plus all user maps I make going forward will be 100% compatible with Megaton since that's where the majority of the user community resides.

In that case simply use Megaton for testing?

[Paul B]

Hi guys,

Just noticed something the other day while running Eduke R5267. I think it was occuring also on older recent builds as well. Anyway, the problem is occasionally when using the Console to input commands the only input the console accepts is the up arrow for the last commands entered. Any new commands or typing directly into the console fails.

Any suggestions? It appears to occur randomly and I'm not sure what brings this on. I am most frequently using setrendermode 0, 3, 4 when I notice that I am unable to enter any additional commands in the console until I close EDuke and restart it.

[Hendricks266]

I encountered the same thing recently but I haven't gone back to reproduce it exactly.

[Paul B]

Hendricks266, on 29 June 2015 - 11:46 AM, said:

I encountered the same thing recently but I haven't gone back to reproduce it exactly.

Well i'm glad i'm not the only one. It appears to be a random occurance. I'll try and pay more attention to the events leading up to it.

Also, here is another puzzler. Here are two screenshots from two different computers. One using Onboard Intel Graphics running Polymer (The first image to left) has some missing letters. The second screenshot with Polymer and an Nvidia 660 Graphics card looks fine (Image to the right). Any idea why the same renderer produces different visual results?

Attached thumbnail(s)

• 
• 

[Paul B]

In my latest map I have a sector which is tagged 10331. It use to play the one time sound when Duke entered this sector. Magically it stopped working. I have other sectors in this map with a 1 time sound and they work just fine. Any suggestions? Using "dnclip" to locate that sector directly doesn't work either the sound just fails to play. The right sector is tagged but it doesn't seem to activate any sound.


Ahh bloody hell i just tried it on another computer and it works. It appears my desktop computer doesn't like to play that one time sound but it plays all others. Not sure what's going on, might have something to do with a speaker configuration setting where the audio might be being delivered to another speaker and not to the headphones. When I figure it out i'll update the post.

No idea what happened. The solution was to delete the entire Eduke game folder download the synthesis again and copy a backup Duke.GRP to the new install folder. The stupid thing was it was the ONLY one time sound that didn't work. Doomed Space Marine one time sound worked among custom one time sounds. Very odd and frustrating!
I was also using the HRP previously in my Eduke folder. So i'm not sure if that messed something up at some point.

Attached thumbnail(s)

•