[pmw]

Yeah it's still under construction.

But right away the next problem -- slopes and sprites.
I havent figured out any reason / solution for this:

[Scott_AW]

Start the slope at the low end, it cuts off because the slope starts high and goes down, start low and go up.

[pmw]

I see. But if I make it go up, then sprites will cut from top -- now from bottom.

[Scott_AW]

Well, you could just create a sliver sector next to it to act as the wall, avoiding the sloped sector.

[pmw]

Yeah, I figured out that too. I was just thinking is there somekind of *real solution* for this. But this is enough for me. THanks for advice.

[MetHy]

Simple question : I know you don't need to put the duke3D.grp anymore in your eduke32 directory if you have the game on Steam.

Is that only the case with Steam? Or does eduke32 automatically detect and use ANY duke3D.grp file as long as you have one on your computer somewhere?

Thanks

[Scott_AW]

I believe it needs to be in the same directory, and you can also have it replaced in the Config file with a different or custom made GRP or similar file. I believe Zips might work, but I use KZip, since it was made with Build in mind. My TC has its own GRP, but its actually called OCD.PK4, runs just fine. You may be able to look in the steamapps directory for Duke3d to find the actual GRP file and use it elsewhere.

[MetHy]

What I mean, let's say I have a copy of vanilla duke on my harddrive, or the gog version. Will eduke32 automatically use the grp from that like it does with the Steam version?

[Kyanos]

No. It's gotta be able to locate it, it looks in the standard steam location and the root game directory of eduke32.exe

It will also load the grp if it's in a joined directory for the case of a mod.

[Hendricks266]

EDuke32 only looks in its search path for duke3d.grp. You can see this list of folders near the top of the log. However, on Windows we check the registry to see if the user has Megaton or the GOG Duke installed, and we add those respective folders to the search path.

[Wolf]

I've got a potentially simple question: what exactly controls monster knockback when you shoot at them?

[Hendricks266]

You'd have to find the code in the source that handles it. You can disable it with the spriteflag/htflag SFLAG_NODAMAGEPUSH.

[Jblade]

Would setting spriteflag 128 (SFLAG_NOEVENTCODE) on a bunch of sprites that don't utilize event_game or event_pregame be worth doing for performance gain, or would it the result be neglible? I'll add it if it does, but if not I'll save the time and effort.

[Hendricks266]

It wouldn't hurt, but I doubt it would help since synchronized code only runs at 30 Hz anyway, and code itself isn't that intensive (except rotatesprite, etc), especially since LuaJIT is a JIT.

[Jblade]

Hendricks266, on 21 June 2014 - 02:16 AM, said:

It wouldn't hurt, but I doubt it would help since synchronized code only runs at 30 Hz anyway, and code itself isn't that intensive (except rotatesprite, etc), especially since LuaJIT is a JIT.

Fair enough - you mention rotatesprite, I've noticed having alot of sprites onscreen using it can affect performance a bit. Would using myospal still be useful in some cases, since it doesn't have as many features?

[Hendricks266]

myospal is nothing more than a shortcut for rotatesprite that passes in some default values, so you might be adding a few nanoseconds by using it instead of rotatesprite.

What takes rotatesprite so long is that the command execution isn't finished until the renderer actually finishes drawing the command. Add the fact that drawing events' execution frequency is the refresh rate instead of the game logic's 30 Hz, and rotatesprite calls take a decent portion of the game's processing time. (Scroll down to "Processing".)

[Jblade]

Interesting information, thanks. I take it there's no plan to add a version that is tied to the gamelogic instead of refresh rate? If that's even feasible or would do anything useful at all. Anyways at least I know a bit more about it now, thanks.

[Chip]

It's been a while since I've been here, but today I would like to play a multiplayer game of Eduke32 with an old friend. However, I don't see any multiplayer options. I've played multiplayer with Eduke32 before long, long ago but it appears there's no option for it now. Has multiplayer been scrapped after all these years or is there some other trick to getting it to work?

Also, what's happened to Eduke64? Was it merged with Eduke32 and had its name reverted back to avoid confusion? Or is Eduke64 downloaded from a separate website to the usual Eduke32 one.

[Hendricks266]

James, on 21 June 2014 - 05:13 AM, said:

I take it there's no plan to add a version that is tied to the gamelogic instead of refresh rate?

That doesn't even make sense . If the renderer is going to draw a frame, it needs to know what to rotatesprite, so it calls the screen drawing events to be told what to do. (It's not rotatesprite that is tied to the refresh rate; it's the screen drawing events for which its use is legal.) The hard work of drawing the HUD has to be done for every frame. The CON command just tells it what to do.

If the user wants to limit the rate of drawing to gamelogic, they can use the console command r_maxfps 30.

[LeoD]

Chip, on 21 June 2014 - 06:04 AM, said:

Also, what's happened to Eduke64?

The project's name is EDuke32, therefore the 64bit versions for Windows 64bit are now correctly named eduke32_win64_*.

[Mike Norvak]

Is there any option to change Gamma in mapster? F11 just changes brightness

[Jblade]

Hendricks266, on 21 June 2014 - 06:04 AM, said:

That doesn't even make sense . If the renderer is going to draw a frame, it needs to know what to rotatesprite, so it calls the screen drawing events to be told what to do. (It's not rotatesprite that is tied to the refresh rate; it's the screen drawing events for which its use is legal.) The hard work of drawing the HUD has to be done for every frame. The CON command just tells it what to do.

Yeah sorry I was on the wrong track and misunderstood what you meant. Thanks for the explanation of this stuff though, I do appreciate it.

[Mark]

What simple step am I missing to create an invisible sprite ?

I gave the sprite a cstat of 33041 which includes 32768 for invisibility, 1 for blocking, 16 for wall aligned and 256 for hitscan.

It is invisible and blocking in the game but does not break when hit. I thought that maybe I had to use a breakable tile number so I chose the glass tile for the sprite but still no breaking.

[Daedolon]

Give it a custom hi-tag.

[Mark]

Thanks. I'll give it a try.

YUP. Dat dood it. Its time to celebrate with a pizza and a beer.

[Daedolon]

My plan exactly.

[Mblackwell]

James, on 20 June 2014 - 11:24 PM, said:

Would setting spriteflag 128 (SFLAG_NOEVENTCODE) on a bunch of sprites that don't utilize event_game or event_pregame be worth doing for performance gain, or would it the result be neglible? I'll add it if it does, but if not I'll save the time and effort.

In my testing it's negligible UNLESS you are processing multiple thousands of objects every tic. Then it shaves off a few fps.

[Jblade]

Mblackwell, on 23 June 2014 - 07:20 AM, said:

In my testing it's negligible UNLESS you are processing multiple thousands of objects every tic. Then it shaves off a few fps.

Heh, definitely not doing anything complicated like that. I'll set it for something like rain at least since it can only help, but otherwise I won't waste my time adding it in. Thanks!

[TerminX]

James, on 20 June 2014 - 11:24 PM, said:

Would setting spriteflag 128 (SFLAG_NOEVENTCODE) on a bunch of sprites that don't utilize event_game or event_pregame be worth doing for performance gain, or would it the result be neglible? I'll add it if it does, but if not I'll save the time and effort.

It's worth it for the performance gain. That's why that flag exists, actually. Use it.

[Jblade]

Well I have 2 guys against and 1 guy for but the for guy is the original author of the port so I'll guess I'll side with TX here