Wall portals

[Gez]

Hey, it's been a week since the last update, any news?

I tried the latest build on Mediafire, and the mirror freeze is back.



(Also just noticed the date on the title bar.)

Also here's an updated test map. Nothing really new, I just put a couple 3D translucent floors in there to see how they work. I was perverse enough to put one of the path of a polyobject portal, ahah. Mostly I fixed the 3D floor lift in the disco room's polyportal.

[ZZYZX]

Hey, it's been a week since the last update, any news?

The two problems that I'm trying to solve right now are the crash and moving polyobjects.
The crash — I still didn't get it to happen in debug mode, and no idea if it's due to the debug binary and VC++ debugger itself (memory is initialized differently there, AFAIK) or simply because the crash is random.
Moving polyobjects — when you slowly walk out of a portal in your moving 3D floor lift, you get stopped (i.e. player's velocity simply gets nulled). Even if the poly isn't moving at the moment. I know definitely that this isn't caused by FPolyObj::PushMobj nor P_TryMove, and it causes weird bugs with entering a moving portal (basically, you enter your lift, walk out in the same direction that the lift is moving in, your velocity becomes 0 and you get noclipped behind the portal surface due to P_PointOnLineSide saying 1 (behind the portal line), since the portal has already moved and you weren't thrown out of it as it should be).

[SyntherAugustus]

This is really neat. Keep up the good work!

[kevans91]

Hey, it's been a week since the last update, any news?

The two problems that I'm trying to solve right now are the crash and moving polyobjects.
The crash — I still didn't get it to happen in debug mode, and no idea if it's due to the debug binary and VC++ debugger itself (memory is initialized differently there, AFAIK) or simply because the crash is random.
Moving polyobjects — when you slowly walk out of a portal in your moving 3D floor lift, you get stopped (i.e. player's velocity simply gets nulled). Even if the poly isn't moving at the moment. I know definitely that this isn't caused by FPolyObj::PushMobj nor P_TryMove, and it causes weird bugs with entering a moving portal (basically, you enter your lift, walk out in the same direction that the lift is moving in, your velocity becomes 0 and you get noclipped behind the portal surface due to P_PointOnLineSide saying 1 (behind the portal line), since the portal has already moved and you weren't thrown out of it as it should be).

You may be able to reproduce the crash by compiling with an /RTC option (s, u, or 1) to turn off initialization of uninitialized local variables -- that may help.

[Tormentor667]

I really can't wait to see this in the next official release or svn release ... this keeps me really thinking about doing a few new maps for Doom again

[StrikerMan780]

^ I hear you, this is pretty fantastic stuff, and opens up a whole new plethora of possibilities.

[Major Cooke]

I'm actually somewhat interested in contributing to this project, but the only problem is, I'm not that great of a coder.

[ZZYZX]

I found a nice video about OpenGL mirror back side clipping (hey GZ, I heard you were telling people to use GZDoom clipping code in software, right? )
https://www.youtube.com/watch?v=ioL7Zi9CRwg

Meanwhile the portal code is getting fixed... Whoever is curious about the progress, see the "current status" part in the first post.

also wat

the only problem is, I'm not that great of a coder

[Major Cooke]

Contributing code, you know? I would love to help with the coding if I wasn't so terrible with it. So if anything I figured I'd help test the daylights out of it instead and report anything broken like what gez has been doing.

[ZZYZX]

Contributing code, you know? I would love to help with the coding if I wasn't so terrible with it. So if anything I figured I'd help test the daylights out of it instead and report anything broken like what gez has been doing.

I understood that part, I just didn't understand what's so bad about the code that you posted

Right now, I'm not sure if anything could be done (since the old code is abandoned and it's not yet ported [where possible] into the new branch), but after that I'm looking for people who could write or at least help writing sound repeating code (i.e. to make same sound appear from multiple coordinate sets) and monster sight code. That's totally not done yet, not even in the original branch.

[Major Cooke]

Graf pretty much redid my Pick submission because it was a memory leak. And you don't need memory leaks from me. >.>

Well, I could try and take a look at replicating the thing you need for playing sounds again through portals. Just don't hold your breath.

[Gez]

A few suggestion:

• I'd take line id from args[4] rather than args[1], so that it's kind of out of the way in UDMF.
• Don't forget that Eternity has its own separate xlat file
• Update the thread more often, just so we know you haven't been shelled when there's no new update for over a month...

[ZZYZX]

I'd take line id from args[4] rather than args[1], so that it's kind of out of the way in UDMF.

it's already that way (although I'll probably move the identifier position to the last arg once I'm sure which exactly arg is the last)

Don't forget that Eternity has its own separate xlat file

Don't forget that eternity.txt is not used by default in any way, and this particular special doesn't conflict with ZDoom's normal functioning so IMO it's okay to put it in the global file.

Update the thread more often, just so we know you haven't been shelled when there's no new update for over a month...

You know, 2 days ago I wasn't sure myself if I'm shelled The amount of critical esoteric bugs was too high (2, and yes, that's high).

[Gez]

Don't forget that eternity.txt is not used by default in any way

I know, but I have some Eternity mods where I've put a ZMAPINFO lump so that they'll load the Eternity translator. It'd be nice to suddenly find out that OPF's skybox works, for example.

There is no drawback to updating the eternity xlat, is what I'm saying.

[Major Cooke]

Anything new on this, by the way?