[Simspelaaja]

As for exiting: this has been a long, long discussion on sites like stackoverflow. The general conclusion is that Android apps should not exit on their own, it goes against Android's philosophy; the OS itself will close an app if it deems it necessary.

Is it really against Android's philosophy? Every (third party) app I have on my phone has an exit button or asks me if I want to exit when I press back button multiple times.

[Rayjingstorm]

icantread49 posted:

You're insanely zoomed out, you can tell by the size of the fluid particles. -snip-
Thanks for the info. Tap 5 times (tap tap tap tap tap quickly), type in "touchpx 20" and tell me if the problem goes away. Some devices are far more sensitive than others when it comes to taps vs. dragging.

That definitely helped immensely, it's completely responsive and usable now.

[ryandaniels]

Darkwater124 posted:

GIF != WebM

Also, WebM is used a lot more often on this forum than GIFs.

Edited:

Also, I'd rather have click-to-play than insta-earrape when scrolling the thread with my mouse on the left side.

Content in this venue is really only effective if it invades the viewers mind; if it simply leaves an invitation, no one is going to be bored or interested enough to bother clicking anything. I know I don't. If there's a pretty picture, I stop and look for a moment, otherwise I move on.

Also, back when we didn't have webm, everyone used gifs. Webm is, effectively, a replacement for gifs .


Garry, maybe you could implement a "fade in" sound effect, where if you mouse over a video, it doesn't immediately start blaring through your speakers, giving you time to mouse out, or at least not be startled.

[ben1066]

Now, is it me or does that look somewhat like ambient occlusion!

[icantread49]

Catdaemon posted:

1) Samsung Nexus S
2) Stock 4.0.3
3) :

Runs pretty well. It starts slowing down the more things you have on screen as you'd expect, but it seems slower than some other physics games I've played.

Just out of curiosity, tap 5 times, type in "drawedges 0" and "drawsky 0". Tell me if it speeds up or not.

The pause/play button was pretty fiddly, might want to make that larger so it's easier to press. Also, why is there a huge rectangle on the right hand side for the menu when android phones have a menu button?

It's actually much easier to drag that out on most devices (especially tablets) depending on how you hold the device. But, yes, I'll add support for the menu button.

Simspelaaja posted:

Is it really against Android's philosophy? Every (third party) app I have on my phone has an exit button or asks me if I want to exit when I press back button multiple times.

Yeah, check this out: http://stackoverflow.com/questions/2...-frowned-upon/ (long thread)

Rayjingstorm posted:

That definitely helped immensely, it's completely responsive and usable now.

I will investigate this and possibly make it the default.

[Hypershadsy]

ben1066 posted:

Now, is it me or does that look somewhat like ambient occlusion!

You've got a freaky plasma effect on the ceiling! Is that what it's supposed to look like?

[VistaPOWA]

1) Nexus S
2) CM9 (ICS 4.0.4)
3) Pretty good, but it has some problems on smaller screens, like touch sensitivity. It selected stuff when I tried to move the camera or move objects. There was a lot of lag when an object got stuck and I tried to force it to move.

1) Samsung Galaxy Tab 10.1
2) Stock Samsung TouchWiz (Honeycomb 3.1)
3) It was really smooth, really easy to interact with. There were some problems with displaying liquids (it cut off like if there was an invisible wall in front of it) at the far right side of the screen. If I scrolled back there (to the right where it got cut off), it would show the liquid again, being static, even if the liquid wasn't there for a long time.

Also, I had a hard time quitting from the app itself, I'd suggest you implement hardware buttons or add a quit option to the menu.

Edited:

If you want me to, I can test it on my ZTE Blade, but I cannot guarantee it in any way it would survive the test.

[icantread49]

VistaPOWA posted:

1) Nexus S
2) CM9 (ICS 4.0.4)
3) Pretty good, but it has some problems on smaller screens, like touch sensitivity. It selected stuff when I tried to move the camera or move objects. There was a lot of lag when an object got stuck and I tried to force it to move.

Interesting, thanks for the feedback.

1) Samsung Galaxy Tab 10.1
2) Stock Samsung TouchWiz (Honeycomb 3.1)
3) It was really smooth, really easy to interact with. There were some problems with displaying liquids (it cut off like if there was an invisible wall in front of it) at the far right side of the screen. If I scrolled back there (to the right where it got cut off), it would show the liquid again, being static, even if the liquid wasn't there for a long time.

I'm going to look into this. If there's any way you can provide a screenshot, that'd be great!

Also, I had a hard time quitting from the app itself, I'd suggest you implement hardware buttons or add a quit option to the menu.

Tapping the home softkey should "exit" the app, no? As for literally quitting the process, see the previous discussion and this link: http://stackoverflow.com/questions/2...-frowned-upon/

[Darkwater124]

ryandaniels posted:

Garry, maybe you could implement a "fade in" sound effect, where if you mouse over a video, it doesn't immediately start blaring through your speakers, giving you time to mouse out, or at least not be startled.

[icantread49]

VistaPOWA posted:

If you want me to, I can test it on my ZTE Blade, but I cannot guarantee it in any way it would survive the test.

Go ahead I'm curious!

[ben1066]

Hypershadsy posted:

You've got a freaky plasma effect on the ceiling! Is that what it's supposed to look like?

That's the result of a shitty random. That's on my "tofix" list.

[icantread49]

Also, I know why your Tab has the issue with the fluid rendering. Thanks for the reminder!

[Tyco RC]

icantread49 posted:

Please reply with:
1) Device information (manufacturer, model, hardware, etc.)
2) OS information (Android version, any custom ROMs, etc.)
3) Result of running Phyzicle (any issues? or does it run well?)
Please be specific!

1) Asus Transformer Prime
2) Stock ICS 4.0.3
3) Runs well I think. Are the physics supposed to be this slow? I feel like everything is in extreme slow-motion but I can't tell if it's like a timestep thing that my device is somehow messing up or if you just intended things to move / fall this gradually. Other than that the performance was fine, but the menus got a little cut off on my tablet:



Also, it's extremely difficult to tell which option at the bottom you have selected sometimes. You might want to make unselected options slightly darker or something.

[icantread49]

Tyco RC posted:

1) Asus Transformer Prime
2) Stock ICS 4.0.3
3) Runs well I think. Are the physics supposed to be this slow? I feel like everything is in extreme slow-motion but I can't tell if it's like a timestep thing that my device is somehow messing up or if you just intended things to move / fall this gradually.

Tap 5 times, type in "dtcoef 1" for real-time. The reason the physics run slightly "slow" by default is for performance on common devices. Higher powered tablets and phones can run real-time physics far better (especially with the fluids).

Other than that the performance was fine, but the menus got a little cut off on my tablet:

Hm, maybe it's just me, but what is that supposed to look like?

Also, it's extremely difficult to tell which option at the bottom you have selected sometimes. You might want to make unselected options slightly darker or something.

Thanks for the report.

[animorten]

Kamshak posted:

[i was told to put this in waywo as there is no lopensteamworks thread:
I have a problem translating a CPP script to lua, it crashes with some stack error

In


Sources
Cpp

CSteamAPILoader loader;
    CreateInterfaceFn factory = loader.Load( );

    ISteamClient008 *steamClient = (ISteamClient008 *)factory( STEAMCLIENT_INTERFACE_VERSION_008, NULL );
    assert( steamClient );

    IClientEngine *clientEngine = (IClientEngine *)factory( CLIENTENGINE_INTERFACE_VERSION, NULL );
    assert( clientEngine );

    HSteamPipe steamPipe = steamClient->CreateSteamPipe( );
    cout << "Pipe: 0x" << hex << steamPipe << endl;

    HSteamUser steamUser = steamClient->ConnectToGlobalUser( steamPipe );

    IClientFriends *clientFriends = (IClientFriends *)clientEngine->GetIClientFriends( steamUser, steamPipe, CLIENTFRIENDS_INTERFACE_VERSION );
    assert( clientFriends );

    CallbackMsg_t callbackMsg;
    std::cout << "Started listening" << endl;
    while ( true )
    {
        if ( Steam_BGetCallback( steamPipe, &callbackMsg ) )
        {
            switch( callbackMsg.m_iCallback )
            {
                case FriendChatMsg_t::k_iCallback:
                {
                    FriendChatMsg_t *friendMessageInfo = (FriendChatMsg_t *)callbackMsg.m_pubParam;

                    //allocate some space
                    char pvData[4096];

                    CSteamID chatter;
                    EChatEntryType msgType;
                    int len = clientFriends->GetChatMessage(friendMessageInfo->m_ulSender, friendMessageInfo->m_iChatID, pvData, 4096, &msgType, &chatter);
                    std::string msg( pvData );

                    if( msgType != k_EChatEntryTypeChatMsg )
                    {
                        Steam_FreeLastCallback( steamPipe );
                        continue;
                    }

                    if( len > 0 )
                    {
                        std::cout << "received chat message from " << clientFriends->GetFriendPersonaName( chatter  ) << " : " << pvData << endl;
                        parser->parse( pvData, chatter );
                    }
                }
                break;
            }
            Steam_FreeLastCallback( steamPipe );
        }
        Session::getInstance( ).update( );
        parser->update( );
        Sleep(10);
    }

Lua

require( "opensteamworks" )

local hSteamPipe = ISteamClient012.CreateSteamPipe( )
local hSteamUser = ISteamClient012.ConnectToGlobalUser( hSteamPipe )
local steamFriends = ISteamClient012.GetISteamFriends( hSteamUser, hSteamPipe, "SteamFriends011" )
local clientFriends = IClientEngine.GetIClientFriends( hSteamUser, hSteamPipe )

while true do
    local b, callbackMsg = Steam_BGetCallback( hSteamPipe )
    if b then
        if callbackMsg.m_iCallback == 805 then
        
            local chatMsg = callbackMsg.m_pubParam
            local friendName = steamFriends:GetFriendPersonaName( chatMsg.m_ulSenderID )
            
            local steamId, msgType, strData, len = clientFriends:GetChatMessage( chatMsg.m_ulSender, chatMsg.m_iChatID )
            print( steamId, msgType, strData, len )
        end
        Steam_FreeLastCallback( hSteamPipe )
    end
end

This happend to me too. Fixed it by adding a virtual function after IClientFriends::SetPersonaName

[VistaPOWA]

App forcecloses on the ZTE Blade, didn't expect much from an ARMv6 device.

[Tyco RC]

icantread49 posted:

Tap 5 times, type in "dtcoef 1" for real-time. The reason the physics run slightly "slow" by default is for performance on common devices. Higher powered tablets and phones can run real-time physics far better (especially with the fluids).


Hm, maybe it's just me, but what is that supposed to look like?


Thanks for the report.

Okay, well I thought it looked odd because I didn't realize that the side menu could be scrolled up and down for some reason so I thought that "Actual Gravity" was being cut off at Actual. Consequently, once I realized that I enabled actual gravity and the simulation behaved much more like I expected it to haha.

I did find something else you might not be aware of however:



For some reason it draws fluids over on the right, that area that is redrawn is relative to the camera too (if I pan too far to the left it disappears, more of it appears when I pan to the right) but it's cut off at what seems to be the boundary of where you're drawing the fully expanded side menu.

It's a hard issue to describe so I hope you get my meaning.

[ShaunOfTheLive]

icantread49 posted:

1) Device information (manufacturer, model, hardware, etc.)
2) OS information (Android version, any custom ROMs, etc.)
3) Result of running Phyzicle (any issues? or does it run well?)
Please be specific!

1) Samsung Galaxy Q, SGH-T589R
2) Android 2.2.2 FROYO.UZKG2 (stock)
3) it shows me a black screen for about five seconds then goes back to app menu lol

[icantread49]

Tyco RC posted:

It's a hard issue to describe so I hope you get my meaning.

Yes I totally understand what's going on, don't worry. I've already fixed it for the next build.

Edited:

shill le 2nd posted:

1) Samsung Galaxy Q, SGH-T589R
2) Android 2.2.2 FROYO.UZKG2 (stock)
3) it shows me a black screen for about five seconds then goes back to app menu lol

Hm ...

[ShaunOfTheLive]

I only have 20 MB internal memory left, would that be a problem?

[Lexic]

The menus look very ungainly on my tablet. They're far too big.
Also, it's not immediately obvious that the side menu has a fundamental function change if you tap on things.
In fact, a lot of features aren't immediately obvious. I hope you have either a tutorial or hint system planned.
I couldn't see any way to:
a) Rotate objects after placing them
b) Rotate oblongs
c) Tell if a shape was static or not just by glancing
Bugs:
If you select the square tool, then drag down with two fingers, it creates a square using normal rules for one finger, and another square from the centre of the simulation for the other finger.
When I zoomed out and panned to the right, the liquid in the centre rendered again

[icantread49]

shill le 2nd posted:

I only have 20 MB internal memory left, would that be a problem?

It's possible.

Lexic posted:

The menus look very ungainly on my tablet. They're far too big.
Also, it's not immediately obvious that the side menu has a fundamental function change if you tap on things.
In fact, a lot of features aren't immediately obvious. I hope you have either a tutorial or hint system planned.

Yeah tap the little book with the question mark.

I couldn't see any way to:
a) Rotate objects after placing them

Typical multi-touch gestures.

b) Rotate oblongs

What do you mean?

c) Tell if a shape was static or not just by glancing

I don't see why this is necessary.

Bugs:
If you select the square tool, then drag down with two fingers, it creates a square using normal rules for one finger, and another square from the centre of the simulation for the other finger.

Thanks!

When I zoomed out and panned to the right, the liquid in the centre rendered again

Hm, I'm a tad bit confused?

[esalaka]

icantread49 posted:

I don't see why this is necessary.

Much more intuitive to use if you just sort of can know if something's static or not.

[icantread49]

esalaka posted:

Much more intuitive to use if you just sort of can know if something's static or not.

Any suggestions on how to mark objects as such?

[ROBO_DONUT]

ben1066 posted:

Now, is it me or does that look somewhat like ambient occlusion!

Sorry to have to say this, but it really doesn't :(
What method are you using?


Going back to page 46, your normals don't look right.
Getting normals from depth should look something like this:

/* back-project depth to eye-space coordinate first */
vec3 coord  = ray * depth;
/* assuming you have eye-space edge rays for the view frustum; */
/* alternative method is to use inverse projection matrix and all that crap */

/* derive normal from partial derivatives of coordinate */
vec3 tangent = dFdx(coord);
vec3 bitangent = dFdy(coord);
vec3 normal = normalize(cross(tangent, bitangent));

Dunno how you write that in HLSL, but the method should be the same.

[Kamshak]

animorten posted:

This happend to me too. Fixed it by adding a virtual function after IClientFriends::SetPersonaName

that works, thanks, new problem arises tho: it crashes in IClientFriends_GetFriendPersonaName with a access violation :/

is there any up-to-date IClientFriends class released?

[esalaka]

icantread49 posted:

Any suggestions on how to mark objects as such?

Different borders or colour schemes? I dunno I'm really terrible at actually planning things like these

[Richy19]

icantread49 posted:

That's odd. Do you know how to get the log, by any chance?

If you know how to use adb, do:

adb logcat -c

To clear the log, then launch the app, wait for it to close, and do:

adb logcat -d > log.txt

Then post/send me that if you can :)

adb isnt findng my device, even tho it is set for dbugging.

[Kopimi]

Richy19 posted:

adb isnt findng my device, even tho it is set for dbugging.

you have your devices usb drivers installed right?

[icantread49]

Richy19 posted:

adb isnt findng my device, even tho it is set for dbugging.

Weird. Well, if you manage to get a log out of it, let me know :)

[Rayjingstorm]

icantread49 posted:

Any suggestions on how to mark objects as such?

Texture them with white noise

[DrLuke]

garry posted:

So what you moaning for then?

I like videos autoplaying. I don't get any issues with sound. I like it the way it is, you like it the way you have it set up. No need to change anything :)

Bandwidth would be an issue whatever I did with HTML5 videos - I'm not going to go out of my way to trick it into not preloading.

Besides the fact that it's only working on chrome, as it plays sound of all videos for 3-4 seconds after opening a page until they're muted, but knowing you, you will just ignore me because you're using chrome and think that it's superior, and therefor everyone should use it.

You're an egoist.

[Darwin226]

Udacity class is teaching me to parse HTML with regular expressions.
Oh dear...

[Averice]

DrLuke posted:

Besides the fact that it's only working on chrome, as it plays sound of all videos for 3-4 seconds after opening a page until they're muted, but knowing you, you will just ignore me because you're using chrome and think that it's superior, and therefor everyone should use it. You're an egoist.

works fine on opera for me..

[ryandaniels]

So I have this function "Flatten"

float2 Flatten(float3 origin, float3 u_axis, float3 v_axis, float3 vector_to_flatten)
{
	vector_to_flatten= vector_to_flatten- origin;

	float u= dot(vector_to_flatten, u_axis);
	float v= dot(vector_to_flatten, v_axis);

	return (float2)(u, v);
}

Which takes a point on a plane (to use as the "origin"), two vectors on the plane (to use as the axis), and finally, a point on the plane which you wish to transform into a 2d uv coordinate.


It works great, however, now I need an function that does the exact opposite; takes a 2d coordinate and turns it into a 3d coordinate.

I'd just work out the math, but I don't know how to solve for a vector inside of a dot product; I know that just given the dot product result and a vector, you don't have enough info to find a unique result.

[ben1066]

ROBO_DONUT posted:

Sorry to have to say this, but it really doesn't :( What method are you using? Going back to page 46, your normals don't look right. Getting normals from depth should look something like this:

 /* back-project depth to eye-space coordinate first */ vec3 coord  = ray * depth; /* assuming you have eye-space edge rays for the view frustum; */ /* alternative method is to use inverse projection matrix and all that crap */  /* derive normal from partial derivatives of coordinate */ vec3 tangent = dFdx(coord); vec3 bitangent = dFdy(coord); vec3 normal = normalize(cross(tangent, bitangent));

Dunno how you write that in HLSL, but the method should be the same.

I noticed, I think I have it now however. I needed to multiply the normals else it ended up looking like crap.

[ShaunOfTheLive]

Hey null:

E/libEGL  ( 8026): call to OpenGL ES API with no current context (logged once per thread)

I/DEBUG   ( 5976): *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***

I/DEBUG   ( 5976): Build fingerprint: 'samsung/SGH-T589R/SGH-T589R/SGH-T589R:2.2.2/FROYO/UZKG2:user/release-keys'

I/DEBUG   ( 5976): pid: 8026, tid: 8032  >>> com.nullular.phyzicle <<<

I/DEBUG   ( 5976): signal 11 (SIGSEGV), fault addr 0000004f

I/DEBUG   ( 5976):  r0 0000004f  r1 524e2a04  r2 5846c363  r3 00000000

I/DEBUG   ( 5976):  r4 8118e28c  r5 0000004f  r6 00000000  r7 00000000

I/DEBUG   ( 5976):  r8 0023d2ac  r9 00000000  10 00000000  fp 00000000

I/DEBUG   ( 5976):  ip afa030c8  sp 524e2a34  lr 8114b9f3  pc 8116ed34  cpsr 60000030

I/DEBUG   ( 5976):  d0  3f8000003f800000  d1  3ff000003f800000

I/DEBUG   ( 5976):  d2  bfd3441350baf6de  d3  c1493000c1323800

I/DEBUG   ( 5976):  d4  4900000040dcb000  d5  000000000006e580

I/DEBUG   ( 5976):  d6  4515b00000000800  d7  408000003f800000

I/DEBUG   ( 5976):  d8  0000000000000000  d9  0000000000000000

I/DEBUG   ( 5976):  d10 0000000000000000  d11 0000000000000000

I/DEBUG   ( 5976):  d12 0000000000000000  d13 0000000000000000

I/DEBUG   ( 5976):  d14 0000000000000000  d15 0000000000000000

I/DEBUG   ( 5976):  scr 60000012

I/DEBUG   ( 5976): 

I/DEBUG   ( 5976):          #00  pc 0006ed34  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976):          #01  pc 0004b9ee  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976):          #02  pc 0002de3e  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976):          #03  pc 000436ea  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976): 

I/DEBUG   ( 5976): code around pc:

I/DEBUG   ( 5976): 8116ed14 eafffff2 e004b500 1613061a d004428b 

I/DEBUG   ( 5976): 8116ed24 78033001 d1f72b00 bd002000 2300b500 

I/DEBUG   ( 5976): 8116ed34 2b0056c3 2300d004 56c23301 d1fb2a00 

I/DEBUG   ( 5976): 8116ed44 bd001c18 4657b5f0 b4c04646 d9022a0f 

I/DEBUG   ( 5976): 8116ed54 42182303 1884d03a 42a01c03 1a25d232 

I/DEBUG   ( 5976): 

I/DEBUG   ( 5976): code around lr:

I/DEBUG   ( 5976): 8114b9d0 6020d007 b0062000 4690bc1c 46a24699 

I/DEBUG   ( 5976): 8114b9e0 4835bdf0 fb90f7e5 28001c05 f023d035 

I/DEBUG   ( 5976): 8114b9f0 af01f99f 1c3a4b31 cb434680 1c28c243 

I/DEBUG   ( 5976): 8114ba00 f0211c39 1c06ef06 d0264580 18281c39 

I/DEBUG   ( 5976): 8114ba10 f0214680 1980ef6c 212e4682 f0234640 

I/DEBUG   ( 5976): 

I/DEBUG   ( 5976): stack:

I/DEBUG   ( 5976):     524e29f4  00000000  

I/DEBUG   ( 5976):     524e29f8  abe08938  /system/lib/libEGL.so

I/DEBUG   ( 5976):     524e29fc  00000000  

I/DEBUG   ( 5976):     524e2a00  002457f0  [heap]

I/DEBUG   ( 5976):     524e2a04  5846c363  

I/DEBUG   ( 5976):     524e2a08  00000000  

I/DEBUG   ( 5976):     524e2a0c  abe087f8  /system/lib/libEGL.so

I/DEBUG   ( 5976):     524e2a10  0023d2ac  [heap]

I/DEBUG   ( 5976):     524e2a14  00000000  

I/DEBUG   ( 5976):     524e2a18  00000000  

I/DEBUG   ( 5976):     524e2a1c  abe02d9f  /system/lib/libEGL.so

I/DEBUG   ( 5976):     524e2a20  abe05c0a  /system/lib/libEGL.so

I/DEBUG   ( 5976):     524e2a24  00000000  

I/DEBUG   ( 5976):     524e2a28  df002777  

I/DEBUG   ( 5976):     524e2a2c  e3a070ad  

I/DEBUG   ( 5976):     524e2a30  811762f4  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976): #00 524e2a34  8114b9f3  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

I/DEBUG   ( 5976): #01 524e2a38  002236c0  [heap]

I/DEBUG   ( 5976):     524e2a3c  80b42727  /system/lib/libdvm.so

I/DEBUG   ( 5976):     524e2a40  0000017c  

I/DEBUG   ( 5976):     524e2a44  002243c8  [heap]

I/DEBUG   ( 5976):     524e2a48  81188980  

I/DEBUG   ( 5976):     524e2a4c  80b42701  /system/lib/libdvm.so

I/DEBUG   ( 5976):     524e2a50  0023d2ac  [heap]

I/DEBUG   ( 5976):     524e2a54  00000000  

I/DEBUG   ( 5976):     524e2a58  00000000  

I/DEBUG   ( 5976):     524e2a5c  8118e28c  

I/DEBUG   ( 5976):     524e2a60  00000001  

I/DEBUG   ( 5976):     524e2a64  00000001  

I/DEBUG   ( 5976):     524e2a68  0023d298  [heap]

I/DEBUG   ( 5976):     524e2a6c  8112de43  /data/data/com.nullular.phyzicle/lib/libphyzicle.so

Edited:

oh maybe my video chip doesn't have OpenGL lmao

EDIT: no wait I'm sure it has ES 2.0, it's a SGX540

[ryandaniels]

ryandaniels posted:

So I have this function "Flatten"

float2 Flatten(float3 origin, float3 u_axis, float3 v_axis, float3 vector_to_flatten)
{
	vector_to_flatten= vector_to_flatten- origin;

	float u= dot(vector_to_flatten, u_axis);
	float v= dot(vector_to_flatten, v_axis);

	return (float2)(u, v);
}

Which takes a point on a plane (to use as the "origin"), two vectors on the plane (to use as the axis), and finally, a point on the plane which you wish to transform into a 2d uv coordinate.


It works great, however, now I need an function that does the exact opposite; takes a 2d coordinate and turns it into a 3d coordinate.

I'd just work out the math, but I don't know how to solve for a vector inside of a dot product; I know that just given the dot product result and a vector, you don't have enough info to find a unique result.

Actually, my brain just kicked into gear and I think I solved it; u and v are scalar projections of a 3d vector onto the u and v axis. So, this is the answer, I believe:

float2 Unflatten(float3 origin, float3 u_axis, float3 v_axis, float2 vector_to_unflatten)
{
	float u= vector_to_unflatten.x;
	float v= vector_to_unflatten.y;

	return (u* u_axis)+ (v* v_axis)+ origin;
}

[Mr. Smartass]

icantread49 posted:

Any suggestions on how to mark objects as such?

Add wireframe / scoring marks to the faces of static objects.

[Natrox]

I can finally create paths;