Windows v6 beta test

[Tigairius]

Made some large changes to the playerlist on Login 2.

Now profile windows will open at the center of your screen, likewise for PM windows and history.

Your Playerlist will save position now too if you have alwayson checked, so if you open it and move it to a certain spot on your screen, then when you log back into Graal, it'll open in that location. If anyone has any issues, please let me know.

Note: These features are only available on Login 2.

[Mark Sir Link]

On testbed, Personal/Mark_Sir_Link/GUITEST

On v5, using this style and changing the font in the profile is working just fine.

http://img834.imageshack.us/img834/1...1295133596.png

On v6, it is replacing the font with something else that doesn't appear to be in the window's blind skin nor is it my default font.

http://img141.imageshack.us/img141/5...1295133197.png

Happening to Jacob_Bald6225 also.

overridestylefont appears to work but isn't allowing the font color to be changed

[Admins]

It's using the font color defined in the wba file (Personality, ActiveTextR/G/B). You can set the profile.overridestylefont to override the font type, or set style = "" or "none" or so to get rid of the wba stuff in a sub-control.

Hmmm the window position things aren't working at all for me on login2, just seeing some windows shifting positions and the playerlist always appear at {0,0}

[xXziroXx]

PHP Code:

File download: guigames_arcmessage2.png (size: 10128) done
File download: guiblue_scroll_noback.png (size: 2543) done
File download: loginserver_en.po not found
File download: login_gradient_blue.png (size: 309) done
File download: graalv4title.png (size: 32000)...
File download: graalv4title.png (size: 64000)...
File download: graalv4title.png (size: 83319)...
File download: graalv4title.png (size: 83319) doneFile download: guigames_arcmessage2.png (size: 10128) done
File download: verdanabd.ttf not found
File download: graalicon_big.png (size: 322) done
File download: microsoft sans serif.ttf not found
File download: guiblue_scroll_noback.png (size: 2543) done 


I get this every single time I open v6, it's rather annoying to download the same files over and over.

Also, and this is a bug that's been around on v5 for ages as well, when all the downloads are done it doesn't close down the window, and forces you to do it on your own. Which is annoying to say the least.

[xXziroXx]

Also, if you press Alt (without any modifier) in the main Graal window or an external one, the game freezes completely.

[Skyld]

Originally Posted by xXziroXx

Also, if you press Alt (without any modifier) in the main Graal window or an external one, the game freezes completely.

I think that is because it is technically selecting the "window menu" (I am guessing if you press Alt and then when it is frozen press the down key, you will see the window menu appear).

[xXziroXx]

Originally Posted by Skyld

I think that is because it is technically selecting the "window menu" (I am guessing if you press Alt and then when it is frozen press the down key, you will see the window menu appear).

Yeah, I know that much, was just hoping for it to be disabled somehow since the window menu doesn't really do much for Graal.

[Skyld]

Originally Posted by xXziroXx

Yeah, I know that much, was just hoping for it to be disabled somehow since the window menu doesn't really do much for Graal.

This is probably more a Windows restriction I am guessing.

[TSAdmin]

Originally Posted by Skyld

This is probably more a Windows restriction I am guessing.

I'm pretty sure it is. If there are no in-window menus available, it treats the ALT key as though you clicked the program icon in the top left of the title bar, which brings up the default "Move, Close, Resize," etc menu. I don't think programs themselves can disable that function. Google Chrome completely removed the title text and icon at the top of their window, yet you can still click that top left corner and achieve the same results.

[-Ramirez-]

Originally Posted by Skyld

This is probably more a Windows restriction I am guessing.

It's a restriction of the rendering loop being in the same thread as the Windows message loop. I highly doubt Stefan is going to go through the potential multi-threading headaches to resolve this.

[Crow]

Originally Posted by oo_jazz_oo

Just discovered something.

Now, I don't know much about particle emitters and stuff, but I found a problem with the snowfall on UN.

Basically, the script has the particle image changing every .5 seconds, so it changed the image for each new snowflake.
On v5, changing the image of the particle would only change it for new particles...

On v6, it is changing the image for every particle, ones already created, and new ones.

Not sure if this is a glitch, or if it was a glitch in v5.
The script changing the image of said particle is:

PHP Code:

findimg(200).emitter.particle.image = temp.newimg; 


Originally Posted by Stefan

It's an optimization, if you have only one particle template and only simple modifiers then it will try to draw all particles at once instead of drawing it one by one. You could eventually use several particle templates (emitter.particletypes = 2, emitter.particles[1].image = ...).

I don't think that using several particle types is a good solution to the problem your optimization creates. Consider an emitter for controlled text particles, each particle supposed to have a different text. How would you do that? Doesn't work quite as easy anymore with v6, does it?

[fowlplay4]

Originally Posted by Crow

I don't think that using several particle types is a good solution to the problem your optimization creates. Consider an emitter for controlled text particles, each particle supposed to have a different text. How would you do that? Doesn't work quite as easy anymore with v6, does it?

The way I see the particle system (or how I use it), it's for situations where you don't want to devote a loop to display an effect. I.e: Sparkle-trail attached to player.

For what jazz is doing on UN he should only have to declare X~ types of particles and assign each particletype/snowflake a different image. I.e:

PHP Code:

emitter.particletypes = 4;
for (temp.i = 0; temp.i < emitter.particletypes; temp.i++) {
  emitter.particles[temp.i].image = "snowflake_" @ temp.i @ ".png";
} 


No need to constant loop and change the images every .5 second.

Could also just set continueafterdestroy to true, and destroy (hideimg) the particle image then re-create it with new values and none would be the wiser unless there's noticeable performance loss.

[Crow]

Well, currently, I have an emitter with a single particle type which emits text (which is always different) as a particle whenever certain conditions are met. The particle count can jump pretty high, too, so there's lotsa text sometimes. I can't really do that with multiple particle types, since I don't want to create a ton of them. With v5 it works just fine, changing the emitter.particle attributes before emitting the next particle. Using v6, all particles update when changing the attributes

Edit: Code example:

PHP Code:

//#CLIENTSIDE
function onCreated() {
  with (findImg(200)) {
    layer = 6;
    
    with (emitter) {
      nrofparticles = 1;
      emitautomatically = false;
      
      with (particle) {
        style = "br";
        zoom  = 0.9;
        
        font = "Arial";
        textshadow  = true;
        shadowcolor = { 0, 0, 0 };
        
        speed = 0;
        lifetime = 1;
      }
      
      addLocalModifier("range", 0, 1, "y", "add", -48, -48);
      addLocalModifier("range", 0.5, 1, "alpha", "add", -2, -2);
    }
  }
}

// called by something
function onStuff(s) {
  findImg(200).emitter.particle.text = s;
  findImg(200).emitter.emit();
} 


[fowlplay4]

Perhaps a emitter.particle.unique is in order to deal with those conditions.

[Crow]

Originally Posted by fowlplay4

Perhaps a emitter.particle.unique is in order to deal with those conditions.

Sounds good. Stefan, would you kindly implement this, please? :3

[Tigairius]

Stefan did that on purpose for efficiency. On kingdoms iphone we have to create a new object each time we want new text basically. After about 30 objects or so, modulus back to 0.

[Crow]

Originally Posted by Tigairius

Stefan did that on purpose for efficiency. On kingdoms iphone we have to create a new object each time we want new text basically. After about 30 objects or so, modulus back to 0.

I know he did. The behavior should be optional, though. I also asked him personally about it, and he replied that it shouldn't even do that with text particles or something like that. He also said I should try two particle types and setting both vars the same, for that would eliminate his optimization, but that didn't work either.

[Admins]

Ok next version will have emitter.particle.uniqueparticle = true/false, will be released this weekend or so.

[Crow]

Originally Posted by Stefan

Ok next version will have emitter.particle.uniqueparticle = true/false, will be released this weekend or so.

Thank you <3

[Fulg0reSama]

Originally Posted by Stefan

Ok next version will have emitter.particle.uniqueparticle = true/false, will be released this weekend or so.

I got a friend to tell bout this.

[Geno]

graal runs on my netbook when using V6.

the last version would not run...so

Thanks!

[MrOmega]

Here's an issue I'm having. I'm using GraalControl.onMouseDown() and call a function from within side this. Now within this function I call another and same again in the next function. The params in getDist() are always the params of onMouseDown(), not the ones I send.

PHP Code:

function GraalControl.onMouseDown()
{

  doActions();

}

function doActions()
{

  someFunction();

}

function someFunction()
{

  temp.dist = getDist( player.x, temp.nx);

}

function getDist( temp.x1, temp.x2, temp.y1, temp.y2)
{

  if ( params.size() == 4)
    return (((( temp.x2 - temp.x1) ^ 2) + (( temp.y2 - temp.y1) ^ 2)) ^ 0.5);

  if ( params.size() == 2)
    return ((( temp.x2 - temp.x1) ^ 2) ^ 0.5);

} 


Basically in getDist the params are that of onMouseDown() regardless.

[cbk1994]

Originally Posted by MrOmega

Basically in getDist the params are that of onMouseDown() regardless.

This is intentional. You have to give the parameters names. Stefan explained it somewhere but I can't find the post.

EDIT:

Originally Posted by Stefan

The params[] array is only created when the script is called (doesn't even need to be a function, GS1 style). That is either on an event, or when you call a function of another object. To get a variable number of parameters either use func(arg1, arg2, arg3, arg4 etc) and ignore the arguments that you don't want, or pass the parameters as array.

Currently accessing the parameters for each function as array is not possible because it would mean a big speed slowdown, although if it's really important to have it then we could add something like temp.functionparams[] which dynamically converts the function parameters to variables.

[fowlplay4]

Originally Posted by MrOmega

Here's an issue I'm having. I'm using GraalControl.onMouseDown() and call a function from within side this. Now within this function I call another and same again in the next function. The params in getDist() are always the params of onMouseDown(), not the ones I send.

Basically in getDist the params are that of onMouseDown() regardless.

Provide an actual working script (especially in Bug Reports) with some output if you want to help get the issue resolved.

Edit: So it's intentional.

PHP Code:

//#CLIENTSIDE

function onCreated() {
  echo("== Created ==");
  getSomevalue(1, 2);
}

function GraalControl.onMouseDown() {
  echo("== Clicked ==");
  echo("MouseDown: " @ params);
  somefunction();
  this.scheduleevent(1, "CheckParams", "");
}

function onCheckParams() {
  echo("check: " @ params);
}

function somefunction() {
  echo("somefunction: " @ params);
  someotherfunction();
}

function someotherfunction() {
  echo("someotherfunction: " @ params);
  temp.getsomething = getSomevalue(1, 2);
}

function getSomevalue() {
  echo("getsomevalue: " @ params);
  echo("getsomevalue params: " @ a SPC b SPC c SPC d);
} 


echos..

PHP Code:

== Created ==
getsomevalue: 
getsomevalue params: 0 10  
== Clicked ==
MouseDown: 0,414,540,1,0
somefunction: 0,414,540,1,0
someotherfunction: 0,414,540,1,0
getsomevalue: 0,414,540,1,0
getsomevalue params: 0 10  
check: "", 


My advice, don't rely on the params array when you can specify parameters in the function. In your case just add a second function.

[cbk1994]

Originally Posted by MrOmega

PHP Code:

  if ( params.size() == 2)
    return ((( temp.x2 - temp.x1) ^ 2) ^ 0.5);

} 


On second look at your code... lol?

[MrOmega]

Originally Posted by cbk1994

On second look at your code... lol?

What? It checks to see if it's getting the dist from two points or just the xdist/ydist. 4 params means its getting it from two actual points...

[cbk1994]

Originally Posted by MrOmega

What? It checks to see if it's getting the dist from two points or just the xdist/ydist. 4 params means its getting it from two actual points...

You don't see anything wrong with:



?

[MrOmega]

Originally Posted by cbk1994

You don't see anything wrong with:



?

Well when you put it in that view I do. But it basically acts as abs(); I could do abs( temp.x2 - temp.x1);

@fowlplay4:
What do you mean add a second function...

Why can't I just recreate the params array when I call a function?

[fowlplay4]

Originally Posted by MrOmega

Well when you put it in that view I do.

@fowlplay4:
What do you mean add a second function...

Why can't I just recreate the params array when I call a function?

You have complete control over what data you pass to it, instead of checking for how many parameters you pass to a certain function, write a function to take exactly how many parameters you need.

I.e:

PHP Code:

function getDist2(temp.x1, temp.x2) {
  // code for just two points here
} 


[WhiteDragon]

Alternatively, write an n-dimensional distance function that accepts an array of tuples.

[MrOmega]

Originally Posted by fowlplay4

You have complete control over what data you pass to it, instead of checking for how many parameters you pass to a certain function, write a function to take exactly how many parameters you need.

I.e:

PHP Code:

function getDist2(temp.x1, temp.x2) {
  // code for just two points here
} 


I kinda did that to save lines, but I get ya. I still have the whole problem of not being able to send the correct param array

[cbk1994]

Originally Posted by MrOmega

I kinda did that to save lines, but I get ya. I still have the whole problem of not being able to send the correct param array

Saving lines is almost never an excuse in programming. Do it the way that makes the most sense.

[Admins]

A new v6 beta version (v5.317) has been released for Windows at http://www.graalonline.com/downloads...al_windows.zip:
- fixed some problems with gani script initialization (onPlayerEnters when re-entering a level)
- fixed update problems of gui graphics with grids
- for avoiding problems with the new particle engine you can specify emitter.particle.uniqueparticle = true; which will disable the optimized rendering and allow different text and images for each particle

[ffcmike]

The issue where on-screen/layer over 3 displays from level NPCs being hidden when the parent npc is off screen which was (seemingly) previously fixed has returned.

A potentially big problem I've noticed with this latest version is that the clientside dir variables of npcs are -sometimes- not setting properly for all npcs within an entire level which is happening within each npc onCreated, I have no idea why though, it's only happening some of the time, I reconnect and enter these levels again and they work fine, can't really find a common condition for it to occur.

Freezing bug for when closing the client is still occuring, any information on such things like specs that could possibly be of interest?

Also may want to remove the ":" from the end of the downlink link .

[Admins]

Anyone experienced problems or has questions about the latest Windows release?

When pressing escape in fullscreen mode it will switch back to window mode (like on Mac). This is scripted so we could change it if it's not wanted.

[ffcmike]

Originally Posted by ffcmike

A potentially big problem I've noticed with this latest version is that the clientside dir variables of npcs are -sometimes- not setting properly for all npcs within an entire level which is happening within each npc onCreated, I have no idea why though, it's only happening some of the time, I reconnect and enter these levels again and they work fine, can't really find a common condition for it to occur.

After further testing I've realised this is happening as a result of rapidly entering a level for the first time and then exiting, which ofcourse is a well known cause of problems, this specific problem doesn't occur in previous versions however, plus I always have the script as "this.dir" so with the correct scope.

Also not related to V6 but I think you may need to take a look at this:
http://forums.graalonline.com/forums...21#post1627521

[MrOmega]

When ever you show an image that hasn't been uploaded to the server it give you the correct error in the log 'File download: foo.png not found'. Then when you do upload the file to server, it doesn't bother to try to download it. Reopening Graal fixes this though.

[fowlplay4]

Originally Posted by MrOmega

When ever you show an image that hasn't been uploaded to the server it give you the correct error in the log 'File download: foo.png not found'. Then when you do upload the file to server, it doesn't bother to try to download it. Reopening Graal fixes this though.

Could probably be fixed with a script like this. My function names are off though but the main idea is there.

PHP Code:

function onLevelFileUpdated() {
  temp.filename = extractfilename(params[0]);
  for (a: allplayers) a.triggerclient(this.name, temp.filename);
}

//#CLIENTSIDE
function onLogMessage() {
  if (logmessage.starts("File download: ")) {
    this.failed.add(filename);
  }
}

function onActionClientSide() {
  if (params[0] == "fileuploaded") {
    if (params[1] in this.failed) {
      downloadfile(params[1]);
      this.failed.remove(params[1]);
    }
  }
} 


or something similar if it's really that big of issue.

[MrOmega]

Originally Posted by fowlplay4

Could probably be fixed with a script like this. My function names are off though but the main idea is there.

PHP Code:

function onLevelFileUpdated() {
  temp.filename = extractfilename(params[0]);
  for (a: allplayers) a.triggerclient(this.name, temp.filename);
}

//#CLIENTSIDE
function onLogMessage() {
  if (logmessage.starts("File download: ")) {
    this.failed.add(filename);
  }
}

function onActionClientSide() {
  if (params[0] == "fileuploaded") {
    if (params[1] in this.failed) {
      downloadfile(params[1]);
      this.failed.remove(params[1]);
    }
  }
} 


or something similar if it's really that big of issue.

That is basically the worked around I'm using, just would like to see the game not give up and never try again. xP

[SlikRick]

I don't know if anybody has reported this, but I don't feel like reading through 59 pages.

No matter what server I am on I am getting lag spurts. I'll be running around, even within the same level, and all of a sudden the client will freeze up for a couple of seconds and it is rather annoying.