[QUOTE=Berkin;48776301]I am happy to report that I have fixed the bug where "dog anus" did not return a proper dog anus color.
[editline]28th September 2015[/editline]
I also added more modifiers.
[img]http://i.imgur.com/u89HDz9.png[/img][/QUOTE]
[t]https://i.gyazo.com/723c7e7c20c821bbd9c3cedfbd0a5503.png[/t]
Pretty sad.
[QUOTE=dexon;48777030][t]https://i.gyazo.com/723c7e7c20c821bbd9c3cedfbd0a5503.png[/t]
Pretty sad.[/QUOTE]
Can't do gradients, sorry :(
SO working on language interpretation (programming, not natural)
I'm gonna make a simple unstructured BASIC dialect with global-only variables and Goto statements (and a library full of built in functions to make it useful.
my basic idea is to just use a hash table to store everything (variables) and variables are untyped.
But does anyone have anything I can read on parsing selection control and loops?
After a few false starts, I finally found a method that works extremely well for what I wanted out of my engine effect system, and applied the camera shake script I wrote earlier in the week. A little camera shake goes a long way.
For the flame itself, it's now a mesh, using an animated texture ([url=http://i.imgur.com/BCLk7hW.gif]from FS2 Open[/url]), that gets scaled based on engine power.
For the lens flare, I found a really cool workaround to my problems with Unity's lens flares. Lens flare occlusion is really tricky to get right after the PhysX update, so I needed some custom (and more controllable) way to do lens flares. The basis is the classic "draw a quad and make it face the camera" trick. The problem with this method is that you get a lot of ugly clipping with the quad whenever it intersects geometry. The trick to get around this is stupid simple hindsight, but amazingly effective. The quad has preset offset that moves it X units closer to the camera. This allows it to draw over a lot of the geometry that it would normally clip into, and but it'll still be occluded by other ships and other parts of the ship geometry.
When combined with fading the quad when the camera gets too far off-angle, it creates a great looking flare that's visible only when the engine interior is visible, and doesn't clip into any of the ship's own geometry.
Finally, I wrote the emitter part of the camera shake system I started earlier. The engines use camera shake "emitters" that cause the camera to shake based on a shake power value, and a min/max range. It's very simple, but super effective.
[vid]https://dl.dropboxusercontent.com/u/15133164/Capital/Unity%202015-09-28%2001-23-10-92.webm[/vid]
[QUOTE=Berkin;48777034]Can't do gradients, sorry :([/QUOTE]
Just make it loop trough the colors
[QUOTE=deyoppe;48778048]Just make it loop trough the colors[/QUOTE]
I would do that if I knew how, but this was really intended to parody those paint chips with the silly names you see in home improvement stores. Needs a bit more work to look convincing, though.
Real slow progress on my NV:MP project, but I'm trying to work this in with my university work and sorting my internship year out. Got animation playback working, tested it with a clientside actor and it works 90% of the time. Combat animations are still buggy but they do eventually work when Gamebyro wakes up.
[media]https://www.youtube.com/watch?v=7iFe086z-ko[/media]
Other player is me before I timed out. Sorry for not posting any actual MP gameplay too, it all works great but I can only test it when someone else plays with me (thanks to Gamebyro only able to run on a single instance) and only 3 people I know has New Vegas. If anyone wants to help me test shit out, I made a [url=http://nv-mp.com/#nvmp]contact page[/url] or you could just message me, it'd help a fuck load.
I've put decent amounts of work into the server too, virtual worlds and custom gamemode rules are in. I implemented a basic "gamemaster" plugin that gives only admins noclip/administration too.
I changed the paint chip appearance a bit. This look better?
[img]http://i.imgur.com/m4ZzRQA.png[/img]
[QUOTE=Berkin;48778337]I changed the paint chip appearance a bit. This look better?
[img]http://i.imgur.com/m4ZzRQA.png[/img][/QUOTE]
Take a look at what actual paint chips look like
[t]https://cdn.shopify.com/s/files/1/0125/8122/products/paint_chip_placemat_green_1024x1024.jpg?v=1335225650[/t]
So you could add a serif font, and a random ID number for show. Maybe have the ID number actually be usable and have it refer to the color
[editline]28th September 2015[/editline]
And maybe not use a pure black font
Doesn't the hex code kind of do the part of an id number though
My refactor is kinda done so I pushed it to the main repository now
[IMG]http://vgy.me/TApzbY.png[/IMG]
I feel clean
[t]http://i.imgur.com/kNfI0KE.png[/t]
doing a shitty moveset editor for my game
[QUOTE=adnzzzzZ;48779251]My refactor is kinda done so I pushed it to the main repository now
[IMG]http://vgy.me/TApzbY.png[/IMG]
I feel clean[/QUOTE]
What's your process for refactoring Lua? I imagine it's a little less straightforward than with statically typesafe languages.
(For reference: I use C#, so I mostly change something and then change whatever breaks to not be broken anymore.)
[QUOTE=Tamschi;48779524]What's your process for refactoring Lua? I imagine it's a little less straightforward than with statically typesafe languages.
(For reference: I use C#, so I mostly change something and then change whatever breaks to not be broken anymore.)[/QUOTE]
My process isn't that different. I just change things to the way I want them and then unbreak them. I don't think the language changes the process that much. You just run the thing, it will reach some run time error and then you fix that. It hasn't happened too much for there to be a code path that is so rare that I'll refactor an entire system, the code path will never be triggered and then it'll be a hidden bug or something. Usually as I'm testing if everything works the way it did before (by playing the game) everything breaks and then I just fix it back.
[img]http://i.imgur.com/9ybxURQ.gif[/img]
oh fuck watch out its a shitty ghost
[QUOTE=No_0ne;48780137]
oh fuck watch out its a shitty ghost[/QUOTE]
Wow, your game looks way too spooky for me.
that movement pattern is really weird
feels like they shouldn't float so high or that the distance they move forward should be shorter
[QUOTE=No_0ne;48780137][img]http://i.imgur.com/9ybxURQ.gif[/img]
oh fuck watch out its a shitty ghost[/QUOTE]
Every good Platformer needs rage-inducing Medusa heads!
[QUOTE=No_0ne;48780137][img]http://i.imgur.com/9ybxURQ.gif[/img]
oh fuck watch out its a shitty ghost[/QUOTE]
How do you set their y coordinate? sorta coincidence that my game has ghosts and I just used a sin wave to map their movement
[QUOTE=proboardslol;48780234]How do you set their y coordinate? sorta coincidence that my game has ghosts and I just used a sin wave to map their movement[/QUOTE]
like this
[code]
void init() {
sprite.setActiveAnimation("GHOST");
startY = getPosition().y;
health = maxHealth;
}
void act(sf::Time deltaTime) {
lifeTime += deltaTime;
position.y = startY - (waveAmplitude * sin(2 * PI_F * (lifeTime % wavePeriod) / wavePeriod));
velocity.x = (sprite.getHFlip() ? -1 : 1) * speed;
setPosition(getPosition() + (velocity * deltaTime.asSeconds()));
}
[/code]
init() gets called everytime they spawn
Working on an Android multiplayer game, i got player var sync, but i'm having several problems with spawning positions, unet it's something really new for me
[vid]https://dl.dropboxusercontent.com/s/wkfm4qivbf8xesv/2015-09-28_15-37-00.mp4?dl=0[/vid]
[QUOTE=Why485;48777810]After a few false starts, I finally found a method that works extremely well for what I wanted out of my engine effect system, and applied the camera shake script I wrote earlier in the week. A little camera shake goes a long way.
For the flame itself, it's now a mesh, using an animated texture ([url=http://i.imgur.com/BCLk7hW.gif]from FS2 Open[/url]), that gets scaled based on engine power.
For the lens flare, I found a really cool workaround to my problems with Unity's lens flares. Lens flare occlusion is really tricky to get right after the PhysX update, so I needed some custom (and more controllable) way to do lens flares. The basis is the classic "draw a quad and make it face the camera" trick. The problem with this method is that you get a lot of ugly clipping with the quad whenever it intersects geometry. The trick to get around this is stupid simple hindsight, but amazingly effective. The quad has preset offset that moves it X units closer to the camera. This allows it to draw over a lot of the geometry that it would normally clip into, and but it'll still be occluded by other ships and other parts of the ship geometry.
When combined with fading the quad when the camera gets too far off-angle, it creates a great looking flare that's visible only when the engine interior is visible, and doesn't clip into any of the ship's own geometry.
Finally, I wrote the emitter part of the camera shake system I started earlier. The engines use camera shake "emitters" that cause the camera to shake based on a shake power value, and a min/max range. It's very simple, but super effective.
[vid]https://dl.dropboxusercontent.com/u/15133164/Capital/Unity%202015-09-28%2001-23-10-92.webm[/vid][/QUOTE]
really diggin that camera shake
[QUOTE=adnzzzzZ;48779251]My refactor is kinda done so I pushed it to the main repository now
[IMG]http://vgy.me/TApzbY.png[/IMG]
I feel clean[/QUOTE]
Deleting code is my favorite thing to do as a programmer.
[QUOTE=Darwin226;48781188]Deleting [b]bad[/b] code is my favorite thing to do as a programmer.[/QUOTE]
This so much
[editline]28th September 2015[/editline]
Just finished my "SparseArray" structure.. basically it's 2D Generic Array structure, that allows negative indexes.
I had most problems with shitty modulo and division in C#.
[QUOTE=Fourier;48781325]This so much
[editline]28th September 2015[/editline]
Just finished my "SparseArray" structure.. basically it's 2D Generic Array structure, that allows negative indexes.
I had most problems with shitty modulo and division in C#.[/QUOTE]
Doesn't have to be bad code. Depends on the definition of course, but any code that is not required for an application to fulfill it's purpose can just fuck right off.
did more work on my c++ site generator thingy because for some odd reason i thought this would be a good idea
it can automatically add certain things to all pages in the site among other things. I use this for things like navbars because im both too lazy to copy and paste the navbar on every page and too lazy to learn how to use a templating engine or something
input:
[code]
int main()
{
Site website("Awcmon.github.io");
{//Generate index
Page index("index");
Element hello("h1", "", "hi im awcmon");
Element desc("p", "", "i used to write c++ and make a lot of people angry but i cant anymore so here's a website bye");
Element jumbotron("div", "class = \"jumbotron\"");
Element jumbotroncontainer("div", "class = \"container\"");
jumbotroncontainer.addElement(hello);
jumbotroncontainer.addElement(desc);
jumbotron.addElement(jumbotroncontainer);
index.addElement(jumbotron);
website.addPage(index);
}
{//Generate about page
Page about("about");
Element hello("h1", "", "about this site");
Element desc("p", "", "i wrote this site in c++ rofl");
Element jumbotron("div", "class = \"jumbotron\"");
Element jumbotroncontainer("div", "class = \"container\"");
jumbotroncontainer.addElement(hello);
jumbotroncontainer.addElement(desc);
jumbotron.addElement(jumbotroncontainer);
about.addElement(jumbotron);
website.addPage(about);
}
{//Generate projects page
Page projects("projects");
Element hello("h1", "", "projects");
Element desc("p", "", "some projects");
Element jumbotron("div", "class = \"jumbotron\"");
Element jumbotroncontainer("div", "class = \"container\"");
jumbotroncontainer.addElement(hello);
jumbotroncontainer.addElement(desc);
jumbotron.addElement(jumbotroncontainer);
projects.addElement(jumbotron);
website.addPage(projects);
}
//Generate common elements
SiteManager sitemanager(website);
Element title("title", "", "Hello World!");
sitemanager.addCommonHead(title);
Element bootstrap("link", "rel=\"stylesheet\" href=\"https://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/css/bootstrap.min.css\"");
sitemanager.addCommonHead(bootstrap);
Element maincss("link", "rel=\"stylesheet\" href=\"main.css\"");
sitemanager.addCommonHead(maincss);
Element navlist("ul", "class=\"pull-right\"");
std::vector< std::array<std::string,2> > navbarButtons;
navbarButtons.push_back({ "Main", "index.html" });
navbarButtons.push_back({ "About", "about.html" });
navbarButtons.push_back({ "Projects", "projects.html" });
for (int i = 0; i < (int)navbarButtons.size(); i++)
{
Element item("li");
item.addElement(Element("a", "href=\"" + navbarButtons[i][1] + "\"", navbarButtons[i][0]));
navlist.addElement(item);
}
Element logo("img", "src=\"namcwologo.png\" class=\"pull-left\" alt=\"logo\" width=\"230\" height=\"28\" style=\"PADDING-TOP:5px;PADDING-RIGHT:5px;\"");
Element logoLink("a", "href=\"index.html\"");
logoLink.addElement(logo);
Element navbar("div", "class=\"nav\"");
Element navcontainer("div", "class=\"container\"");
navcontainer.addElement(navlist);
navbar.addElement(navcontainer);
sitemanager.addCommonBodyHead(logoLink);
sitemanager.addCommonBodyHead(navbar);
GenerateHTML(sitemanager.bake());
return 0;
}
[/code]
output:
[url]http://awcmon.github.io/index.html[/url]
i plan to make a gui designer and shit for this so i dont have to write c++ to generate a website anymore
[editline]28th September 2015[/editline]
disclaimer: im a complete noob to webdev entirely
My game thing has depth sorting. It sorts the object list in the loop that my step is in. I'm not sure if I want to do that because of time complexities.
Right now I'm trying to think of a way to detect when objects have crossed in other object in a way that wouldn't raise time complexity to an absurd level.
[QUOTE=false prophet;48784222]My game thing has depth sorting. It sorts the object list in the loop that my step is in. I'm not sure if I want to do that because of time complexities.
Right now I'm trying to think of a way to detect when objects have crossed in other object in a way that wouldn't raise time complexity to an absurd level.[/QUOTE]
You could do it the same way fully 3d games do it, by utilizing a pixel depth buffer:
At the start of each frame, clear a depth buffer to all lowest values. Before drawing a pixel on the screen buffer, first check if the entity's z-value (distance from the camera, screen depth or y position in an isometric game) is higher than the existing pixel's position on the depth buffer.
If it is, then draw that pixel to the screen buffer and then draw the another pixel in the same position to the depth buffer, but instead of setting the pixel to a colour you set it to the z-value.
If it isn't, you skip drawing that pixel entirely.
This technique avoids needing to perform any kind of sorting.
If you want to do the simpler technique and just sort instead (also known as the painter's algorithm), it should work as long as your entities are relatively consistent in their z-value. You can then use a sorting algorithm which abuses the temporal and spacial homogenity of your entities z-value - for example heapsort, or even insertion sort if you can guarantee that your data will almost always be mostly sorted and thus can avoid the O(n²) worst case.
Today I made it possible for lights to turn on/off using shadow maps, as well as the ability to allow each light to either generate their shadow map each frame or pre-generate them whenever the user decides.
With that, I aim to make it possible to save the shadow texture to a file (ultimately into the map file), in a way that it can then be loaded back onto the gpu and used once again.
However I know not a single thing about how images actually work. The shadow maps just use a single color channel, yet I somehow manage to parse the whole thing out like this:
[t]http://i.imgur.com/loqfanw.png[/t]
Progress, nonetheless.
[QUOTE=Karmah;48784694]Today I made it possible for lights to turn on/off using shadow maps, as well as the ability to either generate their shadow map each frame or pre-generate them whenever the user decides.
With that, I aim to make it possible to save the shadow texture to a file (ultimately into the map file), in a way that it can then be loaded back onto the gpu and used once again.
However I know not a single thing about how images actually work. The shadow maps just use a single color channel, yet I somehow manage to parse the whole thing out like this:
[t]http://i.imgur.com/loqfanw.png[/t]
Progress, nonetheless.[/QUOTE]
Even though it may be saved as one channel, your parser could be splitting it up into multiple channels. That would explain the "layers" of color, because as the depth increases, digits carry over and cause a sudden shift in the color you see here.
Be sure to check the color channel format you're parsing into.
Sorry, you need to Log In to post a reply to this thread.