What Are You Working On? May 2015 - Developers

[QUOTE=Rocket;47674830]How did you screenshot a screenshot?[/QUOTE] I assume PrintScr, which should still dump it into the clipboard even while ShareX is active. [editline]7th May 2015[/editline] [QUOTE=DoctorSalt;47674740]Just finished an atmospheric scattering ray tracer. [...][/QUOTE] Do you even technically need to raytrace that? I assume as long as you don't have density fields like clouds the complexity should be reduced to something that's analytically solvable.

Actually, ShareX still grabbed it and dumped it to file/put the path on the clipboard. Pretty cool.

[QUOTE=Fourier;47673325]Horizon float what[/QUOTE] The camera may rotate around the view axis (tilt), or choose strange paths when lerping. Human necks do have three degrees of freedom, but we usually use them mostly independently and have different limits on each, so a motion where the horizon stays horizontal because there's no temporary roll feels much more natural (unless there's inertia being balanced out in which case people normally [I]would[/I] tilt their heads a bit). If tilts are used in an unpredictable manner it can also easily lead to nausea. If you interpolate across Euler angles independently and apply them in yaw-pitch-roll order that usually leads to the most natural camera movement. The links FalconKrunch posted demonstrate [I]a form of[/I] horizon float (i.e. the horizon not staying horizontal due to accumulated rotations), but the problem with quaternions is more due to the shortest paths on their unit-hypersphere which are different from the ones on the cylinder surface described by Euler view angles, which has a slightly different effect if I'm not mistaken. That said, this depends entirely on the kind of camera controls you want. If you have an untethered camera / player moving forward through 3D space then a quaternion's free movement will feel more natural than an untweaked angle solution, which would require some juice in the form of roll based on the other motion parameters to be believable.

6 Hours later, and I finally have an axis aligned bounding box which automatically resizes when objects are inserted into my application-wide selection stack. It was a bit of a pain to implement, but since every aspect of it is a physical geometrical object I should have no problem performing ray-intersections to determine which parts are getting selected. [vid]https://dl.dropboxusercontent.com/s/j4hlme8f777umup/2015-05-06_23-21-35.mp4[/vid] Next up, implementing dragging of those points. [B]Edit:[/B] Should a group of selected objects rotate each object individually, or a group as a whole by default? Perhaps I'll add a toggle button to the editor's interface for switching between group and individual transformations while selected in a group.

Added tutorials that show up if you haven't completed the first course at least once. [vid]http://webmup.com/bee8c/vid.webm[/vid]

I just decoded my first RLE-4 bitmap! Turns out I had completely misunderstood the format, and after a small holiday I came back and fixed it immediately v:v:v

I finally fixed my data binding so it actually updates properly now, although I had to do unspeakable things to accomplish that Why couldn't they just make a stock observable dictionary, whyyy

So ... I absolutely loath pathfinding. I've spent more time on my project trying to fix the pathfinding than I have anything else. I've literally spent months on it. I scrapped my entity code and rewrote it trying to improve it. For some reason, whenever a unit is stopped, no matter what conditions I check for, it will stop in two places: When another unit is going the opposite direction. When another unit is stopped. Checking anything gives me the same answer for both cases. Checking to see if the other one should stop also gives me the same answer for both. It's driving me nuts because it's the only thing that's keeping me from continuing. If I can't make good pathfinding, then the game sucks. It's that simple. And it's the most irritating thing on the planet when you write a bunch of code that's supposed to check for the difference and [I]none of it works.[/I] Why? Because when I differentiate between an object that's moving from one that's not, for some reason the object will move into a unit that's standing still. When I check for an object that's standing still instead of moving, neither of them move when they collide. It's blowing my mind.

You can use ShareX hotkeys multiple times so you can just screenshot ShareX itself. Pretty neat though, when showing off ShareX to others over the interwebs.

[QUOTE=Berkin;47674652][img]http://images.sodahead.com/polls/0/0/3/7/6/0/8/9/2/illuminati-conspiracy-theory-new-world-order-111843601566.jpeg[/img][/QUOTE] That reminds me: I've been playing with the idea of building an iluminati detector. It would try to find triangles in images and paste the all seeing eye over them. The idea was created whilest sleep deprived at a hackathon. But the more I think about it the more I feel like it would be a neat project. Should be doable with opencv

Waste of time really [editline]7th May 2015[/editline] Or not, it might prove to be great learning time

Or, you know, fun.

Been working on a simple building game... [IMG]http://img.prntscr.com/img?url=http://i.imgur.com/O82t7DU.png[/IMG] Unity, y u scale my shit without me knowing? :suicide:

Random I know, but I discovered this stuff called "Wireworld". It's automata that can be used to simulate electric circuits [IMG]http://upload.wikimedia.org/wikipedia/commons/5/5d/Animated_display.gif[/IMG]

Was bored, made this: [img]http://puu.sh/hEIa7/3d4c62cb42.png[/img] Visuals to maybe follow.

[QUOTE=ZestyLemons;47677320]Was bored, made this: [img]http://puu.sh/hEIa7/3d4c62cb42.png[/img] Visuals to maybe follow.[/QUOTE] Text based FPS RPG? I'd play it.

[QUOTE=Mega1mpact;47676402]That reminds me: I've been playing with the idea of building an iluminati detector. It would try to find triangles in images and paste the all seeing eye over them. The idea was created whilest sleep deprived at a hackathon. But the more I think about it the more I feel like it would be a neat project. Should be doable with opencv[/QUOTE] I actually suggested this exact thing in a previous WAYWO once. :v: You should absolutely do it.

[QUOTE=Berkin;47677457]I actually suggested this exact thing in a previous WAYWO once. :v: You should absolutely do it.[/QUOTE] Already wrote 90% of the code during my lunch break. It's like 20 lines of python code. It's easier then I expected. I could turn it into an android app or something. OpenCV makes stuff like this easy as fuck. All I'm doing it scanning for shapes and counting the amount of corners. After that some simple math to figure out if all angles are within margins.

[QUOTE=Mega1mpact;47676402]That reminds me: I've been playing with the idea of building an iluminati detector. It would try to find triangles in images and paste the all seeing eye over them. The idea was created whilest sleep deprived at a hackathon. But the more I think about it the more I feel like it would be a neat project. Should be doable with opencv[/QUOTE] I've had this idea as well, but I have no experience with image processing and recognition. Glad to see someone have a go at it! Make sure it comes complete with a slow zoom-in and red rectangle that looks like it was drawn in MsPaint: [t]http://rs1img.memecdn.com/illuminati-illuminati-everywhere_o_2666999.jpg[/t] Also, in the case of multiple triangles, you must pick the least obvious triangle, of course.

[QUOTE=Sidneys1;47677525]I've had this idea as well, but I have no experience with image processing and recognition. Glad to see someone have a go at it! Make sure it comes complete with a slow zoom-in and red rectangle that looks like it was drawn in MsPaint: [t]http://rs1img.memecdn.com/illuminati-illuminati-everywhere_o_2666999.jpg[/t] Also, in the case of multiple triangles, you must pick the least obvious triangle, of course.[/QUOTE] I need to find the source of that macaroni picture for my tests. It'd be perfect.

I love me some macaroni sauce.

waywo proves me wrong again [editline]7th May 2015[/editline] I know this is webdev related, but I've been telling everyone since I'm very excited: I've been selected as a nominee at Awwwards! [url]http://www.awwwards.com/best-websites/andrew-mcwatters/[/url] Sorry to shoehorn this in here, but I'm really happy about it. [t]http://i.imgur.com/E0tqPGW.png[/t]

[IMG]https://dl.dropboxusercontent.com/u/46283290/ifs.png[/IMG] I have no idea what this is supposed to be anymore, but I like it! [editline]FOR LOOPS![/editline] [IMG]https://dl.dropboxusercontent.com/u/46283290/fors.png[/IMG]

[QUOTE=andrewmcwatters;47677670]waywo proves me wrong again [editline]7th May 2015[/editline] I know this is webdev related, but I've been telling everyone since I'm very excited: I've been selected as a nominee at Awwwards! [url]http://www.awwwards.com/best-websites/andrew-mcwatters/[/url] Sorry to shoehorn this in here, but I'm really happy about it. [t]http://i.imgur.com/E0tqPGW.png[/t][/QUOTE] Awwwards sounds like it should be an award for cute things.

I finished my first semester of computer science and now I can do this: [img]http://i.imgur.com/AbyRzSM.gif[/img]

[QUOTE=Snickerdoodle;47678516]I finished my first semester of computer science and now I can do this: [img]http://i.imgur.com/AbyRzSM.gif[/img][/QUOTE] That's more than I could do in my first semester; my teacher didnt have us even touch a keyboard until the last 3 weeks of class.

[QUOTE=Mega1mpact;47677511]Already wrote 90% of the code during my lunch break. It's like 20 lines of python code. It's easier then I expected. I could turn it into an android app or something. OpenCV makes stuff like this easy as fuck. All I'm doing it scanning for shapes and counting the amount of corners. After that some simple math to figure out if all angles are within margins.[/QUOTE] It could be like a camera app, and when you take a picture containing a triangle, the eye slowly appears over it as if the photo is "developing". It's a conspiracy theorist's wet dream.

Converted all my code over to the new FSM and Behaviored system, it's super easy to prototype new features like the menu transitions. [video=youtube;xDwDB1XLF6w]http://www.youtube.com/watch?v=xDwDB1XLF6w[/video]

[QUOTE=Cow Muffins;47678501]Awwwards sounds like it should be an award for cute things.[/QUOTE] Awwwards definitely has a cute name, but the nominees, international jury, and public members are pretty fierce and have a steady eye for good design, so it's really all trial by fire. Hopefully by the end of the week my numbers end up okayish. I'm hoping to keep a solid 7 and would be happy with that, but we'll see. It's definitely not 8 quality anymore and it's certainly not 9, I don't think. The design as aged a bit since I made it last year.

I've been using C# for ages now, and I only JUST stumbled upon SendMessage, sweet baby christ how have I not seen or heard of this before, it's like I found gold. I must be a terrible programmer if I didn't know of this until now. :v: