Probuilder is soooo nice for bsp-like mapping:
[IMG]https://dl.dropboxusercontent.com/u/357850863/ShareX/2015/05/2015-05-24_19-52-30.png[/IMG]
[DEL]Any criticisms apart from Unity and ProBuilder's broken lightmapping?[/DEL]
Added some Rocket Jockey to the hovercraft.
[img]http://www.twinfinite.net/wp-content/uploads/2013/11/Rocket-Jockey-4.jpg[/img]
You are supposed to use the pylons to rotate and turn your rocket. However they just sorta whip you around and whatnot. Ive been trying for hours to get it to work like it should but I just cant get it. I don't want to manually adjust rotations as I would rather keep everything physically simulated and realistic.
[vid]http://a.pomf.se/xafwdp.webm[/vid]
Fixed up my wall edge cases! (:
[img]http://puu.sh/hZn0y/2917d01565.png[/img]
Would look a lot better with proper lighting...
You've certainly gotten a lot better at art.
[QUOTE=MadPro119;47796889]Added some Rocket Jockey to the hovercraft.
[img]http://www.twinfinite.net/wp-content/uploads/2013/11/Rocket-Jockey-4.jpg[/img]
You are supposed to use the pylons to rotate and turn your rocket. However they just sorta whip you around and whatnot. Ive been trying for hours to get it to work like it should but I just cant get it. I don't want to manually adjust rotations as I would rather keep everything physically simulated and realistic.[/QUOTE]
Never even heard of Rocket Jockey but after checking out some videos the game looks really fun. I think your implementation of the mechanic seems to be failing due to what seems like two pivot points, rather than one and a solid arm parented to the body.
[QUOTE=Ott;47796925]You've certainly gotten a lot better at art.[/QUOTE]
Thank you very much, that means a lot! I want to redo the floor tiles, but it's hard to make something that doesn't add a load of visual noise to the scene.
Also, I gotta handle the corner cases for those walls; but I might focus on something else first. (;
[QUOTE=andrewmcwatters;47796941]Never even heard of Rocket Jockey but after checking out some videos the game looks really fun. I think your implementation of the mechanic seems to be failing due to what seems like two pivot points, rather than one and a solid arm parented to the body.[/QUOTE]
Thats a very good point. The rope rotates on the body of the rocket itself which is part of the problem.
The game is a ton of fun although old (1996) and not very successful.
[quote][6:38:57 PM] Andrew McWatters: how large is babel now seb
[6:39:32 PM] Sebastian McKenzie: large meaning
[6:39:43 PM] Sebastian McKenzie: code size, popularity or what
[6:39:52 PM] Andrew McWatters: "man, working on large codebases that you’ve written is so much fun"
[6:39:57 PM] Andrew McWatters: i thought you mean code size
[6:39:58 PM] Sebastian McKenzie: 20k lines
[6:40:11 PM] Andrew McWatters: wow dude that's literally larger than my game engine
[6:40:13 PM] Andrew McWatters: lol
[6:40:18 PM] Sebastian McKenzie: lol
[6:40:18 PM] Sebastian McKenzie: tests are 70k
[6:40:23 PM] Andrew McWatters: damn
[6:40:23 PM] Sebastian McKenzie: jesus
[6:40:29 PM] Medcat: comparing codebase sizes now?
[6:40:33 PM] Andrew McWatters: LOL
[6:41:01 PM] Andrew McWatters: yeah seb has more girth, but i know how to squeeze a lot out of very little line count ;)
[/quote]
(warning music)
[vid]https://a.pomf.se/stzpzq.mp4[/vid]
take him out!
[vid]https://a.pomf.se/bzuggr.webm[/vid]
Well I'm gonna say this is finished. I added a nice smooth colour fade using HSLtoRGB, it takes around 50 hits of the ball to cycle through the entire rainbow. For some reason the scanlines weren't showing on the background, so I added some very faint ones. I removed the wizard numbers like that guy said, and it now works at any resolution. Also just a bunch of optimizing the code and making it look prettier.
I didn't feel like actually coding anything else tonight, so I decided to take another crack at my general "Launcher"
(The footage framerate is higher than it actually appears)
[t]http://i.imgur.com/JsKYah9.gif[/t]
I'm going to start adding some more menus to it and change the layout around a bit more, but I think I'm honing in on the general feel of it.
I'm going to try to implement some rudimentary sliding menu animations as well.
[QUOTE=Zelpa;47798013][vid]https://a.pomf.se/bzuggr.webm[/vid]
Well I'm gonna say this is finished. I added a nice smooth colour fade using HSLtoRGB, it takes around 50 hits of the ball to cycle through the entire rainbow. For some reason the scanlines weren't showing on the background, so I added some very faint ones. I removed the wizard numbers like that guy said, and it now works at any resolution. Also just a bunch of optimizing the code and making it look prettier.[/QUOTE]
Looks like there's some collision issues if you touch the ball on the top or bottom edges of the paddles.
[QUOTE=ZestyLemons;47798298]Looks like there's some collision issues if you touch the ball on the top or bottom edges of the paddles.[/QUOTE]
There is, but you can actually use it to your advantage. The ball keeps colliding with the same paddle over and over, so the speed gets fast REALLY quickly, but the chances of it actually going in the direction you want is really low. I like to think of it as a feature.
This is looking a lot better if I do say so myself.
Oh and music of course because Rocket Jockey's is classic
[vid]http://a.pomf.se/dljnlt.webm[/vid]
There we go, put in a sideways sliding menu. I'd say this is quite an upgrade compared to before:
Before versus After
[t]http://i.imgur.com/CvQfgaB.gif[/t] [t]http://i.imgur.com/yJ2aLLV.gif[/t]
I'd make an exiting animation, but generally when I want a program to close, [I]I want a program to close[/I] and not doddle around.
I'm thinking that tomorrow I'll put an alerts feed in the main menu for resuming an auto-saved session that has improperly terminated, so if for whatever reason any tool crashes, the option will be right there at launcher boot-up for resuming said session
So I had this idea that seems really useful and pretty obvious so it's probably been done before.
A documentation system for APIs as well as for code.
The first use case is for documenting a code base. You would write a "walkthrough" of your code in prose-form in markdown or something similar but the key aspect is that you wouldn't include code samples, but references to files and lines of code in them. This would all get nicely rendered with a side by side docs-code view (similar to this [url]http://coffeescript.org/documentation/docs/lexer.html[/url]).
The main difference to the above link would be the fact that you are not bound to the linear structure of code. Your references can to any part of the code base and it would be the renderers job to present it in an intuitive "minimap" sort of way that's easy to follow.
The API docs follow the same system. You add documentation and reference code, but you would mostly chose to keep the references internal (i.e. not show the code in the rendered API docs) because the user probably doesn't care.
Now, the main benefit that both would share is that the system could automatically mark every part of the documentation that's stale. A gradual warning system that would mark paragraphs as "2 commits out of date" or something similar. You would get a nice on-demand report of docs that need to be looked over. In most cases, this would be a trivial task where you just go through the stale ones and click some button to confirm that they are still valid (upon which the reference would update it's lastCommit to the current commit).
I feel like this could be REALLY really great. Comments are bad. As much as it's suggested to comment your code, it very rarely helps with anything because as long as your components (functions, classes...) are small, understanding what's actually happening in them is probably best done by reading the code.
What is problematic is getting the bigger picture.
A great example of this is contributing to open source projects.
I don't need to understand how specific functions work to implement some new feature. I need to understand "where everything is". Where do I put my code? How do I interact with the rest of the codebase without breaking functionallity of (even worse, haha) convention.
What do you think?
[vid]https://a.pomf.se/hcizxo.webm[/vid]
Pong is so 2014, Breakout is where it's at. Look at that smooth paddle movement, so sexy.
edit: how should I go about making the bricks? Could I do a table like
bricks = {
01, 02, 03, 04
01, 02, 03, 04
}
and have those numbers be the amount of hits it takes to destroy a brick?
[QUOTE=Zelpa;47798651][vid]https://a.pomf.se/hcizxo.webm[/vid]
Pong is so 2014, Breakout is where it's at. Look at that smooth paddle movement, so sexy.
edit: how should I go about making the bricks? Could I do a table like
bricks = {
01, 02, 03, 04
01, 02, 03, 04
}
and have those numbers be the amount of hits it takes to destroy a brick?[/QUOTE]
Nah just declare them one by one.
Brick1 = ...
Brick2 = ...
...
Brick400 = ...
Trust me it's way better
[QUOTE=Zelpa;47798651][vid]https://a.pomf.se/hcizxo.webm[/vid]
Pong is so 2014, Breakout is where it's at. Look at that smooth paddle movement, so sexy.
edit: how should I go about making the bricks? Could I do a table like
bricks = {
01, 02, 03, 04
01, 02, 03, 04
}
and have those numbers be the amount of hits it takes to destroy a brick?[/QUOTE]
Yes you can, 0 being a destroyed brick. That's pretty much the game done right there.
wanring: muzak
[vid]https://a.pomf.se/vyxnmg.webm[/vid]
Cool, I managed to get the table of bricks working and drawing. One problem though, the function that memes them is in Love2d's draw loop, so I don't know how to actually give them collision or anything.
[QUOTE=Darwin226;47798610]So I had this idea that seems really useful and pretty obvious so it's probably been done before.
A documentation system for APIs as well as for code.
The first use case is for documenting a code base. You would write a "walkthrough" of your code in prose-form in markdown or something similar but the key aspect is that you wouldn't include code samples, but references to files and lines of code in them. This would all get nicely rendered with a side by side docs-code view (similar to this [url]http://coffeescript.org/documentation/docs/lexer.html[/url]).
The main difference to the above link would be the fact that you are not bound to the linear structure of code. Your references can to any part of the code base and it would be the renderers job to present it in an intuitive "minimap" sort of way that's easy to follow.
The API docs follow the same system. You add documentation and reference code, but you would mostly chose to keep the references internal (i.e. not show the code in the rendered API docs) because the user probably doesn't care.
Now, the main benefit that both would share is that the system could automatically mark every part of the documentation that's stale. A gradual warning system that would mark paragraphs as "2 commits out of date" or something similar. You would get a nice on-demand report of docs that need to be looked over. In most cases, this would be a trivial task where you just go through the stale ones and click some button to confirm that they are still valid (upon which the reference would update it's lastCommit to the current commit).
I feel like this could be REALLY really great. Comments are bad. As much as it's suggested to comment your code, it very rarely helps with anything because as long as your components (functions, classes...) are small, understanding what's actually happening in them is probably best done by reading the code.
What is problematic is getting the bigger picture.
A great example of this is contributing to open source projects.
I don't need to understand how specific functions work to implement some new feature. I need to understand "where everything is". Where do I put my code? How do I interact with the rest of the codebase without breaking functionallity of (even worse, haha) convention.
What do you think?[/QUOTE]
That's a nice idea.
I'm horrible at documenting stuff when there are multiple files to keep track of and updated. Maths-heavy code is a bitch to keep documented, since you've got the mathematical stuff in one LaTeX document, the commented code in another, and the documentation of the code in a third. And if you've made the mistake of optimizing your code, it'll spaghettify any direct correspondence between the LaTeX and your code. And the code comments are pretty much worthless since you can't write anything useful there, you'll just have to reference the other documents.
Text-only coding sometimes seems a bit archaic, when we certainly could have the technology to embed graphs, images, formulae and links as comments. And the linear way in which code is written isn't optimal either. If I want to see a private function and two public functions that use that private function, I could just bring up those functions and not care about the rest of the class, for example.
But now I'm rambling.
Quick question regarding Git, whats the general consensus on using symbolic links?
Specifically linking a file in one repository to point at a file in a submodule that is used to keep all shared assets
[QUOTE=Richy19;47799307]Quick question regarding Git, whats the general consensus on using symbolic links?
Specifically linking a file in one repository to point at a file in a submodule that is used to keep all shared assets[/QUOTE]
[url=http://stackoverflow.com/questions/954560/what-does-git-do-to-files-that-are-a-symbolic-link]here[/url] is the answer you are looking for :) I was wondering about the same thing a while ago.
Formulating idea in my head:
BASIC Interpreter for the gameboy. GBA has ~65k words of memory. That's more than enough for a BASIC interpreter. If I do it in assembly, with some C
I'm regressing. I told myself I'd learn to do C# and SQL but everytime I pick up one of those books I put it down after a couple minutes. High level stuff just isn't that cool to me.
[QUOTE=Zelpa;47799145]wanring: muzak
- vid snip -
Cool, I managed to get the table of bricks working and drawing. One problem though, the function that memes them is in Love2d's draw loop, so I don't know how to actually give them collision or anything.[/QUOTE]
Make a rotating dong! Eh, i mean.. pong! Each time the ball hits the paddle the game field rotates around the center randomly.
[QUOTE=cartman300;47799709]Make a rotating dong! Eh, i mean.. pong! Each time the ball hits the paddle the game field rotates around the center randomly.[/QUOTE]
are you trying to KILL us?
[QUOTE=cartman300;47799709]rotating dong[/QUOTE]
I would just post meatspin but I don't want to be banned.
[QUOTE=MadPro119;47796889][...]
[vid]http://a.pomf.se/xafwdp.webm[/vid][/QUOTE]
Shoot out the rope further near the front of the hovercraft.
If you do it at the centre of mass it's clear there will be no torque on it.
You probably want to put it very slightly past the equilibrium point for taking the curves so you fling a bit, but this may even be velocity- and/or likely thrust-dependent.
I could probably calculate it if I had all system parameters but it's likely easier to just experiment :eng101:
[editline]25th May 2015[/editline]
[QUOTE=proboardslol;47799629]Formulating idea in my head:
BASIC Interpreter for the gameboy. GBA has ~65k words of memory. That's more than enough for a BASIC interpreter. If I do it in assembly, with some C
I'm regressing. I told myself I'd learn to do C# and SQL but everytime I pick up one of those books I put it down after a couple minutes. High level stuff just isn't that cool to me.[/QUOTE]
You could make a CLR-to-GBA compiler :v:
It's very easy to find method references in CIL, so unless the program uses heap object (which is relatively easy to avoid even when using some of the more convenient C# features) it should be (relatively) easy to port anything.
[QUOTE=proboardslol;47799629]Formulating idea in my head:
BASIC Interpreter for the gameboy. GBA has ~65k words of memory. That's more than enough for a BASIC interpreter. If I do it in assembly, with some C
I'm regressing. I told myself I'd learn to do C# and SQL but everytime I pick up one of those books I put it down after a couple minutes. High level stuff just isn't that cool to me.[/QUOTE]
Will you write your own interpreter or try to fit in one that somebody wrote?
Sorry, you need to Log In to post a reply to this thread.
