[QUOTE=hippowombat;52482970]Kind of a weird question - I got a laptop, and I want to be able to collaborate the same project on both my laptop and my desktop since I have to bounce between the two frequently, will it fuck anything up if I'm making changes to the same repo under the same username but from different machines?[/QUOTE]
As long as you push from an up to date copy of the repo, or you change different files from each of the computers, then it will all be fine
[QUOTE=hippowombat;52482970]Kind of a weird question - I got a laptop, and I want to be able to collaborate the same project on both my laptop and my desktop since I have to bounce between the two frequently, will it fuck anything up if I'm making changes to the same repo under the same username but from different machines?[/QUOTE]
I do this all the time. I use SourceTree to manage it.
If for example you continue some work on your laptop but forgot to pull the latest changes from your repo, when you try to push your changes it will let you know that you've fallen behind and you should be able to see the changes and fix it.
Gotcha, thanks for the info folks!
No joke, I've been using Google drive for that. Works really well for all of my projects. I use PyCharm on both.
[editline]18th July 2017[/editline]
But I push major changes to bitbucket
Should probably try to check every little change into git
I've been doing a crapload with flight models and airplane physics lately and but have been neglecting to post it here. It's kind of driven me crazy though because it was all just aimless experimentation. I ended up with [del]3[/del] 4 separate flight models, each of varying degrees of realism and complexity. The most complicated ones used actual wings and required you to build out a plane in a sensible manner, an FLCS loosely based on the F-16's flight control laws, and even had a simple version of transonic drag that you needed an afterburner to break through.
[vid]https://my.mixtape.moe/xuwnzc.webm[/vid]
[img]http://i.imgur.com/erPLLVK.png[/img]
The flight models themselves are pretty cool though. They're all in a place where they would be great starting point for a project to build onto, specialize, and optimize them. However I don't have a game I'm working on, so they're generic and not worth building out any more at this point.
Instead of continuing to go in circles with that, I decided to take a break and just do some gamedev comfort food and make something really simple.
[vid]https://my.mixtape.moe/cfqotl.webm[/vid]
[img]http://i.imgur.com/tJhvaVm.png[/img]
how it all began
[t]https://i.imgur.com/5j9bVlY.png[/t]
what it looks like now
[t]https://i.imgur.com/HzPyW0f.jpg[/t]
damn.
Been off work for a bit to recover from burnout from various things, While I have been (mostly) not doing computers, it is a good opportunity to work on ideas that I really wanted to ( but didn't do because I was too drained when I came back from work )
Currently trying out making a hobby anycast network, been documenting the quirks I've discovered in and out, and making fancy maps that show latency improvements per every set of servers deployed:
[img]http://i.imgur.com/dWAnUaY.gif[/img]
The idea is, anycast is generally hard because of routing policy over the internet is fairly fragile. Small changes can cause your customers to be routed to completely incorrect continents, However it is also very good at dealing with DDoS attacks or network outages because of the use of the internet it self to divide traffic around.
The idea I have basically combines the traditional method ( DNS responses based on your GeoIP ) and combining it with Anycast to get both benefits, with the downside of some added operational complexity.
It seems to actually improve the latency of most requests by the looks of it:
[img]http://i.imgur.com/pBnsB9d.png[/img]
However I still need to figure out if that latency improvement directly means a [I]bandwidth[/I] improvement. Then again, for some use cases you might not care about high bandwidth and really care about low latency.
Hopefully going to wrap up my findings in a blog post in a bit so that the curiosity is not wasted
Just got my [URL="https://smile.amazon.com/Programming-Language-hardcover-4th/dp/0321958322/ref=sr_1_1?ie=UTF8&qid=1500513507&sr=8-1&keywords=c%2B%2B"]The C++ Programming Language by Bjarne Stroustrup[/URL] in the mail.
Side note, does anyone else have issues falling asleep while reading? Pretty much happens immediately for me and I have to really fend it off no matter what material I try to read.
[editline]Edited:[/editline]
I forgot how much you can learn when you spend a few hours doing something interesting.. Haven't been doing any learning outside my job basically since I got it in 2016.
40+ hours of coding for work killed all aspiration that I had before I did it professionally.
Life hit me and I've been pretty busy doing other things, but I managed to sneak in some time to do some engine work.
I want it to be easy for a level designer to make world geometry / "brushes" really quickly.
I made a thing where the user drags 2 points to create an AABB. The 3D cursor intersects with the chosen axis waffle grids, and can snap to the closest intersection from the camera.
If using more than 1 axis, an entire block can be drawn by snaking the mouse from 1 plane to another.
[video]https://youtu.be/HSXbr5L80q4[/video]
It isn't perfect, but the logic isn't too messy at least, with the exception being the face-dragging stuff.
[QUOTE=Karmah;52487801]Life hit me and I've been pretty busy doing other things, but I managed to sneak in some time to do some engine work.
I want it to be easy for a level designer to make world geometry / "brushes" really quickly.
I made a thing where the user drags 2 points to create an AABB. The 3D cursor intersects with the chosen axis waffle grids, and can snap to the closest intersection from the camera.
If using more than 1 axis, an entire block can be drawn by snaking the mouse from 1 plane to another.
[video]https://youtu.be/HSXbr5L80q4[/video]
It isn't perfect, but the logic isn't too messy at least, with the exception being the face-dragging stuff.[/QUOTE]
That's very cool, I wanna look into CSG brushes as well for a thing - I started out making games with purely brush based level editors and it's one of the things that I really missed in Unity when I used it for the first time.
[QUOTE=Proclivitas;52487170]Just got my [URL="https://smile.amazon.com/Programming-Language-hardcover-4th/dp/0321958322/ref=sr_1_1?ie=UTF8&qid=1500513507&sr=8-1&keywords=c%2B%2B"]The C++ Programming Language by Bjarne Stroustrup[/URL] in the mail.
Side note, does anyone else have issues falling asleep while reading? Pretty much happens immediately for me and I have to really fend it off no matter what material I try to read.[/QUOTE]
Yup. Kindle Paperwhite is a surefire way to get me to go to sleep.
I think it's because I always read on the couch.
[QUOTE=Handsome Matt;52487897]
shite video sorry, pulled an all nighter and managed moving the hand around with poor acceleration in Black & White with my irl hand
could probably easily make it completely VR if it wasn't using DX7, I'll have to wrap the calls and translate them to DX9 or something, could be fun to be a literal god.[/QUOTE]
Maybe instead of rolling a custom DX7->DX9 wrapper you could use something like this [url]https://fdossena.com/?p=wined3d/index.frag[/url] which convert the DX calls to opengl and then work from there ?
I finally got bitmap loading working
[img]http://i.imgur.com/ZfOHT5u.png[/img]
took me 2 days of confusion but at least I got 32x32 of pixels out of it
I'm officially ill as fuck again which means I'll be around for a while making random stuff to pass the time until a doctor can figure out what's wrong with me. Luckily this time I'm still able to do programming
I found the 2d platformer I was making to be a bit dry. The actual fun of the game would be from whizzing around at high speed, but in 2d its difficult to make a game thats fun - while something like speedrunners is an option, I didn't really like speedrunners. This made me realise that there's actually very few 2d platformers that I enjoy playing at all, even something like hollow knight leaves me feeling a tad disappointed
But I had a pretty nifty physics thing going on, I actually understand what's going on this time. So, the logical next step was to make an everything simulator!
[video]https://youtu.be/RP3pRaCtWNM[/video]
Solids are not really that... solid at the moment. Its difficult to set up the constraints correctly so that force from the top compresses the structure as a whole slightly rather than shattering the entire thing bottom up. So its more like fluid with some long range elements to it
But its meant to be a game not a proper simulator, so it doesn't matter too much luckily. I was always a big fan of chemistry, but real chemistry is simply absurd (not to mention the computation cost of MO simulation), so the idea is a simplified simulator with specific goals
So while I'm proud of my work on the application proper at work and find it generally satisfying, good god GUI design and programming can die in a hole. I'm using Qt (wish I could just bend ImGui to my will tbh, and still might) which is totally a mixed bag because on one hand there's lots of useful functionality but there's lots of nasty MOC stuff required to use Qt. Not to mention how gross binding the several dozen tweakable parameters present has been, either.
On a better note, I've got two potential job offers for this fall (at my current place of employment). Both full time :saxout:
Nothing solid, yet, but one would have me working on our cognitive (machine learning) radios SBIR since that's mostly C++11/14 work and we've only got one other C++ dev, and the other would be doing more raw C work for our radios in general. I'd much rather get the former, since I've come to love modern C++ and I know the pre-existing codebase for that project is in a great shape since the other C++ dev here is a fucking wizard. I'll take whatever I can get, though. Being able to say "fuck school" entirely after what I've been through seems nice. I wanted to do well in school, and I still enjoyed it for the most part, but [I]goddamn[/I] is my university seemingly determined to screw me over and stop me from taking any sort of major I would want to take. The last upside/note is that I'll be full time for the next round of NASA solicitations. Last time I didn't write any proposals because I was still mixing school and work, but this time I'll definitely be trying to nab some GPU computing related contracts for things like propulsion dynamics and propellant dynamics. Winning one of these would make my primary investigator for them, letting me hire new programmers or requisition current employees to complete my contract. If I do well enough, I could spin up a new division here in a few years :D
I haven't posted in a while, but I have been programming a ton, heeere's a few highlights
Attractors
[t]https://i.imgur.com/VZgFbYm.png[/t][t]https://i.imgur.com/wytWdV0.png[/t]
Brownian motion mountain generation
[t]https://i.imgur.com/vVuGjxR.png[/t]
A fractal program
[t]https://i.imgur.com/tnrxkus.png[/t][t]https://i.imgur.com/u7L6aBP.png[/t][t]https://i.imgur.com/opJUvO9.png[/t]
Some water simulation stuff, still got some kinks to work out
[t]https://i.imgur.com/iDzfY6l.png[/t][t]https://i.imgur.com/ipRLUWO.png[/t]
[QUOTE=Handsome Matt;52491486]Is it open source?[/QUOTE]
I assume Sam means this one, however it's not open source: [url]http://enbdev.com/convertor_dx8_dx9_v0036.htm[/url]
All of yesterday I put off doing any work because I couldn't think of a way to cleanly implement 3D vertex manipulation, but alas today I gave it a go and managed to do it in under 2 hours :v:
My code is still a little bit janky, but this is how I deal with 3D grid snapping:
I define 3 planes at a desired position pointing down all 3 Axes (xyz), and determine if the mouse intersects any.
Generally the closest intersection to the camera is the desired one, however this can be a problem if the plane of the closest intersection point is being viewed at glancing angles.
Imagine trying to drag something across the X axis from -1 to +1, but as soon as it hits the origin (0) it starts dragging down the Z axis towards the camera.
The actual distance of the intersection point to the camera didn't matter, I just needed to know which was the closest but at a reasonable angle, so I modified the reported distance:
Taking 1 minus the absolute dot product of the plane normal against the viewing direction gives me a scaling factor where the value gets smaller when we are looking at a plane more head on, otherwise it approaches 1. Multiplying this value by the reported intersection distance biases points that are at more viewable angles. This change just 'feels' better to use.
I'm super happy with how this turned out.
Small video
[video=youtube;a1bQcs4CIzc]https://www.youtube.com/watch?v=a1bQcs4CIzc[/video]
So I somehow ended up looking through the Postal 1 source code and came across one of the MSBuild files for their projects.
[url=https://bitbucket.org/gopostal/postal-1-open-source/src/f58b127b0202c532be101a5361a30820332efa14/Postal%20Common.vcxproj?at=default&fileviewer=file-view-default]Proceed at your own risk.[/url]
Someone stop me before I fix this abomination...
[QUOTE=CarLuver69;52493605]So I somehow ended up looking through the Postal 1 source code and came across one of the MSBuild files for their projects.
[url=https://bitbucket.org/gopostal/postal-1-open-source/src/f58b127b0202c532be101a5361a30820332efa14/Postal%20Common.vcxproj?at=default&fileviewer=file-view-default]Proceed at your own risk.[/url]
Someone stop me before I fix this abomination...[/QUOTE]
<ClCompile Include="SoundThing.cpp">
<ClCompile Include="thing.cpp">
<ClCompile Include="Thing3d.cpp">
;_;
[QUOTE=Icedshot;52493617]<ClCompile Include="SoundThing.cpp">
<ClCompile Include="thing.cpp">
<ClCompile Include="Thing3d.cpp">
;_;[/QUOTE]
I'm more concerned about this shit:
[code]
<ClCompile Include="ball.cpp">
<Optimization Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">Disabled</Optimization>
<PrecompiledHeader Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">Use</PrecompiledHeader>
<PrecompiledHeaderFile Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">RSPiX.H</PrecompiledHeaderFile>
<BrowseInformation Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">true</BrowseInformation>
<Optimization Condition="'$(Configuration)|$(Platform)'=='Optimized Debug|Win32'">MaxSpeed</Optimization>
<PrecompiledHeader Condition="'$(Configuration)|$(Platform)'=='Optimized Debug|Win32'">Use</PrecompiledHeader>
<PrecompiledHeaderFile Condition="'$(Configuration)|$(Platform)'=='Optimized Debug|Win32'">RSPiX.H</PrecompiledHeaderFile>
<Optimization Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">MaxSpeed</Optimization>
<Optimization Condition="'$(Configuration)|$(Platform)'=='Release(DebugLog)|Win32'">MaxSpeed</Optimization>
<PrecompiledHeader Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">Use</PrecompiledHeader>
<PrecompiledHeader Condition="'$(Configuration)|$(Platform)'=='Release(DebugLog)|Win32'">Use</PrecompiledHeader>
<PrecompiledHeaderFile Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">RSPiX.H</PrecompiledHeaderFile>
<PrecompiledHeaderFile Condition="'$(Configuration)|$(Platform)'=='Release(DebugLog)|Win32'">RSPiX.H</PrecompiledHeaderFile>
</ClCompile>
[/code]
What the actual fuck!?
Made some big improvements on my water system, got the laplacian figured into velocity although I think I did it in a hacky way and should read up on calculating more stuff based on potential energy or something. I should also just study a physics textbook
[t]https://i.imgur.com/07jbbmx.png[/t]
Implementing a memorycard manager for a PSX emulator!
[t]http://imgur.com/Yf5sFd6.png[/t]
[t]http://i.imgur.com/BHSHwtR.png[/t]
[editline]22nd July 2017[/editline]
Also character encoding is a pain in the ass man, wtf
[QUOTE=Karmah;52493336]All of yesterday I put off doing any work because I couldn't think of a way to cleanly implement 3D vertex manipulation, but alas today I gave it a go and managed to do it in under 2 hours :v:
My code is still a little bit janky, but this is how I deal with 3D grid snapping:
I define 3 planes at a desired position pointing down all 3 Axes (xyz), and determine if the mouse intersects any.
Generally the closest intersection to the camera is the desired one, however this can be a problem if the plane of the closest intersection point is being viewed at glancing angles.
Imagine trying to drag something across the X axis from -1 to +1, but as soon as it hits the origin (0) it starts dragging down the Z axis towards the camera.
The actual distance of the intersection point to the camera didn't matter, I just needed to know which was the closest but at a reasonable angle, so I modified the reported distance:
Taking 1 minus the absolute dot product of the plane normal against the viewing direction gives me a scaling factor where the value gets smaller when we are looking at a plane more head on, otherwise it approaches 1. Multiplying this value by the reported intersection distance biases points that are at more viewable angles. This change just 'feels' better to use.
I'm super happy with how this turned out.
Small video
[video=youtube;a1bQcs4CIzc]https://www.youtube.com/watch?v=a1bQcs4CIzc[/video][/QUOTE]
What is your GUI built with?
[QUOTE=Richy19;52496227]What is your GUI built with?[/QUOTE]
looks like imgui
Just made a simple string drawing function with a bitmap font
[img]http://i.imgur.com/g2CN0n4.png[/img]
[QUOTE=kakizc;52496242]looks like imgui[/QUOTE]
Whoops wrong rating
[QUOTE=kakizc;52496242]looks like imgui[/QUOTE]
Yep
ImGui is awesome and lightweight. I wholeheartedly recommend using it if you ever need small UI solution
Making an infinite "runner" game for Android just to get something done and released. You're a fish and you try to collect keys from pelicans and sharks.
[vid]https://zippy.gfycat.com/ImmenseNearGelada.webm[/vid]
I'm aware the triangle bucket beaks look ridiculous but I'm just embracing it V--
Sorry, you need to Log In to post a reply to this thread.
What are you working on? v67 - March 2017
