Yesterday saw the hand in of my dissertation on “Developing a Lightweight Game Engine in C++”, I can honestly say there is so much more I could have included in the paper, but sadly between time constrictions and the somewhat limited length of the paper I could not. But there it is it’s done, complete with a prototype game engine built in c++.

Everything seemed to go wrong yesterday so it was somewhat hectic, The problems began from the moment I woke up, My computer (Where some of my work was temporarily being stored) had decided not to let me log in (The user account had become corrupt for some unknown reason) after an hour of technical malarkey I was able to obtain the work from my hard-drive.

In order to get into university I rely on public transport which is somewhat lacking, I first must get either a train or bus to oxford then from there get a coach to high wycombe, by public transport this usually takes about 3 hours to get from my house to university. Living in a village the public transport is somewhat lacking, after having missed the train (the only train of the day) that morning I had to seek alternative means of travel. A close friend, understanding the importance of me getting into university offered me a lift.

After having got to university one of the first things I did was to go pick up my final bound dissertation from staples, which went smoothly but set me back by around £100.

There were a few other things that needed to be done before I could submit the work, one of those was to copy a cut down version of the engine to a USB to be handed in along with the dissertation, this is were some of the problems began, during the transfer process the laptop decided to crash which very much slowed it down with the deadline creeping up this really slowed down the process, In the end due to my issues with my laptop I transferred to one of the university computers. The process should have took 5 mins but in the end took 45ish mins.

Now the final problem that occurred during this day, the whole final project had 4 expected deliverable s,  2 bound copys of the report, media containing the software that has been developed and an electronic online submission. The problem came with the electronic submission, I was submitting the electronic copy around about the same time that my laptop decided to crash, I had gone through the submission process uploaded the file, only to come back later that day and realize that it had been neglected slightly because of the problems that were occurring at the time, I opened my laptop only to find that the turnitin system that handles the online submission had timed out and as such the electronic copy had not been submitted, after the realization the document was submitted straight away, hopefully this issue will resolve itself but did somewhat a final blow for the day knowing not all the submission requirements had been met. This was the part of the submission that I though would be least problematic but in the end it ended up being the most.

Anyway that’s all for now, Once this course is over I’ll be able to get down to some programming, maybe update the website, upload some stuff it’ll be fun.

Good Day!.

So right what have I been doing, what have I been doing hmmm… well I did go on about uni work and not having enough time for anything it seems very much true, even lost a very beautiful young girl over the matter. But life shall go on.

Right Updates
Coming up to the end of uni, really need to work on dissertation, engines still absolutely terrible and I’m absolutely starting to freak out about the whole deal, but I’m pretty sure that’s the same for most students. As far as the actual engines concerned I’ve done “little” work little in the sense that there’s so much to do at the moment I’ve barely dented the surface. So many problems cropping up, bad code things I regret.

Engine Updates -> About 4 days ago I decided to start making a Minecraft Clone on my engine has been coming together very quickly. Screenshots below from the start to now. Texture Pack credit should be directed to BDCraft.net.

1) Modded/ Hacked the existing TileMap class to add an extra Dimension, For those/ everyone that doesn’t know the TileMap class in my engine used a chunk based system in conjunction with opengl’s diaplay lists (Don’t Judge) to render some pretty large maps.

2) My engine has little in the side of physics at the moment, box2d can technically be used but is not supported, So I added a few new colliders to the engines native physics system, this included an AABB(Axis Aligned Bounding Box) and a Ray. The systems very much brute force at the moment, but is currently not much of a concern as we will be working on a grid area with a limited number of collidable objects. Using these two new objects I went about creating another class to handle the collisions in the world itself a new type known as CubeMapCollider, it takes a reference to the CubeWorld and uses this to basically  work out collisions that may be occurring. When checking for collisions in the world from a ray the CubeMapCollider will work out a set of tiles within the bounding volume of the ray before then checking each tile/ block to see if a collision has occurred, after computing the collisions and distance it will work out the collider object which is first struck. At the stage depicted below, the actual collision algorithm was slightly questionable, at this moment in time I had not worked out how to detect which specific face the ray had collided with on the AABB so for the time being I just added a new tile above which ever tile the ray collided with when the right mouse button was pressed.

3) At this point I was working primarily on optimizations(reducing memory usage, increasing FPS), I made sure only clockwise winding faces are drawn for a start, then I moved onto culling sides of the blocks that where not visible, and various other optimizations. in the end I managed to render a world consisting of over 2 million cubes-> to quote myself on facebook: “current rendering a cube world of 256X32X256 with chunks of 32X32X32 so around 64 chunks with a total of 2,097,152 blocks in the world and getting between 290 – 320 FPS on my computer. Not bad.”

4)The culling of hidden faces left it problematic when dealing with blocks with transparent textures, Its an easy fix but is not fixed yet.

5)By this point I had got the ray AABB intersect perfect next was to work out which face was hit first by the ray and to be able to place blocks attached to said face.

6) Just rendered out 4 chunks of 32X32, messed around a bit as I plan to start to think about chunk + world generation. 

6) Finally got collisions down flawlessly I might add.. 

7)Added functionality that makes it possible to destroy blocks.

Sorry everyone I find it hard to update my blog regularly at the moment, I have a very busy schedule between uni work, social life and my own personal programming so I find i hard to do much else. Not to mention the time I spend travelling (6 hours to uni and back every day I go in).

Updates:

Unis going well got some good grades so far this year hope by the end to have a good classification, very interesting lecture yesterday an X-game designer came in to do a presentation was very informative. Continuing to work on my dissertation and project “Lightweight Game Engine Developed in C++” see nightmare engine blog, hopefully by the end of this month Box2d will be integrated… just finding it hard to find time to get it done at the moment.

Minor update to PSVPAD, yes there was a minor update to PSVPAD, completely changed how the program operates, still very simple application in itself and the codes not totally satisfactory but gets the job done. Managed to play some games very efficiently with a weak connection getting no lag so if you guys can get that to work then happy days its far better than previous versions. It was always planned to be a dedicated remote controller however. I have been discussing with some close friends the possibility that I might add option remote desktop/ play support which is a consideration at the moment. I will definitely clean the code up at some point so expect another update soon.

Thats all for now

have a good day

Just so people know all comments posted are moderated at the moment due to spam, So don’t worry if it doesn’t show up straight away.

I’m working on an update for the application for all those interested… However is not as much an update as a rewrite so it may take some time, I’ve encountered some bugs and problems along the way… The PSV PAD was actually the first piece of networked code I have ever written so it was actually quite poorly written and not very well optimized, hopefully with the update it will have substantially less lag, all tests so far have shown positive results. I’m actually currently having problems with data getting jumbled together along the way which is resulting in the application freezing and crashing at the moment due to desalinization errors on the server side. The new version of the application groups all the data together on the vitas side serializing it then sending the data over the network together. As soon as I’ve fixed the bugs and got it running smoothly I’ll post it up here…