• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

194 Neutral

About ellisvlad

  • Rank
  1. Hi, I'm a developer for a Minecraft Group called Noxcrew. Basically, what I'm trying to do, is get videos with an alpha channel drawn in Minecraft (LWJGL).   Right now, I'm using VLCj to load and render the video to a OpenGL texture, and then drawing this to the screen. When I started, I would get segfault every time I loaded a video. This was because the memory buffer was being freed by Java's GC before VLC was finished with it. After storing the memory of the VLCj MediaPlayer in a static variable, I get a crash after 10 mins to 30 mins of running. I would assume also it's because of the Java GC. I'm guessing it is because the directmemorybuffer is being moved by Java from eden to a more permanent space, but this corrupts any pointers that VLCj stores.   I've been struggling to get VLCj to work for a few weeks now, and have a working-ish solution that just means I have to re-initialise VLCj every 15 mins of running.   The other half of the problem, is the video we are trying to render has an alpha channel. At first, we used just a chroma key to filter out a color as each frame rendered.. But there were problems with the way most video formats compress the video data. Then we tried using the Quicktime 32-bit MOV format, which supports an alpha channel in the video.. But results in a bitrate of ~300Mbps... Manageable as a last resort, but not ideal..   So my two questions are: 1) Is there a reliable way to load and render videos with an alpha channel in Java/OpenGL. Without using a 32-bit mov file with a fixed huge bitrate. 2) Is there a more stable alternative to VLCj to render this video ingame?   Thanks for the help guys!
  2. So I'm making a small game as a hobby project, and I'd like to implement plugin support. I've been trying to think of the best way to do this, and so far, it looks like using shared libraries that are loaded in at run time is my best bet.   So I made a shared library loading system and small API for some of the game features, and it's all going well... But then I made a mistake in my test plugin that caused a segfault, and brought down the whole game with it!   Is there a better way to do plugins that means they can be loaded in at run-time, but don't cause the whole game to crash if there is a bug in the plugin? Or a simple way to isolate segfaults so I can just unload the shared library if it causes a segfault? I'm using shared libraries because it seemed like the logical choice, but is there a better option that anyone can recommend?
  3. OXY - A programming language

    Hi mike, I would love to be able to post a long list of resources that I used when making the compiler, but the truth is, there aren't any good ones out there.. Before starting the compiler, I set out to find as many tutorials, guides, and other resources on how to go about making my own compiler...but I found nothing of use. Most of the time spent programming the compiler, I was offline, on my laptop, so I didn't have resources to go to.. The most important thing I have found with the project, and with many others, is to draw a diagram of how it is all going to work. With pseudo-code dotted around it if you think it will help to develop the diagram. This is how OXY works right now. 1) input a text file with code in it E.g. If (1==2) say(hi); 2) read the text file word by word (stopping at "; (" (those three characters, not the smily)) 3) interpret the read data and add a branch to an element tree with the type of command, and any extra data (parameters) 4) walk through the generated tree, branch by branch, and output as a byte stream. 5) save the byte stream to a file. (This is your executable byte stream. Encrypt/compress/pack however you like now) 6) runner loads the executable byte stream and walks through it, executing what it says. In essence, this is how OXY works, there are of course special cases (functions/classes) but those get coded in later ;)
  4. OXY - A programming language

    :D made good progress this week. A tech demo by the end of the month is a very realistic goal ;) Foreach would be a fun challenge to program in ;) I will need to get arrays working first though ;) Yeah, it's a good idea from a purely educational perspective. Sure, it can become more, but there's a million languages out there so don't expect it to replace C++ or anything like that... I haven't seen any projects like this... you can try, but you've got a lot of free competition. Are you learning? More functional features. Specifically, a good way to mix functional and imperative programming. Pure functional or pure imperative might sound nice in theory, but in practice (particularly for game programming, where speed and code architecture can make a huge difference) are difficult/limiting. Also, the ability to work with C libraries. Good luck!I wasn't sure where I should have put this post. Feel free to move it if it should go somewhere else. So far, the language is purely functional, however, there are some inbuilt functions that just run blocks of internal code, e.g. Creating a window, loading a sprite/font. But I am having great fun writing this, and I am learning a lot about C++ too. (I am also starting to appreciate how clever the people who wrote the third generation languages must have been, it isn't easy!) Great to see positive responses to my post :D. I was, until recently, doubtful that the project would be successful, and felt like giving up so many times, but to read that other people have interest in this language, even if you just want to see it working, is a great motivational boost for me. Thank you!
  5. Hi there, my name is Vlad Ellis, and for the past few months I have been planning on writing a programming language. On the first of this month, I actually started programming the project, and progress so far has been great! There is a working compiler and interpreter, which are capable of compiling and running some basic operations. So far I have implemented setting and freeing variables, if/else/for/while, math (using +-*/() and variables), and function calling. The next challenge now is to add function definitions, objects and included files. So far the language is very fast, and is working as intended. I hope to finish everything mentioned above and test it some more by the end of next month. :D I will then post binaries for both the compiler and runner here for anyone to test and play around with (Windows only for now) In the future, I will remove the need for a separate runner, and the compiler will output .Exe files. I also hope to start writing an IDE for the language some time toward the end of next month, which probably won't be ready for testing by the end of the month, but I will definitely post some screenshots of it. So my questions to this community... - Is writing my own programming language a good idea? It is mainly for educational purposes, but could become more? - I have seen some projects like this in the past, and normally the developers sell their cross platform compiler for a lot of money ($1,000+), and their languages are almost always slow and the compiled .exe is often bulky (4MB for an almost empty project?!) would it be realistic to try and sell my finished compiler? (for much much less of course!) - Am I wasting my time on this? :S - What is important to see in a programming language? (Functions, object-oriented, etc.) - Any other general feedback or comments/questions would be great :D Thank you for reading!
  • Advertisement