My experience with 3.6 is that it works, but you can't exit. If you reset with the button, you have about a 10% chance of getting a system error. I got a system error from this on 3.1 as well.

Since no one else has, I decided to take the plunge and write a version of 2048 in Nspire lua. It still has a few bugs, but it should be playable and fun

Controls: Arrow keys or 8/6/2/4 to move tiles. Enter to reset after winning or losing, or ctrl-del to force a reset during a game. Use any letter to change to a different save slot, to allow multiple games at the same time.

New: Now has graphics based on NBonaparte's modification

Well, if the fps is reduced, it will play closer to the correct speed. However, it won't increase the framerate on the calculator; that will stay low.

I thought it added the association automatically...

Yes, although the player doesn't actually use the Webm container.

Nvid is a video player for the Nspire. It is not the first such project, but it is unique in that it uses vp8 video. This allows long videos to be stored with a relatively small file size and good quality, although it results in somewhat slow playback.

Nvid supports both the Nspire CX, in full color, and the Touchpad/Clickpad in grayscale.

One example video is included, sintel.ivf.tns. To create your own videos, drag any video onto wrapper.bat in the "converter" folder. This creates a file called output.ivf.tns that can be transferred to the calculator.

Nvid does not have a file browser, so you need a version of ndless that supports file association. Run nvid.tns once to install the file association. After that, you should be able to open an ivf file and have it play.

Download here:
https://github.com/pbfy0/nvid/releases/download/v1.1/nvid.zip

I'm working on a vp8-based video player for the nSpire, but i'm getting very strange errors that seem to be unrelated to the vpx library. When I include the block
if(fread(frame, 1, frame_sz, infile) != frame_sz)
            die("Frame failed to read complete frame");, the program crashes as soon as it loads, before that line is even executed. At first, I thought the error was because of libvpx, but the version below, which has everything related to libvpx commented out, still crashes in the same way. This occurs in both the kArmTi emulator and a physical calculator. What am I doing wrong? The full source is available at https://github.com/pbfy0/nvid
/*
  Copyright (c) 2010 The WebM project authors. All Rights Reserved.


  Use of this source code is governed by a BSD-style license
  that can be found in the LICENSE file in the root of the source
  tree. An additional intellectual property rights grant can be found
  in the file PATENTS.  All contributing project authors may
  be found in the AUTHORS file in the root of the source tree.
 */
 
 
#include <os.h>
#include <libndls.h>
#include <nspireio2.h>


FILE            *infile;
nio_console csl;
#define printf(...) nio_printf(&csl, __VA_ARGS__)


//#define VPX_CODEC_DISABLE_COMPAT 1
//#include "vpx/vpx_decoder.h"
//#include "vpx/vp8dx.h"
//#define vpx_interface (vpx_codec_vp8_dx())
//#define clamp(x) (x < 0 ? 0 : (x > 255 ? 255 : x))
 
 
#define IVF_FILE_HDR_SZ  (32)
#define IVF_FRAME_HDR_SZ (12)
 
static unsigned int mem_get_le32(const unsigned char *mem) {
    return (mem[3] << 24)|(mem[2] << 16)|(mem[1] <<|(mem[0]);
}
 
static void die(char *text) {
    uart_printf(text);
    if(text[strlen(text)-1] != '\n')
        uart_printf("\n");
      fclose(infile);
    exit(EXIT_FAILURE);
}
 
/*static void die_codec(vpx_codec_ctx_t *ctx, char *s) {
    const char *detail = vpx_codec_error_detail(ctx);
    uart_printf("%s: %s\n", s, vpx_codec_error(ctx));
    if(detail)
        uart_printf("    %s\n",detail);
      fclose(infile);
    exit(EXIT_FAILURE);
}*/
 
 
int main(int argc, char **argv) {
      uart_printf("asdasdasd\n");
    //vpx_codec_ctx_t  codec;
    int              flags = 0, frame_cnt = 0;
    unsigned char    file_hdr[IVF_FILE_HDR_SZ];
    unsigned char    frame_hdr[IVF_FRAME_HDR_SZ];
    unsigned char    frame[256*1024];
    //vpx_codec_err_t  res;
     
      nio_console csl;
      clrscr();
      // 53 columns, 29 rows. 0px offset for x/y. Background color 0 (black), foreground color 15 (white)
      nio_init(&csl, 53, 29, 0, 0, 0, 15, TRUE);
      nio_fflush(&csl);
      nio_set_default(&csl);
      printf("hello world!");
 
    //(void)res;
    /* Open files */
      if(argc!=2){
         die("Wrong number of arguments.");
         return 0;
      }
      if(!(infile = fopen(argv[1], "rb"))){
         die("Could not open input file");
         return 0;
      }
      //return 0;
 
    /* Read file header */
    if(!(fread(file_hdr, 1, IVF_FILE_HDR_SZ, infile) == IVF_FILE_HDR_SZ
         && file_hdr[0]=='D' && file_hdr[1]=='K' && file_hdr[2]=='I'
         && file_hdr[3]=='F'))
        die("Not an IVF file!");
 
    //printf("Using %s\n",vpx_codec_iface_name(vpx_interface));
    /* Initialize codec */
    //if(vpx_codec_dec_init(&codec, vpx_interface, NULL, flags))
//        die_codec(&codec, "Failed to initialize decoder");
 
    /* Read each frame */
    while(fread(frame_hdr, 1, IVF_FRAME_HDR_SZ, infile) == IVF_FRAME_HDR_SZ) {
        unsigned int               frame_sz = mem_get_le32(frame_hdr);
        //vpx_codec_iter_t  iter = NULL;
        //vpx_image_t      *img;
 
 
        frame_cnt++;
            uart_printf("x%u\n", frame_sz);
        if(frame_sz > sizeof(frame))
            die("Frame data too big for example code buffer");
           
        if(fread(frame, 1, frame_sz, infile) != frame_sz)
            die("Frame failed to read complete frame");
           
            //if(vpx_codec_decode(&codec, frame, frame_sz, NULL, 0))
        //    die_codec(&codec, "Failed to decode frame");
 
       
    }
     
    printf("Processed %d frames.\n",frame_cnt);
    //if(vpx_codec_destroy(&codec))
//        die_codec(&codec, "Failed to destroy codec");
 
    fclose(infile);
    return EXIT_SUCCESS;
}

Edit: I figured out what the problem was. The nspire apparently doesn't like large initializing large arrays. When that block was omitted, the initialization of the frame variable got optimized out, so the crash didn't happen. I replaced the array initialization with malloc, and now it works.