- 6 hours
What have I done this week and how the program has progressed: - Start trialing out "combo" algorithm (running input sequantially first through LZW and then Huffman, should achieve efficiency decently close to DEFLATE (the algorithm zip uses, which is a combination of LZSS and Huffman coding) - Improve binary IO (arbitrary input files to compress should work) - Improve performance and stability for large inputs, add benchmarks for 6.2mb simple and complex files
- Run and document week's benchmark runs in the [implementation document ](https://github.com/xylix/tiralabra-syksy-2021/blob/main/dokumentit/toteutus.md))
What have I learned: - The program is starting to be quite done as far as required features go. - Compairing compression algorithms in a table is not trivial because there are multiple dimensions for performance and they change differently based on input data's characteristics.
What has been left unclear or is hard: - Nothing this week.
What to do next: - Not too sure. Hoping to get some ideas from peer reviews.