🚀 go-pugleaf

-> By the way, the basic algorithm of LHA is in my public-domain pre-LHA
-> experimental archiver ar002 (which I wrote when I was a high-school
-> teacher).  It's in CompuServe's IBMPRO Library and also on the disk

I was wondering if you have considered any improvements to LHA's
compression, or any increase to its speed.

What do you think programs such as ARJ are doing to get compression
improvements over LHA?  Simply increasing the buffer size from 8K to 24K
can't be all it is.  Perhaps there is a better way of encoding the
huffman tree and outputting it - there may also be a faster way of
generating the huffman tree.

Also, do you know of a faster way to locate matches (than the digital
hash trie used in AR002)?  It doesn't matter if it takes more memory;
I'm just looking for speed improvements in the match finding process.

AR002 is a great program!
--
Canada Remote Systems  - Toronto, Ontario
World's Largest PCBOARD System - 416-629-7000/629-7044