Sorry to post such an irrelevant question for today's compression research, but I'm trying to compare older compression standards and I ran across PKZIP 1.x's "implode" method and am wondering if there is a description of it anywhere. I've RE'd it for close to an hour and, unless I'm mistaken, it's very close to deflate except it uses a 4K dictionary instead of deflate's 32K. Is this correct, or am I missing something? The PKWARE APPNOTE.TXT and the Deflate RFCs don't mention PKZIP 1.x "implode" anywhere

(In case you're wondering why I'm looking at ancient history, it's because I wrote a decompression routine for 8088/8086 CPUs as part of my hobby and I want to compare it to the pkware DCL's performance. And yes, I believe my code to be the fastest of its kind ever written for the 808x but of course I want to back that up with results before making everything available