jigg is a concept that implements nice features that squid should implement but doesn't, and perhaps never will. jigg runs on the same machine a web cache does (squid). Users configure their web browsers to point to jigg.

jigg will compress HTTP responses where it can, and when it sees a TEXT/HTML response, it will look for any IMG SRC links and tell the web cache to go get them (this is called "looking ahead")... so that by the time the client browser requests those links, the web cache will already have them. It will request the most occurring links first. See the diagram follow:

/-------------\                     /------------------\
  web browser   <----------------->   jigg <---> squid
\-------------/                     \------------------/
users machine                          web cache server

My boss asked me to write a C program from a Perl concept he developed (zipproxy.pl) based on Randal L. Schwartz's ideas. My job was to implement "Transfer encoding" plus the additional feature of "look aheads"-- to precache links ahead of the client requesting them. The idea is to speed dialup users web request/responses.

jigg is an experimental concept (but hopefully a useful working utility). I've developed jigg to the point now where I can say, "yeah the concept works" and go to my boss and give him my result.

01 Apr 2003: jigg (a year in the making), never got put to use. Two days after I installed it on our main Web Cache servers, Arafura Connect Pty Ltd was put into administration..