Ok I Took A Look At This And Heres How It All Works...
First Of All It Looks As If The HTML Is Returned In Like GZip Encoding.. After You Get Past That You'll Find That The JPG Is Secured By A Dif Cookie Seq/Query Seq Of Chars.. Whats Really Interesting Is That The Codes Chg On Each Reload. So If You Get The Pic Once And Are Able To Copy Shorcut Then The Code Is Prob Invalidated. So Once The Pic Is Downloaded You Must Prob Reload The Page To Get New Codes WITHOUT Redownloading The Img, Also The Order That The Several Dif Images Seems Important Here. The Codes Sent Back Prob Represent The Useragent, Or A PHP Session Etc...
The Server Is Apache, Running Mod Perl, PHP 4 Etc...
Ok So The Final Thing Was To Simply Socket Scrape This. I Have The Img But Whoever Created This You'll Have To Look At The Bytes. I Had To Cut/Paste The Bytes In HEX Into A Txt Doc, Then Wrote A Quick Script That Converted Back To JPG.
It Is Eitheir That You Are Doing Chucked Encoding On The Return Or I Screwed Up One Of The Bits(Aprox 82Thouand Bytes)..
So The Pic Is A Little Screwy Of The Bottom However If You DL The File You Will See That The Bytes Match To The File On Your Server.... Here Is The URL:
http://www.candidhosting.com/indexS.html
Bobby - Senior Programmer Posting For Candidhosting