Date: Thu, 01 Aug 1996 02:09:50 -0400 From: Dan Wallach <dwallach> Subject: Dilbert Hack Page - Technical Details To: cyberia-l@listserv.aol.com
I've been advised to moderate my opinions in a public context until I resolve things with United Media, but I can clarify the exact technical details behind how the "Dilbert Hack Page" works.
Back in the old days (about a year ago), United Media always had the current Dilbert available as "http://www.unitedmedia.com/ comics/dilbert/todays_dilbert.gif", and people inlined it all across the Web.
Sometime in December, they changed all the comics to have the date coded into their name. Shortly thereafter, they appended a pseudo-random four-digit number after the date (example: a filename for the July 31, 1996 Dilbert image might be: dilbert073196-9876.gif). This thwarts any casual attempt to inline images from their server.
My script visits their server once every evening, reads their HTML, and copies the image *names*. Further, since they slightly change the image height by a few pixels each day, my script loads enough bytes from the images to read the image sizes from the image header and then promptly closes the connection.
That's enough information to generate my page.
When somebody visits my page, they're receiving the images directly from the United Media site, as directed by my page.
I won't discuss my opinions on the legality of my page in a public forum, but I'm in the process of collecting feedback, especially from lawyers. Once I assemble a quorum of legal opinions, I'll follow their advice.
At the present time, I don't have the financial resources to properly defend myself in court. If anybody can address this issue, please contact me as soon as possible.
Thanks,
-- Dan Wallach Princeton University, Computer Science Department dwallach@cs.princeton.edu http://www.cs.princeton.edu/~dwallach/ PGP Ready