As I looked at things a little more closely, I did notice that all of the comic titles all those broken links belonged to were ones from the GoComics website. So I thought there might be something wrong with their server. When I tried calling up the comic images individually, I'd get a 404 error so I knew it wasn't a server problem at least. Maybe there was a formatting issue that screwed up how the file names were spelled out. Nope, re-typing the file name to match the day before's -- which I know for certain had worked because that's how I read all of them -- resulted in more 404 errors. I did some more checking. 404s for all the individual titles' directories. 404 for the directory all those directories were residing in.
I got a 403 error when I got up to the server level itself, which meant the server was still online. But the 404s meant that those individual files were gone. They didn't change acccess levels, they removed the files and directories entirely.
Thirty years ago, when newspaper syndicates first started putting all their comics online, they did so in a pretty basic fashion. They had a directory for each title, and uploaded each new strip with a file name that was a combination of an abbreviation of the title and the date. They needed to do something structured like that so they could keep all of their work organized and the whole server didn't become a jumbled mess after a week. The problem, though, is that structure was also easy to understand and therefore easy to develop a script for. This mean that just about anyone with a rudimentary knowledge of virtually any coding language could write a script to automatically take that day's comic and drop it into their own website. Which, if you make your money by selling subscriptions to that exact content, means you'll have more people linking to your content for free rather than paying for it.
They eventually realized this, and re-worked their website so that the images presented there were given a randomized file name when it was served to the user. They kept the naming structure internally so they didn't lose that organization, but they had a script that effectively anonymized that for public presentment. This would've been back in the late '90s, I believe.
My guess is that at some point, the syndicate told newspapers they were going to sunset that server and they gave them a window of a month or so to switch their websites to refer to the new server with anonymized file names. I expect more security was built in, so they needed a password or access key or something. And the newspaper all switched over because readers tend to get vocally upset when they can't read their comics, and there was no way an editor was going to risk that backlash over a stupid tech issue.
Except the syndicate never turned off the old server. Or the script that was copying files to it. It was no doubt set as an automated task, and maybe the programmer who put it together in the first place left; maybe they forgot about it; maybe they got busy with the next project... Eventually, Uclick was fully absorbed by Andrews McMeel's Syndicate in 2009; I'm sure there was a massive staff shakeup and any of the folks who were even still left by then were likely let go. So that old system just kept chugging along in the background, without any institutional knowledge at Andrews McMeel that it even existed.
It would have been pretty inconspicuous. Back then, there would've been few 'hackers' interested enough to do anything with the information. And you couldn't get access to EVERY comic they were running; mostly just well-established strips. So I'm sure most folks who did know about it eventually opted for other methods of getting to those comics. I know I certainly stopped referring to them on at least two separate occasions, and only returned when that new source broke and I'd find myself digging across old code I had written years earlier. So the amount of traffic that server would still be getting would be neglible at most. And as hard drive space has gotten cheaper and cheaper and cheaper, the amount of space these duplicate image files would be taking up relative to their main sites would be rounding error.
So that server has been quietly hosting a slew of comics in the background, available essentially for free to anyone who knew they were there, for the past thirty years. "Hacking" access to those images wasn't even Web 1.0 levels of difficulty, so much so that I'm not even sure it even is considered hacking.
I've long since figured out other ways to get those strips I was reading, so it didn't take long to rework my personal comics page and drop in other sources for everything. And I don't really have a point other than to mark the very end of an era. There's very little in the way of Web 1.0 remnants floating around and still being updated -- the only comics outlier I know at this point is, fittingly, the Zippy the Pinhead website, which creator Bill Griffith still seems to update by hand every day -- and given the state of the world, that's a shame. Because that level of technologicial naivety was charmingly refreshing in an age when every website wants to track everything you do online and you get security alert notices on a daily basis about how this app or that website has been compromised, and all your personal information has found its way to the dark web once again. That that old system had been just doing its thing in the background, largely ignored, for three decades was amazing in its simplicity. RIP to the remnants of the original digital frontiers.
0 comments:
Post a Comment