back
46 comments
Deleting Geocities was a horrifyingly destructive act by Yahoo. They took thousands of sites that had been built by other people and just deleted them all. And this isn't an isolated incident. The industry seems to think that taking gobs of user data and trashing it is fine, and it happens on a regular basis.

We need to change the culture of the tech industry to view users' data as belonging to the users. Until then, consider donating to the Internet Archive https://archive.org/donate/

That's an awful move by itself, but what I personally cannot get over is their decision to rick-roll every user who clicks on the outdated link. That simply insulting.

Here's a random example: www.geocities.com/scythemantis

What the fuck. They are literally rick-rolling people. And Yahoo is a publicly traded, supposedly reputable company. Also, I thought the Internet Archive was mirroring Geocities. Why don't they redirect people there, with a legal disclaimer, or just give the Internet Archive the domain name. There is so much additional baggage with the name, Yahoo could never reuse the name for something else.
This series of concerns is close to my heart. Some time ago, I built a toy site that lets people make Gifs out of YouTube videos[0]. I put no limits on the service since I'm lazy and limits would have required extra work.

This means there's no limit on the size of gifs created, nor any limit on their length. Additionally, a user can make as many gifs as they like. I think one can imagine where this is going.

Since it's inception, people have made tens of thousands of gifs, and now I have hundreds of gigabytes of gifs I have to host. Recently, to save space I had to go through and delete pretty much all the gifs that had been created before a particular time. I downloaded them to my personal computer to ensure that some record of them remains, but still, I potentially broke tens of thousands of links.

In the future, I'll be putting all sorts of limits on user creations, but I'm curious how bad action like this is?

[0] - http://gifmachine.xwl.me/

You could join #archiveteam-bs on efnet and talk about it? Also if you want to send me a copy, my email is in my profile :)
Actually, I'd love to!
Yahoo gave plenty of warning and allowed users to export their data (I didn't use it so I can't say how complete or easy the export process actually was). What else would you like them to do, keep the sites up indefinitely?
That's one option. Geocities in particular was a bit over 1TB, so it wouldn't have been too expensive to host it read-only. Or they could send the data to an archive like IA.
Their whole user built data was 1TB ? 50$ worth of hard drive .. sad.
Plenty of people archived Geocities; if you google for it you can find a couple. A friend of mine made some effort to do this, I don't know if his version is still up, but I'm sure he has the tarball somewhere.
Plenty of people got partial grabs of Geocities. No one knows how much was lost. This is the most complete single collection I know of, pieced together from several people. https://thepiratebay.se/torrent/6353395/Geocities_-_The_PATC... If you just want to browse, Reocities is probably the best interface. http://reocities.com/neighborhoods/
I have cherished, fun and nostalgic memories of learning HTML as a kid and building sites on GeoCities. It seemed so amazing at the time. Breaks my heart to think back on it and know that it's all been deleted.
Are they just supposed to keep the servers and infrastructure which costs them money running forever, no matter how unprofitable?
It's hard to believe that an archive of sites meant to be consumed on dial-up modems, most abandoned, would today require particularly high bandwidth costs. Yahoo has been faltering for a long time, but it's still worth tens of billions of dollars. So yes, I think they should have kept a (read-only) archive running forever, despite not making any money from it, as an act of good internet citizenship with respect to such a historically important website - or at the very least not have forced the Internet Archive to race to manually crawl as much as possible before the deadline.

edit: As a general note, this is IMO one of the fundamental technical deficiencies of the Web today. Keeping website information around shouldn't require active maintenance by the publisher. Imagine if everyone could no longer read the contents of a book once the publisher stopped printing it - what that would do to our access to historical culture? In an era where most users have abundant storage at least by the standards of text, it wouldn't be that hard to invent a protocol where visitors to websites automatically mirror them (the text at least) and mirror them to other clients, forming a decentralized filesystem. Compared to this, the WayBackMachine is an impressive effort, but it's limited by its centralized nature, managed by a single, relatively small non-profit organization.

You should tell your users upfront if you're going to delete their data. Something like "Upload your photos, share them with the world, then never see them again after a few years." https://web.archive.org/web/20140810180036/http://www.fotope... That one was grabbed by the Archive Team, who gave the data to the Internet Archive so you can still browse it. 2TB of photos, information, discussion, nearly deleted forever. Edit: fixed link
https://web.archive.org/reporter/magazine is ironically a 404.
Oh, thanks. I guess something on the page changes the URL in the address bar after the page loads.
According to this post it looks like they were given notice: http://www.businessinsider.com/yahoo-final-notice-geocities-...

Generally shutting down a service means you won't have access to it and the data on it anymore.

Having notice is better than not having notice (like the AOL Hometown shutdown). But it's not an excuse. It's not your data. It would be one thing if you literally ran out of money, which Yahoo obviously didn't. And many startups that delete user data are "exiting" or "pivoting" and don't have that excuse either. In addition, you should be transparent with your users if it looks like they might lose their data in the forseeable future.
Maybe it's the users' fault for relying so much on client-server architectures. There are some initiatives for P2P/F2F social networks but they need help.
People have been using Facebook and complaining when it spills out their info years after Facebook got bad publicity for privacy related offenses. People still get offended and complain when their iOS app gets rejected years after Apple got bad publicity for needlessly censoring apps. People still use tumblr a year after Yahoo bought it and deleted many blogs to clean the site for advertisers, and that after they shut down the previous social network they bought. The users have a choice, despite how annoying those sign-up nags can be. They chose to use services which have reputations anyone can learn about if they care to do so, so yes it's partly their fault.
There we go, blame the user.
Most of these comments and similar above sound great on HN, but terrible ideas when you are actually operating a business which is trying to make money. You drop your unprofitable services and move those resources to other areas of your business which will make you money. Google gets heat and does this all the time, and they do it for smart business reasons.
If an unfunded team can grab 9TB of Reader data, and the Internet Archive can host it forever, Google really has no excuse here. https://archive.org/details/archiveteam_greader
I think the grandparent would argue that they shouldn't solicit user data unless they're prepared to pay reasonable costs of archival
Great headline, but there is not much to the article. I clicked around for a few minutes thinking I missed something. Am I an idiot?
Nope, thought the same thing. I had an awesome Duke Nukem 3d site on geocities. It had gameplay tips, easter eggs, and tips and tricks on how to build your own maps and scripts for Duke Nukem. This may have been my first attempt at programming, and was easily my first glimpse into network protocols
The page itself uses the garish late 90's web design gimmicks, but the article has amusing links in the sidebars to other galleries.
Olia Lialina's work on the terabyte of Geocities that was salvaged [you might want to turn down your speakers]:

http://contemporary-home-computing.org/1tb/

If you miss GeoCities, try http://neocities.org/
Edit: I have turned this into a blog post: https://neocities.org/blog/we-are-the-future
This is nice.

It would be helpful to have an RSS feed on the main blog. RSS is still widely used too!

Not sure this was intended, but I'm on a bus wifi, and the background sky image is loading one row at a time, feels more reminiscent of that time than the animated images.
"Vaporwave" music and art is all about those late 90's and early 2000's graphics and sounds. If you're interested in this stuff you should check it out: https://www.tumblr.com/search/vaporwave
It's not really for me but I'm not a big fan of this stuff. I went to a JS conference recently and a whole day of it was devoted to "LOL marquee tags and 3D text GIFs". Silly me, I thought we'd talk about writing code.
My very first website was on GeoCities. After a while my site got too popular so I moved to paid hosting.

It did a fantastic job for what I needed it for and was surprised that Yahoo managed to completely fuck it up.

Oh, I thought we'd moved on to the 2000s? I was just about to post an old collection of MSN and AOL portal screenshots on my tumblr. That was close. Any idea around when the 2000s will be in-vogue? Some of my case studies indicate it should have begun already. My baseline is 1000 notifications a post so call me when it's time please.
I don't know if I should up vote you or down vote you.

I agree that articles like this one are a pain. They are pitching a misplaced nostalgia. The "good ol' days" weren't good. Even then, we knew that these sites were bad. We just didn't know what else to do.

I used to be torn on the Tumblr blogs that create the content that ends up in articles like this. On one hand, link-baity articles that keep hammering on one subject, over and over again, wouldn't exist without them. But on the other, I understand that need to explore and create things, even things that aren't a good aesthetic. I did an entire series on taking cellphone photos of prisms that had lasers passing through them. Hundreds of photos that were mostly junk. But there was something in there that I couldn't articulate, and the process of posting helped me figure it out. I did it for the same reason I played Sudoku for a year straight, but haven't touched it again in 7 years: I had a need to figure it out.

It's one of the reasons why I think there is no meaningful distinction between art and science. Your brain subconsciously recognizes that something is there, gets snagged on it, and you use the tools at your disposal to test, test, test and figure it out.

I know very few working, successful (even if marginally so) artists who make exactly what they want. Most of them are trend following on how the market reacts, and they aren't exactly cognizant of the precise elements that make their work popular.

It's our fault for continuing to buy it.

But regardless, I also don't think it deserves such a flippant reply. HN gives you all the characters you could want to write a reply.

Well you probably should have up voted me but it's my fault for being unclear about my opinion while trying to be funny.

I don't think that this aesthetic is stupid or anything, it's only that what's represented in this article is the same kind of content you would see when tumblr had almost just been created. People were very attracted to these aesthetics, some focusing more on the uncanny aspects while others presented a more quaint and nostalgic picture. It gave rise to a whole plethora of visual and even musical sub-genres. People developed sometime very fine and subtle aesthetics.

Reading this article makes me feel the same way an average HN'er would feel reading a 2015 article praising the merits of that new paradigm shift called Object Oriented Programming. You would expect at least one or two sarcastic comments.

Even *ing Rihanna 'borrowed' the style 3 years ago [1]. And we're talking about the most mainstream US singer since Michael Jackson. Still I know techies aren't exactly the most art-voracious demographic so maybe I'll try to dig up some google stats to prove my point.

[1]: http://www.spin.com/articles/azealia-banks-rihanna-seapunk-b... (2012)

The aesthetics of Geocities was admittedly pretty questionable :) But it is still interesting. http://contemporary-home-computing.org/1tb/
There is no accounting for the amount of time I spent trying to pick the perfect background music for the websites I built in the late 90s. It never occurred to me that you'd even want a silent site.

But, I guess that's the nature of youth. We're supposed to be a bit embarrassed of ourselves, or else we never grew.

Thanks for that link. Fun memories.