David Gelernter predicted Dropbox in 1996 | Hacker News Reader

David Gelernter predicted Dropbox in 1996

blog.garrytan.com

42 pointssaurabh35 comments

The issue we face isn't about predicting in broad-strokes technology that would be useful. It's all the minute decisions that turn the broad-strokes into something great.

When something becomes popular, people often look for people who "predicted" it. I remember the article about how Microsoft's Nathan Myhrvold "invented" the iPhone in 1991. Except that he didn't. Microsoft made a hand-held computing OS before the iPhone and Palm Pilots existed a decade before the iPhone. In broad-strokes, they're the same thing. However, they weren't popular (and not for lack of trying). The issue wasn't the broad-strokes, but the minute details that made the iPhone (and other devices since) a joy to use. His work probably led to Windows CE, PocketPC, and Windows Mobile. He didn't invent the iPhone.

Similarly, before Dropbox we had options like XDrive from AOL, WebDAV servers, FTP, iDrive from Apple, and many others. We were awash with crappy software to store our documents on a server and access them from any computer we were on. Even Microsoft offered a slew of sub-par options that would get re-done as a Dropbox competitor as everyone started recognizing that Dropbox's model was wonderful.

Giving David Gelernter credit for predicting Dropbox is to show a lack of appreciation for the decisions that made Dropbox great - and better than the army of options that came before it that simply weren't good. Dropbox did what many before it couldn't do: create an online storage system that people love. I remember trying a bunch of services like XDrive back in the day. Transfers would fail, the transfers were synchronous (so you'd get a pause until the write had happened), things weren't cached locally so load times were poor, the mount was finicky, etc. They all satisfied the "making documents available everywhere" criteria similar to Dropbox, but they simply weren't good. Dropbox worked hard to make it good. They made lots of little design/technical decisions that make it Just Work. Many before them had tried and failed.

> Giving David Gelernter credit for predicting Dropbox is to show a lack of appreciation for the decisions that made Dropbox great - and better than the army of options that came before it that simply weren't good.

I was extra impressed by was how close Gelernter was to predicting the cost and nature of such a service. It's an impressive thing to guess about $12.50, and that it would be a simple and reliable monthly service. This was an era where 1 GB hard drives cost $1000, mind you.

You're right, Gelernter couldn't have known the specific details that would have made Dropbox successful. If you could know that, you would have created Dropbox.

> If you could know that, you would have created Dropbox.

No, you probably still wouldn't have. That's just one of many factors that allowed a number of people to create Dropbox.

Besides a vision you need the right timing, addressable market, execution skills, presentation, determination and so on.

> Similarly, before Dropbox we had options like XDrive from AOL, WebDAV servers, FTP, iDrive from Apple, and many others. We were awash with crappy software to store our documents on a server and access them from any computer we were on.

Amen! Don't be too hard on third-party vendors, though. Dropbox could only have been created after file-monitoring APIs became commonplace in OS kernels. If I've got my history straight these APIs were introduced because first Apple (Sherlock, Spotlight), then the GNOME guys (starting with Eazel's Medusa), and later Microsoft (after WinFS failed) wanted to do nearly-near-time full-text indexing. I remember that in Linux land these APIs were quite contentious and it took several iterations before they got them right.

I'm not sure if the ideas underlying Dropbox were _that_ unique at the time, but their execution sure was exceptionally good.

Still, some companies _were_ in a position to invent Dropbox before Dropbox did. The obvious ones being Apple and Eazel, since they had products that were monitoring the entire user file system _and_ they were selling WebDav storage!

Many already did. Ray Ozzie started Groove Networks in 1997 to do file sharing and collaboration, later bought by MS. MS also bought a file sharing company in 2005 to power Live Mesh. That MS failed to nail this market on their own platform is kinda' pathetic. There were a billion other companies before Dropbox. I've never understood the enthusiasm for Dropbox. It's all been done before.

I felt the same until I started using it. As boring as it sounds, just working and being easy enough for your grandparents to use is apparently enough to get people excited. The fact that it's been done (or tried) so many times before really just makes you appreciate it more.

Yeah, MS, Apple, etc. all should've had this figured out years ago and yet they didn't. So boo on them, but it's probably for the better because now we have Dropbox and you can use it on your Mac or in Windows, on your iPhone, Android, or whatever...

A former boss of mine has a theory that I endorse wholeheartedly- if you want to make a hugely successful business, start with one that is totally unaffordable in the current climate, and wait for change.

That's how YouTube started- when it launched everyone was blown away by the sheer amount of bandwidth they were eating though every day. Lo and behold, bandwidth costs dropped. The same goes for Dropbox- 2GB for every user, for free? Are you insane? But storage costs dropped and it's an affordable model. So it's one thing to predict Dropbox, it's entirely another to actually execute on it.

I do wonder what the next company will be to follow this pattern. I'd like to believe that it's Spotify, but their limitations aren't technical, they're the entire industry. So there's no guarantee that they'll fit the model.

It's called chasing the cost waterfall from monopoly to commodity and making sure you own it with economies of scale (from the book How We Got Here: A Silicon Valley and Wall Street Primer (A History of Technology and Markets) by Andy Kessler).

Larry Page has also talked about the same phenomena (http://www.youtube.com/watch?feature=player_detailpage&v...). Recommend watching the entire talk.

Follow falling costs - they drive adoption.

> The same goes for Dropbox- 2GB for every user, for free?

While it was a contributing factor to Dropbox success, it was hardly the most important one. 2GB was perfectly sustainable even back then if you consider than only a fraction of users would use all of their 2 Gigs. It's the overall execution and the focus that mattered. I'd also guess that aggressive affiliate marketing helped quite a bit too.

That wasn't a prediction. He, like anybody other academic connected to the Internet of that era had been enjoying drop box like services for years. Both at work, and if you added the correct digital service in the home. I lot was done with FTP at the time and there were a lot of specialized clients that did some really neat stuff. So the statement was more like: other people are really going to enjoy some of the luxuries I already have.

You're right. In 1996 my home directory was stored on an Andrew File System folder that allowed me to securely authenticate from untrusted networks using Kerberos, and then access my files remotely, from multiple machines, and it used local cache used to keep things speedy.

It was incredibly strange that these sorts of capabilities weren't meaningfully commercialized, even though they were ridiculously useful.

I would also ask "so"? Everybody can imagine lots of stuff. Its far more important who will make them real and economically viable. I was dreaming (e.g. predicting) YouTube in 1998, but that does not mean that I consider that I've done something big or I've been the first.

So? [1]

  [1]http://en.wikipedia.org/wiki/WebDAV#History

and [1]

[1] http://en.wikipedia.org/wiki/Network_Computer

Douglas Engelbart predicted the whole modern computing world, including cloud services, in 1970s. It still took hard work of thousands of people to realize it in a way useful to a wide population.

http://en.wikipedia.org/wiki/Douglas_Engelbart

http://en.wikipedia.org/wiki/The_Mother_of_All_Demos

Someone's going to make a service that will make us forget about Dropbox overnight.

You can quote me on that whenever it eventually happens.

Not sure how impressive this was. My recollection of 1996 was very much a lot of talk about Java terminals, e.g. machines dedicated only to running some dumb/thin client with backend servers doing all the work/persistence. That was a "future" that was only supposed to be a couple years out.

Ok, what do you say about this prediction? http://en.wikipedia.org/wiki/America-Lite:_How_Imperial_Acad...

Thanks for the reference. I enjoyed reading some of the reviews on Amazon that summarized the book.

Edit: I just bought the book.

I just finished tips book, pretty bad. In my opinion he sounds like a sexist and his attacks on Obama are really over the top.

This part jumped out at me:

"As a solution, Gelernter proposes moving all of human knowledge to online servers so that the in-person college experience can be replaced by user-driven self-education."

The most important way people are making money on the Net is on the model of the yellow pages or direct mail companies, by providing a service that you get for free and making others compete for placement.

"If you are not paying for it, you're not the customer; you're the product being sold."

- http://www.metafilter.com/95152/Userdriven-discontent#325604...

So in this case scenario, how is dropbox making money off me? Note I have one 1.9GB true crypted file that i host on drop box for years. please tell me, how am I being a product that is being sold?

I think ronilan point was that DropBox's type of business (where customers pay) will always be second tier to Facebook's type of business (where customers get it for free).

"We scanned your documents and found you're a 25 year old man who really, REALLY likes self-help books about depression"

I am getting tired of hearing that argument. That is a gross generalization, which is far from always true.

http://en.wikipedia.org/wiki/Win-win_game

Isn't this a quote from "The Filter Bubble"? [0] Fantastic read by the way, for anyone who is interested in (or terrified of) behavioral advertising.

[0]: http://www.amazon.com/Filter-Bubble-Personalized-Changing-Th...

The quote on Metafilter (August 2010) precedes The Filter Bubble (May 2011).

The company he is describing is not dropbox, it is Evernote. At least today. Dropbox does seem to be going towards adding value on top of the stored documents instead of just being a cloud FS. So, Evernote and Dropbox are on a collision course.

When did Microsoft start trying to build Dropbox? Iirc they made at least half a dozen attempts (that failed) before Dropbox came along. Presumably one of those attempts started around 1996...

Come on, people have been predicting utility computing for decades. See Multics, for example.

I predicted that too. So what?

David Gelernter is awesome. Linda distributed memory, Lifestreams, etc., etc.

I had a crazy idea about ten years ago to implement a simple, pedantic version of Lifestreams for a book project and someone at his company more or less said that would be OK.

Anyway, they guy is full of wonderful ideas.

It just occurred to me that Evernote has become, for me, a form of Lifestreams. I use the app on my droid to capture lots of pictures, scan important paperwork and documents, record interesting and potentially useful content on the web, etc. My only issue with Evernote is a small concern that all of my personal computing devices use SSD drives, and Evernote'ing a lot of stuff makes me hope for more and more storage capacity.

Reply on news.ycombinator.com