This is actually one of the ironic casualties of the cloud. In the old days you bought your packaged software and it was pretty much guaranteed to work as long as you didn't change your setup (OS, hardware, etc). With cloud services the change is out of your control, so there's nothing you can do to freeze your setup.

This is probably a reason why there may be a growing market for personal clouds.

And this led to a lot of other problems, for example, you couldn't keep up with new technological innovations because you needed older systems for some programs.

It's an interesting question, which practice is better.

A boss at a job I worked at a few years ago still keeps his contacts in an ancient version of Lotus organizer without VCard export support. As a result he's pretty much locked into it unless he ponies up $100 for the latest version of the software (released in 2003) which has VCard export. He's too cheap to do that so he'll have to stick with it as long as he is still on a version of Windows that supports 16bit software.

http://theopenphotoproject.org or https://openphoto.me or https://github.com/photo

Disclaimer, I'm the lead dev

Without diving into the source, my primary complaint was that the whole thing felt unpolished. There were display errors that just needed some TLC, documentation that sorely needed the eyes of a proficient English proofreader, and issues that made me wonder if there were going to be underlying security concerns. (For example, sometimes the path in the URL would be encoded, sometimes not... just makes you wonder what's lurking under there.)

[Sparrow](http://www.sparrowmailapp.com/) is a great example. They were acquired soon after the release of their iPhone app. That kinda soured me and many others.

EDIT: The same could be done for other cases, obviously. Take usage rights on social photo services as a completely hypothetical example. :)

EDIT2: In fact, it could be something YC requires their funded companies to do. That must be pretty popular considering all the Instagram/Flickr debacle.

When was it that Y Combinator decided to ban companies supporting SOPA/PIPA or something to that effect?

Imagine running a cloud service, where you make $10/month/user, and having to read through a separate contract for each person. Does that make sense?

The reality is, you have limited control over a product that you've paid $5, $50, or even $500 for. At most you can get a refund if you're unhappy with the service you received. You cannot force the team of developers to continue work on something, unless you're paying the millions of dollars that that costs.

And it's not difficult to see how adopting it can have benefits: http://blog.flickr.net/en/2011/05/13/at-flickr-your-photos-a....

A lot of people are abandoning Instagram not because of the ToS per se, but because they were reminded that it is now owned by Facebook. And Facebook's reputation is horrible.

Look at what Tumblr are doing by posting their policies on GitHub, which lets everyone see the diffs: https://github.com/tumblr/policy.

Goodwill doesn't have to be legally tethered. But it can be.

Your own.

Supposing that they don't butcher the migration, what Snapjoy service do you love, which wouldn't benefit from a close integration with Dropbox?

My only beef with Dropbox is that the gap between free users and paid customers is too wide, price wise. I understand that customers have to subsidize the hordes of freeloaders, but $120 a year for 100GB doesn't feel right. Maybe they shouldn't try to be profitable yet.

Other than that, they've successfully sneaked a usable file system in iOS, they support all my devices including Linux, they "just work" unless you try to share writing rights, and they're going to get us rid of the dreadful iTunes, iPhoto and no doubt one day iBook/Kindle. That's super impressive.

* either lower the entry price and limit the features of the free version if they need to be profitable now, getting more paying customers with a lower ARPU;

* or lower the entry price a lot, lose money and increase both free and paying market-share, if market-share is their current priority.

But the current gap from free to paying is too damn high IMO.

Great question. I think in this situation, a company like Dropbox is probably safer, as they're a lot bigger in certain measurements (amount of funding, amount of revenue, etc).

I think the companies that have the most to lose are the safest to invest time and money in using. A small startup that just started 3 weeks ago is more likely to be sold than a company like Dropbox, in my opinion.

I agree with the rest of your post but what you said above doesn't make sense to me. A 3 week startup is probably more likely to die than be bought whereas Dropbox would likely be a fantastic purchase (for anyone who could afford them). The aim of such Startups (at least if they have VC investors) is to sell the company at some point, whether that be to the public (IPO) or another company (acquisition).

Exactly. My point is that a 3 week old startup is more likely to disappear quickly (fail, get bought, etc.) than a company like Dropbox. Dropbox is a good acquisition target, but I think a company like that would take longer to do so.

As you said, a company like Dropbox could also go for an IPO, in which case, theoretically, they still provide the product that you're paying them for.

I think there will always be early adopters to help startups get their initial traction.

Even in that case, you will still find that no company keeps their products or services static over decades. In most cases you probably wouldn't want them to. Imagine how it would feel to be stuck on Windows 95.