… until someone got tired of paying for the server and turned it off
… until a hard drive crashed or corrupted data and holes in your backup strategy were revealed
… but your friends never used it because it was hard to share things, the desktop app was hard to install, etc.
The cloud model is different. It has pros and cons, and nothing of value will come from starting a conversation on the false impression that it's universally worse but people somehow don't see that.
* If the software stops being supported and isn't patched, you can still run the old version; you could even patch it yourself.
* If someone turns off the server, you migrate to your own. (You're keeping backups, right?)
* If the hard drive crashes, you replace it. (You ARE keeping backups, right?)
* If your friends don't use it, you can try to improve the software, lobby harder, switch to something else, etc.
None of these things are easy, but it's at least possible - and at the very least, you don't suffer the same sort of emotional ennui due to it being completely out of your hands.
When working with the cloud people just assume it will be there forever as long as you pay (or someone pays in a freemium model) money.
Well that's something of the past. We have things like Apple Time Machine now.
This is also why I like apps like Day One which can operate with purely client-side apps storing their data in a user's Dropbox folder. There are a lot of limitations to that approach, but benefits too.
In my experience, that window is typically one or two years, and the longest I remember it ever going is 6 years.
Given that they've committed to keeping Panoramio supported for 2 years after first announcing it was done and yet another year for migrating if you couldn't be bothered to in the first two years, it seems to me like Panoramio has outlived a typical software download's window of life.
I think I've been using gmail and flickr longer than any media based software or download I've ever had.
Most regular non-technical computer users can't use the software distribution model of emacs for local applications. Hell, I'd probably be safe saying most programmers can't do it either, let alone use emacs.
A better comparison might be Gallery[0], which I've been using for 13 years. They stopped development in 2014, but tons of people still use it.
Certainly some data is more valuable than other. Is a hundred IT labor hours cheap or expensive compared to all pieces of data? Who knows best, the end user or big brother?
I'm not sure what I'd do with data from Panoramio. I certainly don't have a local application on my computer that can read or edit that data.
In general, most cloud services I've seen shut down do the same, often on tighter deadlines. While it may have happened at some point, I've never seen happen what you suggest. I've never seen a cloud service makes a judgement call about how valuable someone's data is, and not allow it out when the service is EOLd.
But as a 2 year old article on this very topic put it: "The decision shows that communities ultimately are the property of the companies that run them, not the members who supply their content." - 'Panoramio founders try persuading Google not to kill site', CNET, 24 Sep 2014 (https://www.cnet.com/news/panoramio-founders-try-persuading-...)
Of course, I've had to purchase new versions of any applications I wanted to keep running, if the new versions were available, which in some cases weren't. I was pretty sad when my outdated version of Shake would no longer run on OSX.
It certainly shouldn't go without being pointed out that we are comparing apples to oranges. Cloud services on the whole gereally provide different values and different kinds of features than local apps.