Here are my steps to getting pictures into the “system”. Basically,

1. Save pictures to some folder on the desktop of whatever computer I happen to be using. Remove the obvious “bad” photos.
2. Add the library to my Gallery2 library that’s hosted by Dreamhost. Gallery2 is one of the best online photo-view apps I’ve found, and since it’s hosted on my Dreamhost account, I have unlimited space to put pictures (unlike Picasa Web Albums which caps you out 1GB)
3. However, Dreamhost is by no means a secure backup location (they don’t claim to be, and they recently lost a bunch of user data), so I don’t feel very safe having my photos only stored there. Therefore, after my photos are uploaded to Dreamhost, I rsync the photo directory onto Dropbox. Dropbox is a great online storage and file-syncing utility, so it doesn’t matter which computer I’m using — the Dropbox folder is always the same on all of them. When I rsync to my Dropbox folder, the files are automatically backed up online and synced to the other computers that have Dropbox installed. If you’re curious, here is the rsync command I use:
   
rsync --verbose --progress --stats \
--compress --rsh=/usr/bin/ssh \
--recursive --times --perms --links \
/source/directory ~/Dropbox/Photos/.

   Note: That’s a space between /source/directory [space] ~/Dropbox/Photos

   From here, all of my computers have Picasa installed (since it works on Linux, OSX and Windows), and the Dropbox photo directory is set as a “watched” directory. This means that as new pictures are synced to the Dropbox folders on other computers, Picasa on that computer picks it up and adds it automatically.

4. The final, and probably unnecessary step is rsyncing the files to Amazon S3 via Jungledisk. This doesn’t really offer anything in the way of additional backup security, since Dropbox is S3 based on the backend anyways (so if anything happens to S3, both will go down). The reason I do this is because I look at S3 as my long-term “vault” and Dropbox as a shorter-term, more convenient storage place. The rsync itself is the same command as above, but replace the Dropbox destination directory with your Jungledisk directory.

If you’re on Windows, you can replace the rsync steps with a straight copy — you’ll still get most of the functionality. Rsync works fine on OSX and Linux.

Clearly, it’s an involved process when I add pictures. However, it’s not too bad — a bulk upload onto Dreamhost and then running 2 commands from the command line to get the photos onto Dropbox and Amazon. You can even remove the Amazon step if you want. Last time I checked there was no way to add photos to Gallery2 using a script, but if and when that becomes available, this whole process can be turned into a quick script to run in the background.

Follow these steps, and you’ll have secure, accessible and useful photos on all of your computers.

The reason platforms are so important is that they allow developers to use massive technological leverage. To illustrate this point, lets use the example of Dropbox, an outstanding online backup application. Dropbox creates a folder on a user’s computer where they can drag files, and then stores those files on Amazon’s S3. S3 is a network that programmers can easily use to store files on Amazon’s servers. By simply paying a low fee per-gigabyte of disk space used, developers have access to an unlimited amount of secure, redundant disk space online on top of which they can build their own program. Dropbox is a derivative application built on top of the S3 platform, and utilizes the S3 backend to intelligently sync and back up files on individual PCs. Without S3, Dropbox and others would have to invest enormous resources into creating and maintaining a huge storage datacenter, which would almost certainly be prohibitive for the rapidly-growing startup. However, since they are able to leverage the S3 platform to handle the storage issue, the developers are able to focus on the product instead of the nitty gritty needed to get that product running.

The ability to abstract away the “nitty gritty” by using a platform is immensely powerful, because it allows entrepreneurs to build on top of existing technologies instead of reinventing the wheel for each venture. Facebook applications have been incredibly successful because they allow a developer to leverage the existing networks people have already built, rather than fighting to get people to sign up for a new website and build a new network. Before Facebook Photos, several sites attempted the tagging functionality that Facebook mastered with virtually no success because there was no way of enrolling every user that might be in someone’s photo. Since Facebook already has a huge portion of internet users, Photos (and other apps) can simply abstract away the “building a network” aspect — a huge obstacle to building a networking application — and focus on creating new and interesting functionality.

The other cool aspect that platform creators get “for free” is marketing. If a developer creates an application on Facebook, they are likely to spend time and money marketing that application. However, since that app is nothing without Facebook, any marketing that the app developer does is by default marketing for Facebook as well. While its not always this obvious, it’s always the case that the platform benefits as the child application grows — as Dropbox gains users, for example, Amazon naturally gets more business without really needing to do any extra marketing. It’s all done for them, and all owed to their killer platform.

Platforms are an amazing innovation, and we’ve seen (and will continue to see) some incredible applications arise from them. They lower the barrier to entry for entrepreneurs, so programs that used to require a team of engineers to handle “support” aspects of an application now only need one or two core developers. As a result, more products can come into the market quickly, and great ideas can rise to the top even more freely.