I’ve looked at clouds from both sides now
From up and down, and still somehow
It’s cloud illusions I recall
I really don’t know clouds at all
Joni Mitchell, Both sides now
When I first heard the term “cloud computing”, some years ago, I assumed it was a variant on an idea which had been briefly current. Someone tried to persuade organisations that the way to go was to have just one multi-user copy of each program (word processor, etc.) their staff needed, and then the computers on staff desks need be nothing more than cheap terminals. I assumed that they were going to try to persuade us that the world only needed one copy of each program. This seemed like not a very good idea, so I thought no more about it.
Now the cloud has descended into our offices and living rooms, and it is something different, more about storage than running programs. How can we best use it?
One thing that has been on my mind for quite a while is long-term storage of files. I have lots of old photos, documents, etc., which I would quite like to keep, not for general posterity but at least for my grandchildren. Archiving data is a problem for several reasons:
- The media wear out. No storage medium (even paper and ink) guarantees storage for ever. Often there is a trade-off between permanence and accessibility.
- The hardware becomes obsolete. A few years ago, expensive high-quality compact discs seemed the best option for file archiving. But my last several computers have not included CD slots.
- The software becomes obsolete. This is already happening with dvi, designed by Knuth to be a permanent unchangeable format; but it lacked the convenience of pdf, and dvi readers/printers are harder to find now.
So you have to keep “refreshing” the files and converting them to new formats when necessary.
Can the cloud help? It has its own strengths and weaknesses. The strengths include easy accessibility. But, although I haven’t yet had any files fade into the mist, even big computers can lose data, and even big organisations can disappear.
Perhaps the best answer to archiving is to store the data in as many different formats, and as many different places, as possible. Of course, this makes updating, and keeping copies synchronized, much more of a pain.
How to use the cloud?
There are many different places where you can store data. Among them:
- Dropbox gives new users 2 Gigabytes free storage; they have good facilities for sharing, either with the world or with a specified group of users, and the offer more space for introducing new users.
- Ubuntu One offer 5 Gigabytes. I signed up but haven’t yet explored the sharing facilities.
- WordPress offer about 3 Gigabytes. You could use this for storage if you were so inclined; you can upload “media”, and I don’t know what the restrictions on type of media or visibility are (if any).
- Then, of course, there is the arXiv, which (once you have been accepted) stores your preprints as you upload them. I suppose you could upload your back catalogue of preprints to the arXiv, but you had better stick to academic papers.
- Finally, academics are fortunate that our employers are usually fairly relaxed about what we put on the network. Once there, as with all the others, you can keep it private, and access it from anywhere.
None of these is enough for simply dumping all of my files. I have somewhere around 20 Gb on my laptop now, more than half of which is pictures. So I should specialise, and store particular things in different places in the cloud. I haven’t thought it through carefully yet; I don’t yet know whether there are good reasons for using one place rather than another. But I have started building a library of my own and other people’s papers on Ubuntu One.
Of course, it is not inconceivable that the Web itself will disappear – there have been several recent high-profile cyber-attacks. Maybe something different will have to take its place, and we will have to relocate our files again.