Back it up!

This post is from Dr Ali Daws and originally appeared on his most excellent blog, Writing on Writing. Ali is quite extraordinary, not the least for the fact that he is the only person I know to go and do a Masters degree AFTER he got his PhD ( just for interest – like you do).

This post is reposted from his blog by his permission because I thought the topic of backing up was both important and interesting (well, Ali makes it interesting). Enjoy. 

I will always remember one scene in Wonder Boys, even though the rest of the plot has become hazy for me over the years. In this scene, a tortured writer (played by Michael Douglas) watches as the hand-typed pages of his manuscript blow away. All twelve thousand of them. The thought is enough to send chills down my spine.

Just a plot device, I hear you say. It wouldn’t happen in the age of computers. Oh really?

What’s your backup plan?

Some of you may be thinking to yourselves, “but Word saves automatically, doesn’t it?” Others, “what’s a backup plan?” But there are some of you who know exactly what I’m talking about. The ones who’ve lost data.

I was busy in the lab one day writing my Honours thesis when the fire alarm went off. I assumed it was a drill. I kept on writing. That is, until the fire warden found me. He said the lab next door was on fire and told me to get the hell outside with everybody else. I stared at him, then at the ageing Apple Macintosh computer with all of my precious words painstakingly hammered into place with two fingers. (This was before I could touch-type.) Then I looked at the jars of extremely flammable fixative and solvents and God-only-knows-what-else lining the shelves. (This was also before occupational health and safety was given much credence.)

I can tell you one thing—Word’s auto-save feature didn’t give me much comfort on that day. I fought off the fire warden long enough to unplug the computer from the wall and disentangle it from various peripherals. Then I carried the damned thing downstairs in my arms.

That was when I started backing up my work religiously.

I developed an intricate system when I was working on my PhD thesis. It involved saving each chapter on its own floppy disk at the end of the day. I actually made two copies, and took one set of floppies home in a box so that when the lab inevitably exploded I wouldn’t lose everything. Some of you are no doubt smiling at the memory of little boxes of floppies we all used to carry around. For the rest of you, floppy disks are what we had before we burned things to CD or DVD or USB drives. You know that icon you hit to save your work? The square one?

That’s a floppy disk.

You’re welcome.

Today my backup schedule is a little different. For one thing, disk space is ridiculously cheap. Backing up no longer means saving each chapter to a different disk. But the biggest difference comes from the ingenious backup features built into Scrivener. Every time I close a project, Scrivener zips the whole thing up and saves it in a designated place. The Preferences pane gives a range of options for this kind of regular backup, including how many backups to keep (in my case, the ten most recent) and where to save the backups.

Scrivener also supports a version of off-site backups, which means that if my laptop suffers a calamity—possibly involving coffee, given my writing habits—I have a zipped copy of my work to restore onto whatever replaces my trusty MacBook Air. Once a day I go to the File menu and choose Back Up > Back Up To. This menu option lets me choose an external drive as the target for my backup instead of the folder on my laptop used by the automatic backups. I have a shared network drive at home. (editor’s note: if you want to read more about the word processing program scrivener, visit this earlier post)

You can also use something like Dropbox that hosts your files in the cloud somewhere. (Just be careful to read the terms and conditions of online file storage services—some include the right to copy and even sell your documents.) The advantage of these services is that you can access your files from anywhere in the world.

My backup strategy has changed over the years. When I adopted Scrivener for all of my writing it took a giant leap forward. If this post has made you think twice about the way you backup your own work, then it was worth the time it has taken me to write. Save yourself the trauma of losing everything. Backup your work.

Have you ever lost a bunch of data or writing in a digital catastrophe – or had a near miss? How do you manage version control between all your files? We’d love to hear some of your data management strategies!

Related posts

Don’t type format c:

Researching then and now

This entry was posted in Uncategorized. Bookmark the permalink.

33 thoughts on “Back it up!

  1. My sister was telling me middle of the year how the family’s lap top started flickering on the screen and one day died an ignominious death. I was all ears. I had not heard of computers failing spectacularly without a real good reason… like coffee spills, fire etc. I quizzed her thinking of my own lap top. About a month ago I got a small flicker top left of my screen. I dashed out and bought a 2nd hard drive. My other one is as big as a house brick, this one fits in my palm.
    I had just submitted my exegesis to the examiners so I started off loading everything to both hard drives and drop box. The flickerings were intermittent and did not bother me. About 2 weeks ago whilst writing my oral exam part the computer just shut down blank. I started up and all work was gone from desktop and illumination on screen turned to dark. I had been saving to my external hard drive which had not been affected. Eventually it got very regular, the shut down and flickering. I took to Uni and was told there was no problem. Home again and it kept happening. Back again and when I said the electric chord light now flicked from green to orange and back a lot, the tech said aha! My electrical ‘something’ has broken inside and is too expensive to be repaired.
    I borrowed a friends PC and kept preparing my oral exam. This week I opened the computer and all word documents had a virus blocking them from changes being made. I had copies on usb but my computer rage was high. I now own a new desktop and am still saving to 3 different places. I now think of my computer as a box that can type and do marvellous things but it is just a piece of plastic. At any moment it can go wrong so everything must be kept in copies.
    Ps Apple Mac book pro dropped by $249 this week! cheaper than the educational price. New product coming out I guess

  2. When I try to back up, I end up with 10-12 versions of my different sections. It becomes frustrating to search through them and try to remember which was the final version. I sometimes save 3 on the same day which increases my confusion.

    • You should be able to see the time you saved them as well as the date, which will give you the latest one painlessly. Or, sort them by date, and the most recent one will be at top or the bottom of the list.

      I back up from three computers and an ipad to the cloud, and also to our local time machine overnight. I have found that Sugarsync works well. The amount of space I need for my whole PhD – including Endnote library with articles, NVIVO files and many many word files (most in folders marked ‘old’!) – is still within the free allowance they provide. Wherever I am I can get the latest copy of whatever I’m working on.

    • When I was doing my editing course, I was taught to re-name every version with my initials and the next number.
      So thesischapter1.docx becomes, successively:
      thesischapter1_gw1.docx
      thesischapter1_gw2.docx
      thesischapter1_gw3.docx
      etc.
      I wish I had done that for my thesis, but I didn’t know about the trick.
      Every time I open the document, I immediately save it as the next number. You have no idea how many times this has saved me losing documents. If version 2 is okay but version 3 fails, you still have version 2 to work on.

  3. When I was doing my doctoral research, I was worried that a plane I was on might crash and I’d have to jump out if it without my computer–so I had my USB Port on a chain around my neck–just in case. If I died in a crash I didn’t really care–but the thought of surviving one WITHOUT my data was truly frightening. I made sure I always carried a recent version of my research with me in case my house burned down, in case of an earthquake, or in case of–well anything!

  4. I had three HDD failures during my PhD. I had backed up, but probably not as religiously as I should have. Thankfully I have IT friends who rescued me….otherwise I may have just thrown it in.

  5. I’ve always kept good backups of important documents. I was very happy to have these when I murdered my iBook by spilling the equivalent of a whole glass of Coke on the keyboard … it was dead, dead, dead! My backup allowed me to continue working on my thesis using my husbands PC for a few days until I got my replacement MacBook and upgraded all the software (I was running very old versions, intending not to change anything until the thesis was done).

    Today I use Apple’s Time Machine to automatically back up everything to an external drive. I also copy my thesis folder to a USB regularly, just in case.

    Even so, I have occasionally lost half a day’s work or so due to software failure or stupidity on my part, e.g. wanting to go back to an older version and accidentally overwriting something I actually wanted!?!

  6. Jeremy Stout, our associate online director here at AAU is fond of saying “your data is your tuition money made tangible”. Your work at school = your portfolio = the only way you can prove to an employer that you learned something. For you non visual thesis people it’s pretty much the same right? No papers = no proof that you did all the thinking!

    For our students that means their data is worth roughly 60-80K. Which is an incentive to treat it as the valuable resource it is…

    • I like this way of looking at it. In my case, time is my most precious (rare and with multiple demands) resource so perhaps I should work out my hourly rate to calculate the financial cost of my data! I’ve got better at version control and having backups in different locations but this is a timely post to remind me not to get complacent.

  7. I’ve been using Time Machine backup system on my Mac, which theoretically makes the transfer of ALL the data and program files to a new Mac seamlessly easy. However, my dear Mac fell off the desk during an earthquake and, although it looked just fine, it failed to complete a Time Machine backup after the fall. Little I knew that the mere attempt corrupted the backup drive, so when I tried to plug it into my new Mac – it didn’t work 😦 Turned out, the deeper backups were unaffected, so I could still rescue most of my data from that external HD. Take home note: do not rely exclusively on backup software; good old “copy-paste” is safer at times.

  8. One day I did something brilliant! I typed it up, printed it out and found my supervisor in the tea room. We were both very excited. A few days later I opened up the file, and the section where I had typed my fabulous idea was gone. After a while I realised that I had edited another version and saved over. My supervisor could not find the print out. I had to go back through my notebook and find the scribbled calculations all over again.

    Only cost me a few days work, but it was a enough to learn from. My problem was not backing up, but version control.

    I have a ‘live’ version on a usb stick. I then save to what ever computer I am using (home desktop, laptop, Uni desktop). I also have a back up usb stick. I carry the ‘live’ version in my wallet wherever I go.

    • I used to have a ‘live’ version of my honours thesis on a usb stick. I once overwrote it with an older file. Argh.

      Then I set up a Bazaar repository to keep track of all of the versions of my files, on the USB stick. Then I got lazy – I backed up onto my home computer and my uni computer, but I worked directly off the USB key. When the USB key carked it (from overuse?), I lost my day’s work.

      More recently I seem to have an almost identical setup to that which welf describes in his/her link below . I have Bazaar repositories on my home computer and my uni computer, and they synchronise via a copy on the heavily-backed-up university server.

  9. Not long after I started my PhD, one of the other buildings on campus caught fire. PhD students were rumoured to have lost up to 7 years of work (back in the day when a PhD could take as long as you liked). I backed up regularly to CD, which I took home with me.

    During my 2nd year, I went on a writing retreat with my supervisor and two collaborators on a project. We planned to work really hard, while simultaneously walking in the woods and collecting mushrooms and blueberries. Only my supervisor had a laptop with the software we needed, and fairly early in the week (after we had picked mushrooms and barbecued them), we settled down with a few small beers to start the working really hard part of the trip. A small bottle of beer got knocked over. Into the laptop. “But I haven’t backed up for a month!” wailed my supervisor. Fortunately, the hard disk was recoverable (although the motherboard was not). I backed up even more regularly.

    Now, I use Time Machine to back up hourly to an external hard drive, which sits in my office. I also sync my Mac with my MacBook on a daily basis, and take the MacBook home with me, so that should the University burn down overnight, I have a copy with me. If the fire alarm goes off, I try to take my MacBook with me… But there are still holes in my backup plan. What if the fire alarm goes off when I am not in my office? I have copies of some things in Dropbox, and an second hard drive at home with a back up from about two weeks ago…

    My PhD supervisor also stressed the importance of naming conventions. She suggested that each project should have a short, snappy title, which is the name of the folder in which all material for that project is kept. Then, when working on the manuscript, you give that a short snappy title too, such as SexyProject.doc. When you want to make major changes, or every so often, you rename SexyProject.doc to SexyProject1.doc, and continue working on SexyProject.doc. The next rename will be SexyProject2.doc. So SexyProject.doc is ALWAYS your ‘live’ version, and the numbered version with the highest number is the most recent previous version. It’s worked really well for me, and keeps my folders looking nice and tidy.

  10. Version control: Software engineers have had this solved for years.

    Perhaps the most popular program is Subversion which is easy to use (it works as a “versioned” folder that you “check in” and “check out” of a central repository), takes mere minutes to setup and (best of all!) FREE.

    And it’s easy enough that the marketing people at my day job use it on a daily basis.

    I use for for both my real job and my thesis. It sure beats labelling everything “this is the latest version….. no this one!”

  11. For those who mentioned version control, there is a companion piece (the second half of the blog post on my blog) about exactly that. How do you keep track of the myriad tiny changes you’ve made in your writing? This is particularly crucial during those final weeks (or months) of revision, where the changes are scattered throughout the manuscript like buckshot.

  12. Like most of the other posters here I have friends who have had catastrophic losses of data, one poor student in the first semester of my Masters had a hard disk failure and lost all of her work (no backup), another had a laptop stolen (again, no back up and she also lost a lot of company data ). I have a somewhat obsessive back up process, I have an Apple time capsule which dutifully copies my HDD every hour or so, that lives in my study under the desk, I have a external HDD I run SuperDuper on about once a month – I would recommend SuperDuper as a good backup program by the way, easy to use and reasonably quick, takes about 5 hours for the 250GB on my laptop. Then, to legislate for for the absolute worst case total loss of not just my laptop but also my house, I have a subscription with Backblaze which is a cloud based backup company, I port all of the data from my laptop to their site, similar to time capsule except it is in a cloud server, so if everything in my study were to disappear one day I could restore everything from that back up. Both Time Capsule and Backblaze run in the background so you barely even realise they are working which is also a nice feature.

    This could be seen possibly overkill I know, but for the cost of a couple of bits of hardware and a cloud subscription I sleep so much better, and even on those odd occasions where my laptop takes a bit longer to start up and my subconscious starts wondering if it is going to play nice today, I don’t get stressed at all.

    On the subject of version control I put either a revision code on the end of the file name, such as xxxxxx V1.docx, xxxxxx V2.docx etc if it is a short term document, longer term documents like my thesis which is running into years rather than days or weeks, get the save date as the version number concatenated into ddmmyy in place of the V number, yes, that does leave a lot of versions floating around the folder but at least I know how they all relate to one another. This comes from years working in the engineering profession where revision control is an obsession.

  13. I find the easiest way to back up writing when I’m out and about (without a USB or external hard drive) is to just email the files to myself. then the versions are automatically sorted by date and time in my inbox!

  14. I’ll briefly describe my own system, just in case others find it helpful.

    –My MacBook Pro backs up to a Time Capsule at home. So I (should) have a complete backup history going back several months. And a quick way to recover everything on my Mac if it were to crash. Thankfully, I haven’t had to test recovering much from that backup. And the Time Capsule doesn’t protect against theft or fire.

    –All my current research (and everything historical except for especially large data files) lives in a research folder within my Dropbox folder. So Dropbox maintains a current copy and roughly a month’s worth of historical copies in ‘the cloud’. My Dropbox folder synchronizes automatically from my main laptop to my 2 other machines, so everything is implicitly backed up to machines at home and at my work office (a bit of additional off-site backup, but without any versioning).

    –All my projects (subfolders within Dropbox/research) are checked in to a Subversion (SVN) hosted at school. That ensures I have some version history of all documents. And the server hosting SVN is backed up regularly. So everything of importance is available at another physical location.

    –I do all my writing in LaTeX, which is a plain-text format, so the SVN version history allows meaningful comparisons between historical versions. I’m in Computer Science, so it’s important to write in LaTeX. It might not be as practical in the humanities. An additional benefit is that the SVN repository is available to my collaborators as well, so we collaborate on writing and editing through SVN commits instead of emailed documents. You could accomplish the same thing with Google Docs, but this system works well for LaTeX editing. And the SVN history is obviously useful for code and scripts as well.

    –Most of my code is checked into a Google Code repository (our code is copyrighted, but open-source – if I worked on proprietary code, I’d have just stuck with the existing private SVN repository). I clone that repository regularly onto at least one machine other than my MacBook Pro, so I could theoretically recover all the version history even if Google Code disappeared on the same day my house burned down, although I that scenario is pretty unlikely. 😉

    The initial setup took a bit of doing, but wasn’t too onerous:
    –Move all projects into the Dropbox/research folder
    –Move code to Google Code
    –Import initial versions of all projects to SVN (from subfolders within Dropbox/research)

    The whole system is probably overkill just for backup (maybe even verging on paranoia), but the end result is that everything of importance is available at at least 3 physical locations, and I have version history going back several years. And once it was setup, it doesn’t really take any work to maintain – Dropbox synchronizes everything across several computers (and to the cloud), and I just commit any significant document changes to SVN and code changes to Google Code. Which is a useful practice anyway, even if I weren’t concerned about possible data loss.

    Hope that gives someone else a useful idea or two.

  15. I’ve lost data and as a result become paranoid about backing up.
    I used to do this using Windows Live Mesh to sync all my research files from my laptop to my uni pc, and to the cloud. That is, until Microsoft decided to mess with a perfectly good service.
    No:
    All my notes (onenote notebooks) are stored in Dropbox.
    I use scrivener to write, and it is backed up to Dropbox at the end of every work session. Every couple of weeks, I move older backups to an archive folder.
    I now use an external drive to sync files between my laptop and uni pc, using SyncToy, but I’ve also just discovered Synchredible, another free utility that does the same job. The external drive is always on me, just in case there are fires/floods at home and at uni at the same time 🙂
    I use a very clunky system of version control, but it works for me. when working with word or excel files, I save using todays date, title, and a version number plus notes (e.g. with expanded methods section) The version number only changes if there are major edits to the document, but basically every day I have a unique document name. In my research diary (onenote) I enter the filenames of files I’m working with each day, and what I did to them, in order to keep a track of what was changed when. It’s a matter of searching for the unique file name or just skimming through my records to see what changes were made, when, and which document to open to find them.

    It’s clunky, and difficult to explain, but it makes sense in my head and it works for me.

    You can never be too paranoid about backing up

  16. Pingback: How NOT to hand in your PhD | The Thesis Whisperer

  17. Pingback: researcher organise thyself | The Thesis Whisperer

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s