Data safety

In his “Fast Forward” column in the Business section of today’s Washington Post, Rob Pegoraro discusses the Google service interruption that I wrote about here earlier.  He points out, quite correctly in my view, that people can become fixated on outages of Web-based services (such as Google Mail, or Amazon’s S3), but that there is another danger to the health and well-being of users’ data, stored on their own PCs,  that they often don’t think about.

That risk is  having your data stored in an undocumented, proprietary file format, such as the format used by Microsoft Outlook to store E-mail:

Using a format that one company keeps to itself, at worst, can prevent you from ever taking your business elsewhere. More often, you’ll have to sit through prolonged and complicated file-export procedures to get your data into another program.

Being “locked in” to a particular company’s  proprietary software is bad enough.  As Pegoraro points out, even if it’s possible to move your data to another format or program, it is usually a painful process.  Even worse, the vendor of the software you are using might discontinue support for the product, discontinue the product altogether, or just go out of business.  (How are those old XyWrite files working out for you?)  I have seen this happen with both individuals and businesses.  In one case, the vendor of the accounting software that the organization was using went out of existence on a few days notice.  One could say that the users should have been paying attention to the health of their suppliers, but that is to some extent beside the point.  The only alternative they had was to get someone to reverse-engineer the proprietary file formats, and then manually fix up any errors from their paper archives — which they fortunately still had.  In a follow-up post to the column on his “Faster Forward” blog, Rob Pegoraro says he has concluded that this is a bigger risk than the risk of outages of Web-based services:

On reflection, I decided that the second problem is worse. Companies that don’t have death wishes usually get their Web services back up after an outage pretty quickly, but some of the closed formats I mention in the story have been around for a decade or more and show no signs of being opened.

I agree, although he is surely right, too, that each user has to evaluate these risks in light of his or her own situation,.

There’s another aspect of using Web services, such as Google Mail, that I’d like to introduce to the discussion.  And, for a change, there is actually some good news on offer.  Google Mail can, of course, be used entirely as a Web-based E-mail service, accessed via your browser on Google’s servers.  But it is also possible, at no extra charge from Google, to use Google Mail in conjunction with a PC- ot Mac-based E-mail client (such as Mozilla Thunderbird).  Assuming that the client you use stores mail in an open, documented format, you avoid lock-in, and still have your E-mail history available to you if the Web service is temporarily unavailable.  This kind of setup also gives you one important benefit, compared to the PC-client-only solution, where all data is stored on your PC: you automatically have an off-site backup of your E-mail history, sitting on Google’s servers.  And if I am reporting honestly, I have to say that I have seen far more data loss attributable to a lack of a sensible (or any) backup procedure than to failure of a Web-based service.