It’s been said that those with the good fortune to survive near death experiences often take a step back and actively reconsider the direction of their life. It’s also been said that there are two types of people in the world: those who back up their data and those who have yet to lose a hard drive. The first I can’t speak to – a near miss with an 18 wheeler up by Queechy Lake years ago aside – but the second, at least in my experience, is accurate.
Please note that I am in no way trying to equate the death of a person with the loss of a hard drive. I’m merely trying to point out that the consideration of finality, whatever the context, can often inspire some thoughtful introspection. A rethinking of where you’ve been, where you’re going, and how you you plan on getting there.
So it has been in my case. With my primary computing platform, that is, not my life. Mark apparently did not require such a dramatic wake up call, but then I’m not as sharp as Mark. As has been previously documented, approximately one month ago I managed to corrupt the primary partition of my primary desktop operating system, which was Linux. Gentoo Linux, to be specific. Since that time, I’ve been forced to eke out a meager existence on top of a storage challenged backup partition of Windows XP. The loss of the data, and my subsequent experiences on top of Windows have been as instructive as they have painful. The replacement hard drive has in fact arrived (along with an unexpected but very welcome replacement machine), but the lessons of this little episode should linger for a while.
My epiphany, such as it is, was simple: I no longer want to have to care about my operating system. For those of you short on appreciation for subtlety, the italics were for you: the ‘have to’ modifier is the most important part of that statement. I may choose to care about certain aspects of the underlying operating system – be that compatibility with a specific device, love of a vendor or project, belief in free as in freedom offerings, or the way it makes you feel – but I want no part of a platform that rigidly requires me to devote more than a modicum of attention to it.
So the good news is that I’ve figured out where I want to be. The bad news is that I have a ways to go before I get there. If my month long tenancy on Windows has taught me nothing else, it’s that I’m not particularly operating system agnostic. I continually find myself seeking out Windows replacements for the tools I’m familiar with. Huge portions of my data are tied up in proprietary applications and proprietary formats. And frankly, I expect that will continue to be the case – at least for the forseeable future.  Much as I’d like to subsist entirely in the network cloud, I’m not there yet and don’t expect to be soon. For those that do like to think of yourselves as relatively operating sytem agnostic, and argue that your reliance on thin clients makes operating systems irrelevant, I suggest a simple exercise. Replicate my experience by destroying the hard drive of your primary machine. And don’t think to cheat by backing everything up before you do if that’s not your usual habit (unlike Sandy, who is a backup zealot) – I’ll know if you do. I think you’ll discover that you are in fact dependent on a particular operating system, as much you might not like that fact. Now before the SunRay people jump in and claim that this would be no problem to them, let me say that if my primary system was a SunRay you would not be reading these words right now. They’re coming to you from a train with no connectivity. There’s always a price.
As for me, I’m content to throw total operating system independence into the out-of-sight, out-of-mind “strategic” bucket. I’m headed there in the long term, but am planning for a near term without it. What’s my tactical plan in the meantime to improve my platform flexibility? There are several different components to the strategy, and some might be of interest to you:
- Embrace Cross-Platform Applications:
In our podcast discussing the recent Microsoft/ODF news, I speculated briefly on the possibility of cross-platform application availability becoming a differentiator in the wake of the seemingly inevitable support for both ODF and PDF within Microsoft Office. While I wouldn’t care to generalize to far on that point, I can say that two of my most used applications – Firefox and OpenOffice.org – are cross platform and have followed me to Windows from Linux. I have addictions that I’ll need to ween myself from – hello, iTunes – but I’m growing more and more reluctant to commit dollars, time or both to applications and their ecosystems which would choose an operating platform for me.
- Open Data & Open Standards:In a perfect world, all of my important and even non-critical data would be wrapped up in a nicely documented open standard that supports multiple implementations across a variety of operating systems. These open standards would permit me what Simon terms the Freedom to Leave, Jon Udell earlier termed an Exit Strategy, and what I’ll simply refer to as the logical counterpart to my Barriers to Entry meme – low Barriers to Exit.
Unfortunately, it’s not a perfect world, and competiton via exclusive ownership of a format is still considered a viable strategic approach in some circles. This, in a manner of speaking, is what’s behind the recent defections from OS X to Ubuntu of Mark and Cory. I don’t yet have an accurate read on whether or not their concerns are canary-in-a-coal-mine-ish or the fringe, but I do know that I personally share them. Whether it’s my email, my documents or even the wiki we’ll be running off internally, my primary goal for information storage is to be as loosely coupled to my operating system as possible. Applications too.
In other words, I want my data divorced – forcibly, if necessary – from the tools and platform they run on. While your average consumer might not care to choose different applications or another operating system right now, they might want to in future. And I did yesterday.
- Software as a Service:
The shortest distance between an application and cross-platform support is not, IMO, VMs like Java’s or Mono’s or multi-platform GUI toolkits such as GTK or QT, but the approach variously described by terms such as Software-as-a-Service or web based software. I do not happen to agree with Luis’ anonymous contact from GUADEC who claimed that ‘open source is doomed’ because of the steady migration of applications to the web, but I believe do believe the objection is sustained rather than overruled. As Havoc, Luis and the others plan the future of GNOME – and other vendors and projects ponder the same question – I’d offer only this sage bit of wisdom: the network and the applications that run on it are the future. As Cote put it well yesterday (and Ian helpfully amplified):
What most desktop applications lack now-a-days are features that are fully web-enabled, in a bi-directional sense.
The operating system / desktop environment that figure out how best to support and compliment them will be in good shape. It’s not a zero sum game, in my view: both will have a place. But they need to figure out how to work better together. Operating systems are still acting as if network connectivity is a condition that is indefinite and uninterruptible – where I’d definie it as the opposite. I’ll get into what this means from an OS perspective momentarily, but for a hint think on the next bullet.
My colleague has astutely pointed to the next logical evolutionary stage of web applications being synchronization, therein coining the term “synchronized web.” While I think we still have aspects of that to parse I think he’s got a point. It would appear that Google thinks so too. The initial indication was the Google Browser Sync extension for Firefox that propagates one’s bookmarks, cookies, and so on around to any machine (importantly) that will run Firefox 1.5. But that was but a mere prelude, if the Google Operating System blog is to be believed, to their followup act: full network attached storage, with a variety of value add services such as collaboration and – fittingly – synchronization. Some users are probably noting that Google Desktop already provides some of these file sharing capabilities, but here’s the important difference: according to that entry, GDrive will be cross-platform unlike the current Google Desktop.
Assuming the speculation vis a vis GDrive is true, what does that mean? Well, it would make the operating system a little less relevant for me – particularly if I can get a substantial portion of my assets converted over to open formats. If I had that sort of capability now, I simply could have ordered a new drive, imaged it with Gentoo, and been off to the races with Gdrive dropping most of my data back in.
There you go: one man’s plan for reducing his overall operating system depedence. While there are more, I think it’s time to be moving on before I lose the few of you still with me. So let’s take a minute to see if we can distill out of those macro-level concerns any suggestions – requirements, if you will – for those would contend for the title of next generation operating system or desktop environment.
If I do adhere to the above plan, and I intend to at least try, it’s clear that my needs from an operating system are not
terribly extensive. Here’s what I’m thinking would be nice from my future operating system:
- Be easy on the eyes (translated: pretty is a feature)
- Stay the hell out of my way (translated: don’t take 5 minutes to recover from suspend, don’t crash, don’t get infected with malware, don’t let applications suffering from memory leaks lock up my desktop)
- Make my life easier (translated: don’t make me go anywhere to install applications, don’t make me become an admin, and anticipate things so that I don’t have to)
- Just work (translated: support most hardware out of the box, think of the little things for me)
- Be popular (translated: if you’re popular, the software/drivers/etc availability will take care of itself)
- Work with the network, not against it (translated: design for intermittent and multi-speed access, build aplications and platforms that are network aware at the lowest level, understand that I may use Flickr and Gmail as much or more than any client application, and design for multi-device connectivity and sync)
Again, this is one man’s list, and many of you will have different priorities for your platforms. But I think it’s likely that some or many of you would agree with a point here or a point there, if not more, as we’ve all been frustrated users.
 I do find it interesting, however, that I seem to care about applications a lot less than do folks like Bray, de hÓra, Leung, or Pilgrim. Where most of them have a small but critical list of thick client applications that they can’t do without, with the exception of an addiction here or there (damn you, iTunes) I’m not terribly concerned with application availability, because most of my day is spent in a browser.