On Mark's un-Switching, and Lemmings

Now that most of the fuss about Mark Pilgrim's un-switching has died down and that he's made it abundantly clear that he wants to keep his data usable for the next half a century, here's my take on it.

My Life is not in

I have very similar views to Mark's where it regards my data and proprietary formats. In fact, I've never relied on any of the applications since I had a .

  • I don't do movies. But the current H.264 Open Source efforts seem to be mature enough to encode stuff with, and thanks to the recent boom in video sites there is a lot of interest in editing tools, so if I get around to it, I'll probably have no trouble.
  • I don't trust with my photos. I never did. I just use it to import, rotate/crop/apply minor tweaks, and then cull the bad ones and . Tagging and organizing them is not a priority, but I believe in storing my metadata in my files and know my way around .
  • is just a (very good) front-end to a DAAP server, as well as a "good enough" ID3 tagging tool. For CDs that aren't in CDDB, all it takes is drag and drop and some patience.
  • All my mail is in an IMAP store - and that includes all my finished documents, which means I only have to index one thing. Storing them as attachments implies some overhead, obviously, but disk space is cheap and the extra metadata from mail headers comes in handy more often than not.

I'm not too concerned about documents. Those I write for work are stored at work (and ensuring those are future-proof is not entirely my own concern, although I make a point of generating PDFs and HTML versions of nearly everything), and those I create for personal purposes are few and far between. Other than , I can replace with just about anything at a moment's notice.

In fact, I've taken to doing stuff in HTML and render it as PDF, very much like A List Apart but with less panache.

I've also dabbled in doing some stuff in XML and Flash for generating presentations (instead of S5), but that went nowhere for the simple reason that I don't do presentations off work - even if I did, delivering them usually entails using someone else's computer, and that in turn pretty much mandates the use of .

Maybe I Should...

But let's take the who would follow in Mark's footsteps head on, shall we?

Remember my ? Well, I've been sorely tempted to make it real of late, both due to 's amazing unreliability and to the fact that I find myself wanting to go fully location-agnostic - i.e., what I really want is a secure server with a very fast and efficient remote desktop protocol and an ultra-slim laptop that I can carry around anywhere.

I don't want to configure my applications more than once. I want to leave a location, travel, arrive at another place and have them pop up on my screen as I left them. I want to leave my laptop behind and connect from another machine, anytime, anywhere.

Yes, I want a cheap, mobile, lightweight Sun Ray (I've ranted about this at length often enough), and the ability to hook up from whatever computer I have to use. Borrowing a Sun motto, I want to have the network be my computer, but with a desktop environment instead of that crufty Web 2.0 junk.

I can already do that with for my work. Now I want to do it with the rest of my computing life.

As it happens, I am making it happen now, using , Gnome and an UMTS connection from my . And it rocks - I am typing this in the Gnome editor inside a 16-bit desktop, and it's more than fast enough to keep up, even inside an wrapper.

It's doable. The tools are everywhere, and it has been inside my reach for years, even before UMTS access became ubiquituous.

So what's the hangup?

Maybe I Shouldn't...

The fundamental thing keeping me back is that provides such a damn good desktop experience that I can't see myself letting go of it anytime soon. No matter how good Gnome is right now, I still find myself grappling with the usability issues (and eyesores) of Gnome (or worse, KDE) applications on a regular basis.

It's not just the aesthetical aspects, it's also the integration (or lack of) between the applications themselves and the operating system, as well as the ad-hoc nature of what passes for UI consistency in those environments.

Plus there is no shortage of things that I simply love on that add up to the long list of nearly intangible niceties that make up the experience - and most of them happen to be free, built in to the OS, or tremendously effective through leveraging -specific features.

Things like (which is, for me, the quintessential experience), , (and its great Address Book integration), VoodooPad Lite, , you name it. There's no shortage of useful stuff out there.

And they all just work. There are some particularly annoying repeat offenders like , mind you, but the truth is that there are very few (or Windows) applications that match the average degree of quality and polish of software.

As to development (and brushing aside for the moment) not only can I do everything I want under , but I can also do it without any fuss.

There is the occasional trouble when trying to compile something that is badly packaged or that doesn't take portability into account (as a lot of -centric Open Source projects happen to do), but I end up chasing about the same number of dependencies and obscure libraries as I used to do under .

So yes, I find it to be a superior environment for and web development. And whenever I step into Cocoa for a bit, my mind expands just that little bit more.

For me, ignoring 's peculiarities is easy - there are just as many (or worse) commercial tie-ins and restrictions in software, and I'm not a zealot by any means.

Hardware

Choosing your own hardware, now that's something that struck home, given my interest in getting hold of a when the kinks are ironed out. It's an investment in technology, and I am extremely picky about any sort of investment that significantly devalues after less than a year (as all computers do), so I've been looking wistfully at "regular" laptops as well.

Ignoring the usual annoyances associated with being a user outside the US, I find the (oft-cited) argument of hardware being too expensive to be complete nonsense. I can't find a decent laptop to match the on both features and price, and especially not when I throw in the degree of hardware/software integration only can deliver right now.

So although I've been tinkering with the idea of picking up a Toshiba R100 secondhand (which would be more than enough to make my Network Computing vision come true), the only reason I even consider that option (besides liking that particular machine) is that I know my way around Toshiba hardware.

Besides, I can't help but feel that Mark's timing was a bit off. Sure, once installed doesn't go wonky like (unless you're the adventurous kernel-upgrading type), but every time I install on a laptop I feel like I was back in the early 386BSD days, digging up obscure patches just to get it to run, deal with power management, etc., etc.

It's a pain in the ass, and even hardware standartization hasn't helped much (how many of you have a laptop with an Intel or Broadcom Wi-Fi chipset?).

All The Best, Mark.

Still, he's sure to be able to pull it off successfully. For starters, he's bought a desktop, which is sure to go a long way towards lessening any hardware-related issues.

But, honestly, anyone with an ICBM meta-tag on his site and an X-Status: It's hard to find a new hobby. on his headers is sure to not only know his way around technology, but also enjoy tinkering about with until it runs flawlessly on his new Lenovo.

Let's just keep in mind that it was his personal choice and stop annoying him, shall we?

This page is referenced in: