Okay, I’m almost done with this paper that I’ve been writing for a few weeks. Deadline is tomorrow morning, I think I’ll send it in later tonight. A whole lot of things to talk about though.
Why .Mac could have been great
I just ended my subscription to .Mac. No, I am not rich, and no, I am not crazy. When .Mac was first announced, sometime in 2002, it was claimed that it’d be very much integrated with OS X. I was a subscriber since day one. While this did happen (an API was released in May 2005, when 10.4 came out, if I recall correctly) not much more did. Developers didn’t embrace it. Charging $99/year for some web space, an email account, demo versions of some decent software (if you can find it), discounts on software you’d never touch (with a polestick) and — of course — the capability to sync stuff between your macs is definitely not worth it if you ask me.
So, I wonder. How come no one has written a Sync API compatible, syncing backend that makes use of, say, different storage backends, from the local network file server, to Amazon’s S3 or — really — any old WebDAV account to sync data? It’s not that hard and seeing that I don’t have .Mac anymore to sync stuff between my macs, I might just be tempted enough to write one. And now that I mentioned this, let’s revisit….
The web application nightmare
I was listening earlier today to vrypan’s podcast [in hellenic]. Ok, I lied, I didn’t listen to the whole thing, but I did read the shownotes. And I watched the video on Scrybe. Interesting. But come on! Surely this should
I am getting increasingly frustrated when people just jump on the ‘Ohhhh, Web 2.0’ bandwagon. Don’t get me wrong. I’ve got many computers that I use and I would like to share stuff between them. I — literally — spend all day, 7 days a week, 365/6 days a year in front of a computer, so I do care about application and data ubiquity. I do want to be able to check my email, my data, my photos, my everything from everywhere.
I just don’t think you should trust a web application that exploits a fashionable trend (AJAX), but is really a kludge, or — to be honest — a commercial online service that may go bankrupt sooner than you can say ‘bubble’ with your productivity tools, or, indeed, your data.
What I would much rather prefer to see is a configurable, free, .Mac-type syncing service on steroids. A service that syncs data. That’s what I need. Not a full fledged web application. But a full fledged, configurable syncing service. I want to point it to my WebDAV server in Texas. Or my home ADSL in Athens. Or my friend’s ADSL in London. Or my company’s server in Aachen. Or whatever. And then I’d like Open Standards. Everywhere. That’s not a pipe dream. We have it for the most part. XHTML, JPG, PNG, MP3, OGGs,…OTF. It’s all there. We just lack the syncing services built into our software. Apple did it. But they kept it under lock and key, ’cause they thought they’d rather keep on doing what they’ve been doing since 1984: lock users down to their ( in this case paid and miserable) service.
Once we’ve had a good syncing service in place we’d just need the desktop applications to leverage it. Good, clean, managed, easy to develop, gorgeous applications. ‘Rich’ in ways ‘Web 2.0’ can only — possibly — dream of its distant successors having: animation, sound, 3D, shadows, ‘intelligence’. The applications development industry, with the exception of the Mac and some Linux programs, has stagnated. Windows (and linux) makes your life hard, while the CPU gauge of your brand new Core 2 Duo is at 1%. Strike that, 0.5%. Yet Microsoft has been innovating with its .Net platform, the CLR, C#. How is everyone else answering Microsoft’s future roadmap? With Web 2.0? Why is an inferior choice like malfunctioning, temperamental and plain slow Web 2.0 application considered to be the solution?
Do we really want net applications? Yes. Well, sit down and write a proper standard for them instead of abusing XHTML. (If you think this is a rant you’re probably right. If you think it bears some relation to the ‘Lost Priorities’ article, you’re probably also right).
Storage
In his podcast (see above) vrypan mentioned Amazon’s S3 as a potentially good choice for backing up stuff. I disagree. Burning a DVD+R is faster, cheaper and safer than uploading data to Amazon’s servers. Sure, it’s probably a bit more of a hassle, but I don’t think the odds are on Amazon’s side here. In fact, the problems with storage are great and, unfortunately, unknown to most people. Our rate of generating content is much higher than our rate of improving the performance of our storage media. In addition to this, our storage media is rapidly going to reach the point where it will start failing. U.S. Agencies have already started experiencing some of the problems associated with magnetic storage. And most people are blissfully ignoring the fact that for the first time we do not have a storage medium that is going to last for centuries, although we generate more information in a year than people did in centuries a while back. There was a very interesting article on this on IEEE Spectrum (if I recall correctly) from a couple of years ago. I’ll try to find it sometime.
Anyway, that’s the end of my rant. Gotta go back to writing about Prolog, DAI ‘and stuff’.
P.S.: Oh, by the way, cosmix.org is 3 years old! Yes, I know you’ve noticed. 🙂