2004 Predictions, Revisited


Last year I wrote a fair share of 2003 reviews and predictions, and as the inevitable media reviews start lining up for year's end, I've started drafting a similar article for 2005.

It turned out, however, that cross-checking my predictions was so entertaining that it deserved a standalone post. So let's revisit my predictions, one by one -

Debian finally gets its act together and releases a decent, grandma-compatible installer with no traces whatsoever of dselect. Please!

Well, it sort of happened. Besides the usual anaconda stuff, we got Ubuntu, which is probably the best possible option (a Debian-based project that is easy to install, modern and without the self-serving Debian bureaucracy). Not bad for starters, even if it can't quite match up to Fedora.

I finally get a Fedora kernel to compile and run on a Cobalt box.

Nope. Not yet. And I'll probably only have time to try Gentoo next year. Still, it amazes me that there seems to be zero concerted effort to support old Cobalt hardware on any sort of distribution, given the amount of boxes out there.

KDE gets ported to Qt/Mac (thanks to RangerRick), and hordes of leetnix switchers will bombard us with almost-non-UI-guidelines-compliant ports of every single little KDE applet, which we'll gladly Fink to our Macs. Chaos ensues as we repeat the "choice is good" mantra and sort out the good stuff from the bad.

Base KDE stuff has been working fine on Mac OS X for months now on native Qt. Benjamin has been doing amazing work on it, but there isn't a single new cross-platform application running atop it, nor have there been any ports to speak of.

It looks like the Mac developer community is sticking to Cocoa and has no real interest in porting existing applications - which is odd, since a lot of newcomers to that community must come from UNIX backgrounds. Maybe they were simply assimilated and decided to learn Cocoa.

OpenOffice for the Mac becomes increasingly irrelevant, since the Aqua version keeps getting delayed.

Yep. Even though NeoOffice/J is making some progress, a completely native port of OpenOffice is still in the realm of Sci-Fi, and it's nowhere near the level of polish Office has.

We'll get yet another slightly incompatible flavour of Wi-Fi, and a lot of WISPs will either be bankrupt, bought out, or both. Consolidation will probably save Wi-Fi as a service, but carriers will add more security measures to the point where it looks pretty much like a cross between wideband, short-range GPRS (with SIM cards, subscription plans) and ADSL (logins, traffic limits, terms of service, etc.). Anyway, free Wi-Fi will still be around - but with limitations.

This one turned out to be a lot more complex that I could ever predict. The fact is that we are pretty much where we were last year technically:

Wi-Fi standards and chipset "features" became far too many to keep close track of during one single year. Everyone will point to 802.11g as being the only thing worth using at this point, but there is so much fragmentation on security (with WPA becoming a sensible add-on, but still not the solution) and fine-tuning (radio optimization for mixed b/g networks, etc.) that there are literally hundreds of different setups out there.

Add to this the impossibility of doing real-life deployments of SIM-based authentication (mainly due to the installed base of older access points and cards), and you have an incredibly messy environment that will take decades to sort out (people who remember the dawn of Ethernet and what AUI ports looked like will know what I'm driving at). Adding WiMax to the mix also helped further confuse matters, with a lot of companies splitting their efforts across both technologies.

As to the market itself, Portugal enjoyed a fair increase in the number of Wi-Fi service offerings, with lots of partnerships springing up and a few cross-market deals:

  • Competition sprang up (the Optimus/Novis/Clix service, for instance)
  • Vodafone Portugal further expanded their coverage.
  • Pretty much everyone implemented Vodafone's SMS-based authentication and billing method (nice to see my projects inspire our competition).
  • The incumbent operator - started tying together its own Wi-Fi service with its cable and ADSL offerings (by tacking on extra fees and/or promotions).
  • National (Oni/Vodafone) and international roaming agreements kicked in.

Most of this was overshadowed by the ruckus concerning the portuguese ADSL market, which was cornered by the incumbent until Sonae's Clix started offering 4 and 8Mbps traffic atop their own infrastructure. Still, the two markets are closely coupled, and I expect to see a lot more movement in both next year.

SCO and Eolas are eventually thrown out of court, despite the fact that there will be no real patent law reform in the US (heck, they've practically exported it to Europe this year, so why spoil things for the big corporate lobbies?).

Neither of them was actually thrown out - yet. Still, patent law came under much closer scrutiny everywhere, with a few irresponsible attitudes added to the mix. This does not bode well for next year.

Nokia comes out with more surprises just when people say their designs are starting to look jaded (I'd prefer they focused on getting the 6230 out soon, to give the T610 some real competition).

Bingo. Nobody (even if they had access to roadmaps and early previews) could have foreseen anything like the 7280, and some of the new designs make it look tame by comparison.

Push-To-Talk is heralded (and pushed) by all major carriers as the next-generation VoIP service that will drive up their revenues, until people realise that Europeans are much more polite and discreet than Americans on the phone and prefer MMS or SMS anyway. Vendors make a bundle selling it, but a lot less than previous years. Trials go on until the end of 2004 (and beyond). The US eventually go wild with it, causing yet another major cultural rift in mobile culture.

Well, this one was almost 100% correct on this side of the Atlantic. In fact, several analysts pointed out that Nextel (the operator most used as an example of Push-To-Talk's success) derived most of its success from a cunningly designed tariff plan and not from selling annoying phones. Add to this the cultural backlash against loud cellphone chatter in public locations, and you can understand why European operators aren't that keen on launching it wholesale.

The T630 becomes my phone of choice, despite the fact that it still has an extremely shitty QCIF camera.

Technology got ahead of me on this one. The Z1010 and the V800 quickly replaced my T610 until I got my hands on a Blackberry. Nothing even comes close to that at a professional level, even though the V800 gets a fair amount of use as a data modem and video telephony terminal.

I leave the technology field entirely and take up a position in corporate finances (just kidding).

Well, I won't ever kid about that again. The management course I took during the first half of this year made me realize that no matter how comfortable I feel in technology, it's time to move on and start doing other stuff

So besides changing my work focus towards more business-related aspects, I'm contemplating my options for next year (relocating to another country has been on the table for a while now).

All in all, the predictions turned out OK. From what I knew last year, I could never have foreseen stuff like IBM getting out of the PC business (but then, I wasn't looking), the fuss about the iPhone or Samsung rising in the ranks of handset manufacturers (I would have personally have betted on Sharp, who seem to be missing the boat this year).

I should, nevertheless, have foreseen the G5 iMac (if I thought it worth writing about - it was always pretty much of a given), Motorola's downfall and the successful launch of 3G services (if I had been able to write about that, of course).

We'll see next year.