Mobile thin clients.

Inspired by Martin again, I picked up a thread of thought he abandoned the real fun began: couldn't we look at mobile devices as thin clients?
He predicts that the shift is bound to happen as voice traffic gets wrapped in the communication logic that is bound to get implemented in the coming years.
Beyong the obvious issues of network management that he does get into, there's the whole problem of, you know, fitness for a particular purpose. And beyond the fact that mobile voice/video devices will always need some kind of internal codec logic, there's a good bit of thinking to be done about computational versus network: I'm reminded of it everytime I see satnav software struggling to re-optimize a route on an anemic smartphone/converged device. At the root of the issue the speed/latency/energy-expenditure compromise that needs to be decided upon on a per-application basis.
Why can't my desktop provide RPC services to my smartphones?

Other aspects:

• Application control: Network ops and service providers would probably love the mobile thin client paradign, allowing them to control what the users can run, and at what price, but the customers will push for the thick, all-you-can-eat computational buffet that they've gotten used to on the desktop. Desktop appliances didn't fail without reason.
• And if your telco ever smells the fresh application-provider meat, it'll never accept to be reduced to a dumb-pipe provider, at all, ever.
• The usual worry about DRM, software renewability, and the ability to run your apps on hardware you own. The latest TiVo debacle shone light on that issue already, and the TiVo appliances are not that thin yet.
• Location-dependant prices for location-independent services. While I'd love to have my smartphone to be sync'ed to a web service automatically, I don't think I'd like being charged extra for my calendering apps when roaming abroad.
  

Webloog as a Blogger.com alternative

On the advice of a roommate, I've given Webloog a try for the turn-key publishing solution I use for this blog. I've since stopped, because it's garbage. Well, not really, but it's not quite up to par with Blogger's reliability and ease of use. Here's another fancy list, because we really never can have enough of those around:

Webloog is cool because:

• It has categories, Blogger doesn't.
  
• It doesn't impose that omnipresent bar at the top of your blog
• It has niftier themes, although the colors could be better.

Blogger is much better because:

• The WYSIWYG editor actually works.
• Spellchecker.
• Quick-preview.

Now go play outside.

Google Talk a bit more in-depth

After using Google Talk for a bit, here's the stuff I can come up with as to how it compares with other VoIP apps, and how it fares in general. Mossberg has (as usual) a very nice review in layman's term, so I'll stick to the more in-depth stuff.

Good Stuff

• Compatible text-IM protocol.
  
• Quiet, simple interface.
  
• No obnoxious smileys (looking at you MSN)
  
• No fancy text formatting (Don't you love getting 12points Comic Sans MS from your teenage sister? Thought not.). Low key sounds.
  
• Doesn't desperately try to look like a phone (Skype's bad for that)

Bad Stuff

• No client for alternative OS'es (Skype has those)
  
• Voice protocol is closed.
  
• Doesn't use native widget sets (Don't get me started on Windows XP Themes management, but from a usability standpoint, it's a bit annoying, considering the rest of the app is pretty flawless)
  
• No gateway to analog networks (They need G-Money setup before they can setup paid services on G-Talk)
  
• Doesn't use the system default mailer to send emails, even though there's an option for that.
  
• Multiple open chat windows get a weird system to switch between your contacts, where both the top and bottom of the window contains active contacts. Tabs are a much more discoverable, consistent solution.
  
• Not compatible with Ultramon (the Multi-monitor management system).
  

In short: it's too bad about the lack of cross-platform voice and the non-standard look. But we'll live.

More on content management.

Martin "I'll make your network feel stupid" Geddes has a pretty thought-provoking post about DRM and the non-financial costs of finding digital music retailers.
He (basically) points out that finding non-restricted music has very high non-financial (namely search) costs, and besides a couple of good points, he indicates that the newbie's guide to DRM by the EFF actually decreases the non-financial costs of DRM'ed content. This is only partly true, because:

• most elements in the EFF guides are restrictions, not features (decreasing perceived value)
  
• the EFF guide points to DRM-free alternatives (decreasing their search cost)

So the net effect could swing either way, really.

In the ensuing chat, he pointed out that DRM could also be used as a market differenciator for feature-conscious consumers, in a manner similar that what airline passengers are used to. Again I don't agree, because there is no infrastructure upkeep problem in the digital music market, and the marginal cost of a digital tune is about nil. As Martin put it, there are lots of buts (such as the actual availability of unencumbered alternatives), but in the digital market, we might all end up flying first class.

As an aside, there might be intermediation money to be made in this market as well.

Scobblandora

The resident Microsoft-does-not-suck blogger just tried Pandora radio (a super webapp that'll feed you music based on a band you tell it about) and fed by his conviction that thick apps are coming back, it needs a conpanion thick (desktop) app indeed, where you can actually buy the tracks it'll recommend.
Nevermind that fact that this flies in the face of the everything MS is about when concerning ourselves with music (aka trying to figure out ways to screw their customers so Apple won't get so much damn attention), that app has a name, and it's called iTunes.
Those kind of apps are what the music industry should concern themselves instead of running after en users. Which strategy has the best ROI? Paying $toomuch-an-hour lawyers to drag end users in front of judge, or having a couple of music nerds meet a couple of info science nerds and build something like Pandora?
Of course, were anyone like the RIAA to build anything like Pandora, you'd end up having to install MS Media Player 10 (with FREE DRM) to be able to listen to it.

Telecom as utilities

James Enck made a comment about a fund manager saying "The problem with the incumbents is that they are still unwilling to admit that they are utilities."
There's much to say about that statement, but my opinion is that if you don't have a regulation authority limiting the incumbent's behaviour to that of a utility, you're screwed. Looking at the Netherlands' situation, I don't think our regulator's been quite strict enough towards KPN, but at least they limited the damage. They did drop the ball totally on cable TV, creating what amounts to a couple of local monopolies, controlled by an oligopolistic bunch of telecom companies. Results: marginal costs are about €5/month/home, with the actual price being closer to €10-€15. Thanks, guys.
KPN knows how to recognize a juicy situation when it sees one, and introduced its DTV offer not so long ago. It costs €14/month/home, but only €7 if you happen to have KPN phoneline and/or DSL connection. How does that naked DSL sound now, you bastard customer?
Edit: Wadda you know, OPTA just announced KPN would have to share its network with other voice providers. No word on the pricing structure yet, but I'm sure there'll be enough bickering about that.

British pigopolists and customer value.

British Telecom finally has finally unveiled its Bluephone service: A mobile phone that uses an IP connection to provide landline-rates calls when you're at home. Awesome! Until you read the fineprint, of course.

• Your IP connectivity must in fact be a BT broadband link. No third parties at this party, no sir. And mobile connectivity is through Vodaphone only.
  
• You'll be charged the normal landline tariffs for outgoing calls from home (even though you're already providing and paying for the IP link).
• You're being billed the tariff relevant to the situation your started the call in. The call can switch, the billing won't.
  
• People calling you will be charged the to-mobile tariff, regardless of where you are. If you're at home they end up paying a premium to reach an IP-phone.
  

So BT is in fact double charging both you and people calling you: providing cutomer value indeed. And the best part is that when this fails you can be sure they'll blame it on something fancy like lack of cutomer confidence for new-gen communications, while their ex-customers happily keep yapping into their Skype headsets.
Addind insult to injury they have proved that voice over bluetooth can work, and so can seamless network hopping. How long until someone comes up with a smart Skype gateway? Then you just plug a BT dongle in your desktop, and off you go.
Linky-linky:
James Enck explains in numbers why it's doomed.
Martin Geddes also has an interesting commentary on value.
The Register syndicates an article from newswireless.net pointing out an extra couple of technical shortcomings and billing insanities that I didn't know about, the best one being the ability of the base station to do Wifi, but not for placing calls. This effectively turns your lovely mobile phone into a wireless device with a range inferior to the average phone cable.

You've got MAIL!

I was reading a pretty entertaining piece arguing that most applications (and productivity apps in particular) were in need of a small scale revolution ("remix" sounds cooler, but whatever).
After spending "nearly a decade" studying email and how people use it, the IBM Collaborative User Experience guys came up with Remail, their take at a what a next-gen email client could look like. It's got lots of fancy features, and it looks *more* laborious to use than simpler incarnations like Outlook Express or Thunderbird.

The one feature I love in modern client are saved-search folders: a really easy, really flexible way to present information, it stays out of your way should you happen to prefer the old-school way of working with email, but helps you get a detailed, synoptic way of what's going on in your inbox without forcing you to do all the work (classifying mails, or inputing meta-data so your client can do it for you).

Mail classification and metadata input are comparably boring, repetitive tasks, but metadata is not limited to email: the average PIM software makes extensive use of it.
Of course you see where I'm going: can't we implement some sort of saved criterion sorting for agenda and todo items? It seems pretty hard to implement with todo's (it flies in the face of context-lists and hierarchical lists), but for something like switching event colors, it could very well prove useful.

Another feature of modern mail clients is bayesian filtering, which can also be used for sorting (pdf): this presents a pretty good opportunity to improve PIM software (in addition or combination with user-defined searches), but it's unlikely to get implemented anytime soon. Bayesian sorting offers one great advantage to the email saved-search folders I talked about above: users can actually dump emails in a bayesian-sorted folder, which you can't do with a search folder (it's just a container for a search's results).

Yahoo Music & Co.

Everyone has been talking about Yahoo's first foray into the digital music business, and a couple of things came to mind when looking at it, and at the initial comments it raised:

• It's absurdly cheap, giving Mark Cuban a good reason to firebomb the RIAA again.
  
• Renting music is not my idea of fun, but it's very much in line with how it's consumed.
• It's compatible with about 10 devices. An Open DRM platform, is a pipedream.
  

The thing about MS is that they're basically selling what no-one is willing to buy, the DRM infrastructure. The music shops can worry about getting the pigopolists on board, and device makers will need to buy into interoperability. Of course, the whole thing's still busted.

On the other side of things, Apple's got their act sorted out, but they've also set the bar much lower: they make the only supported range of hardware players, and they also control the music shop. Their DRM scheme's obviously been busted but at least your iPod will play the damn music.

Edit: Microsoft's music-for-rent efforts only seem to be limited by the copyright holders, who do not want people having bought iTunes-DRM'ed songs to be able to get them for free on MSN's service. When one pigopolist's profit-driven call for interoperabilty gets bashed by another, it's irony in the making.