plaxoed!

I saw Om and Scoble’s light banter about thick vs. thin clients and it got me thinking about my vision for the future of computing – plus, I’m on a flight to Seattle again with no Internet connectivity, giving me plenty of time to think and write about the topic

As Om notes, many applications are moving to thin clients, more devices are becoming thin clients, and those devices are becoming more ubiquitous. Very true, but before the Internet, there was no reason for most consumer users and even many enterprise users to have thin clients. As many have pointed out in the comments of both posts, Office doesn’t need to be in a thin client; why should it? Even if it was fully responsive while online, what about the offline scenarios? The connected scenarios that people talk about (online collaboration, document sharing, software/template update delivery, etc.) are all available today with the application residing locally on your machine.

On the other hand, as Scoble notes, there are many applications that you probably will always want a thick client for – Photoshop being a good example. Generalizing a little bit, any application that requires a lot of processing power and demands a responsive interface will need to have some tie to the desktop, at least in the near future.

So yes, there are thick clients and there are thin clients… but is it really that black and white?

What I see in the future is most applications becoming some shade of grey – thick apps will start getting more connected, thin apps will start bleeding over onto the desktop. Let’s take one of the most pointed examples: e-mail.

I have 3 main e-mail accounts: blog, work, and personal. For my blog address, I use Thunderbird exclusively; for work I use Outlook/OWA; for personal I use GMail. The Thunderbird client vs. GMail access highlights the thick vs. thin client pretty readily. With Thunderbird, I can read and compose mail offline, I get extremely snappy response, and I can use a familiar drag/drop interface. With GMail, I can read my mail wherever I go, the “software” is always updated, and there are interesting places where integration can occur (see Google Talk’s use of the GMail address book).

Now take a look at Outlook/OWA… for the same e-mail account, I have both a thick client and a thin client. This begs the question: why isn’t there just one hybrid client? Or rather, why don’t they improve OWA to the point where Outlook can just be a browser with some offline store (cached mode) capabilities? Better yet, why isn’t that online/offline store a WinFS store? Why isn’t Outlook rendered through Avalon? Wouldn’t this hybrid Outlook communicate using Indigo?

Aha! The WinFX vision lives

As you can probably tell, these new frameworks really excite me. Why? Because it makes the thick/thin client debate obsolete and gives the users the best experience. Want snappy thick client response? You get it with an offline WinFS store and client-rendered Avalon UI. Want a sweet travel scenario? You get it with an online WinFS store and a remote scripted web app. Want to connect up these scenarios? Indigo and WinFS replication to the rescue!

The really frustrating thing is, as exciting as these scenarios are, they are still at least 5 years away. Sure Avalon, Indigo, and WinFS will probably ship within the next 2-3 years, but the majority of users won’t have these frameworks until the version of Windows after Vista. That means most application developers – including Microsoft themselves – won’t be able to use these constructs until then.

In an extremely ironic move, Office has been resisting dependencies on the .NET Framework. Why? Beats me. If I were BillG, I would’ve said, “Hey Raikes and Sinofsky, guess what? The latest version of .NET Framework is going to be included in the Office installer and future development on Office will start using the .NET Framework.” What happened to Office being the best dogfooding of the Windows platform?

Oh sure, there will be complaints. The .NET redist is too large, it’s too slow, we’ve got too much legacy code. Erm, since when did that stop Microsoft from releasing Office 95? If I recall correctly, Windows 95 wasn’t the snappiest OS on the existing hardware when it first came out; my old Pentium running Office 6 on Win 3.1 readily spanked Office 95 running on Win 95. But back then, Microsoft was willing to make a big bet on their own platform – they said, fine, it might be slower, but we’re looking at the future here.

Currently for $500 you get a 2.8 GHz processor, 512MB of RAM, and a 40GB disk. I don’t think that hardware is going to have a hard time running Office inside the CLR.

Microsoft’s in a dominant position right now. What do they have to lose by transitioning Office to .NET?

Ok, so that went off on a tangent, but hey, the flight is 2 hours, so I’ve got time to explore tangents 

Anyways, back to the original topic, in my opinion, the vision of WinFX is spot on. Syncable online/offline stores, service oriented communication layers, and client-rendered rich UI. I think that’s the future of application development – it’s neither a thick client nor a thin client. It’s just a connected, rich client.

Will Microsoft be the one who enables this? Will Google? Yahoo? Who knows, but the developers of the world are waiting for someone to do it

This past weekend, Rose, her parents, and I went to Victoria, BC to check it out. We spent Friday night in Vancouver and then took the first ferry (7AM! ) to Victoria for the entire day on Saturday. Here’s us on the ferry:

The ferry going to Victoria was pretty cool. It had a cafeteria, a buffet, and a work lounge ($7 admission, but you get free continental breakfast). The only thing not cool about the ferry was that all the seats had arm rests so you could lay across multiple seats and take a nap. That’s probably by design to stop people from monopolizing multiple seats when the ferry is full, but it’s quite annoying when you want to nap and there’s no one else on the ferry using those seats…

When we got to Victoria we headed to Buchardt Gardens, quite a photographer’s dream! It was a bright and sunny day so we got some great shots. Here are some from my camera:

Our cameras were filling up quickly so we really had to start being picky about which shots to take; I can easily see someone filling up gigabytes of flash memory here

After Buchardt Gardens, we went to downtown Victoria for a stroll. Along Government St (which is the main street in Victoria), there are lots of shops with big stuffed animals in front. We’re not sure why it’s all the rage there, but we took some pictures of them too

We walked around everywhere, taking in the sights and sounds. It’s quite a beautiful little downtown area. Quite touristy, but oh well, that’s what we were anyways.

While taking all these pictures, I started playing around with the more advanced controls on my Canon Powershot S400 (which I think has been discontinued now). It’s got some cool adjustments for those who like to tinker - exposure, color balance, ISO modes, auto focus settings, etc. But I quickly found myself frustrated with the menu interface (which requires you to use the screen and is also slow to navigate) and the inability to precisely control the auto focus. It makes me think that perhaps it’s time to upgrade, but that presents me with a major choice: is it time to go dSLR?

My co-worker Terry is quite knowledgeable about cameras, optics, and photography in general so I consulted him. He convinced me that in order to get the best photos, I’ll need to use a dSLR. The problem though is that those things are huge, especially if you’re also toting around a few lenses. I keep asking him why I can’t get awesome photo quality out of a compact digital camera and he keeps telling me that it’s the physics of it all. You just can’t get good optics in a compact point and shoot. Argh!

Anyways, I’ll probably end up with a Nikon D70s or a Canon EOS 20D if I decide to go dSLR. If I decide to stay with a compact, Terry highly recommended the Casio Exilim EX-Z750, so I’ll probably check that out.

Here are a few cool pictures I got my S400 to take, check out the rest of my Victoria set at Flickr if you’re bored at work

Rose took the last one of the big government building - pretty sweet shot

Rose linked me to Slide yesterday and I got around to checking it out today. I’m not really sure why it’s all that great yet, but maybe I’m missing something… for me, the product would be a million times better if it integrated with Flickr. My pictures are up on Flickr and so are many of my friend’s pictures.

The desktop app is pretty slick and I like how they have feeds. Flickr has feeds and could use a slick desktop app like Slide… I wonder why didn’t they just build on top of Flickr’s API?

SUCKS! Now I know why there’s a traffic problem; most public transit routes are ridiculously stupid. Yesterday, my flight got into San Joes airport (SJC) at 8AM. No one was available to pick me up, so I decided to take public transit - after all, I figured my office was just down 101, there’s gotta be a quick way to get there.

So I look at some maps and see that there’s actually two ways I can get to Mountain View: I can ride CalTrain or I can take the VTA Light Rail. However, there’s a problem: there’s neither a CalTrain nor a VTA Light Rail station at SJC. Instead, they have “Airport” stops that are too far to walk to; you have to take a bus.

Fine, it’s just a quick bus I figure. So I go to the bus stop in front of the terminal and wait… and wait… and wait. One of the buses on the schedule never arrives and the next one arrives about 5 minutes late. In total, I have to wait about 30 minutes at the bus stop. The first one that arrives is a #10 to the CalTrain station. Looks like it’ll work so I get on.

Now this #10 bus decides to take the long way to the CalTrain station, winding its way through an airport employee parking lot and not making any stops, but driving extremely slowly. It’s bad enough that there’s no CalTrain station at the airport, why does the transfer bus have to take a roundabout path to the station?

I get to the CalTrain station and since the #10 was late, I had just missed a Northbound CalTrain. I had to wait another 30 minutes for the next one. For some reason, they decide to drop people off at a random CalTrain station instead of a main station. This means that only half the trains decide to stop there. Why would they do this? Don’t they anticipate people coming from the airport and wanting to get on CalTrain?

At this point, I was thoroughly annoyed, but what other options did I have? And to make it worse, there’s absolutely nothing by the CalTrain station. A Starbucks or a small cafe would have totally capitalized on me at this point.

The Northbound CalTrain finally arrives and I get on. It takes me to the Mountain View CalTrain station where there’s literally 20 company shuttles waiting around - none of them mine. There were 2 or 3 Yahoo! shuttles, an Apple shuttle, a Microsoft shuttle, and shuttles to some other random companies around the area. So again, I sit around and wait.

Anyways, the shuttle finally arrives and I finally get to work. Total travel time: over 2 hours. That’s ridiculous. I’m never riding public transit again… maybe I’ll buy a huge SUV just out of spite

Well, I’ve been in Seattle, Victoria and Vancouver all weekend and what do I come back to? About 150 comment/trackback spams taken care of by Spam Karma. On the one hand, I’m really glad that Spam Karma took care of them all; on the other hand, I’m wondering why the spammers suddenly decided to turn their spam engines on my blog…

Anyone know how you get onto the spammer’s lists? Or better yet, how to get off?

I’m on a flight to Seattle and I’m bored; I wish these they would hurry up and get Internet access on planes already. I know Scoble, Chris Pirillo and some other people were able to check out the new Wi-Fi enabled planes courtesy of Boeing a few weeks back, but they need to deploy that stuff stat!

Anyways, one of the thoughts I’ve been mulling over the past few days is a strange inconsistency with the way Microsoft has been setting up its strategy for MSN. First, let me put out an assertion:

One of the smartest things Microsoft ever did was create an awesome developer platform for Windows. It was pure genius; they created an ecosystem around the Windows platform and it is widely recognized as one of the reasons that Microsoft has an OS monopoly today.

And then look at how MSN is operating:

From what I can tell, MSN has been operating in a walled garden. Most of their services do not have accessible developer APIs (there are exceptions I’m sure, Virtual Earth comes to mind). They have put the onus upon themselves to create the Internet experience that they think consumers want. The MSN backend infrastructure is completely closed from what I can tell; there is no Win32 of MSN.

Here’s what I’d love to see:

Microsoft takes a strategy from their own playbook and replicates it for MSN. They stop focusing so much on providing all the services themselves; instead, they start focusing on enabling other companies to create services on top of their platform. Lastly, connect the MSN platform to the Windows platform. Make it easy for desktop apps to use MSN as a store; make it easy for web apps to access things on the desktop.

Option 1: Provide the platform for free. Build an ecosystem of developers who write connected apps, using your infrastructure as their backend. Once the ecosystem is built up, create MSN apps that are worth paying for – such as MSN Money, MSN Music, mobile sync, etc. These MSN apps are on the same playing field as 3rd party apps; the ecosystem works the same way as the Windows platform ecosystem works today.

Option 2: Charge consumers for enhanced platform access. For all users, provide a base level of functionality for free. For Windows users, their experience gets better. For those who want the best experience, they can pay a subscription fee and get access to enhanced platform functionality. This enhanced platform may enable cool MSN developed apps and/or it may enable some more complex 3rd party apps. Maybe you give 2GB of online storage space and have to pay for more. Maybe you’re allowed to do 1Gb of transfer a month with the API, and you have to pay for more. I’m sure the marketing and business guys can think of something

Option 3: Consumers get it for free, business have to pay. Make it free for consumers so they get on your platform. Then, businesses who want to play have to pay. Basically, they are paying to get access to the users. The market for this is potentially huge; if the MSN platform is an extension of the Windows platform, we’re talking about allowing companies to get access to all Windows users through the MSN platform.

If MSN were to do this, and I worked at Yahoo!, Google, or AOL, I would probably pee my pants. Out of all the big players, Microsoft is the one best suited to build up a platform like this - an Internet OS. They’ve got all the pieces: the experience, the bankroll, the talent, and the infrastructure.

I wonder why they haven’t done this… there are lots of incredibly smart people at Microsoft and many of them work on MSN. I’d be shocked if BillG hasn’t thought about this very idea before – he’s an incredibly smart guy. Do the people at Microsoft still think the future of computing is on the desktop? Do they not believe that the future of computing is in the network, not on the desktop?

Sure some things work pretty well today on the desktop – word processing, spreadsheets, graphics rendering, and (for some strange reason) IM clients. But you’ve got to wonder, one day, when everyone has Gigabit connections, won’t it be easy to have a remote hosted copy of Microsoft Word? These ideas may not have been feasible back in 1996 when people were still on dial-up modems, but within the next decade, I think we’re looking at a rapidly changing computing landscape.

The Netscape dream will come true sooner or later, you can delay it, but you can’t escape it

Melody Toyota called me today and told me that my car was fixed (a day early!) and the total charge was only $450. Not bad for a new transmission and clutch (true, the tranny was covered by warranty, but a clutch job alone usually costs more than that).

I asked them to save my old clutch and took a look at it. The mechanic agreed that it wasn’t too badly worn for having 45k miles on it. They probably could have put the old one back in, but since they were already in there, I figure why not get it fixed up all nice.

Now the clutch is super smooth and my steering seems a little bit tighter (not sure why that is…). The horrible noise is gone and things seem to be working pretty well.

I’d recommend Melody Toyota in San Bruno. They seemed pretty reasonable and were willing to work with me on bringing the price of my repairs down. I chatted with the mechanic a little bit and he was really nice about explaining what he found and what he recommended I do. They gave me the rental car for free for 4 days, courteously answered my barrage of questions, and even finished early. Either they are really nice, or they are really good at pretending to be nice. Either way, I’m currently a satisfied customer.

For the past few days, I’ve been driving around a Camry as a loaner car and I must say, if I was looking for a nice budget family car, the Camry would be a strong contender. At $20k, it’s a bit pricey, but you get what you pay for. The Camry was smooth - road noise was minimal, acceleration was weak but consistent, handling wasn’t as bad as I assumed it was going to be, and steering was decent. Cavernous trunk space and good interior room round out the car and make it quite a good overall package for its target market.

In contrast, when traveling I often will rent a car from Avis and they usually have a GM fleet. So I’m quite familiar with Grand Ams, Aleros, Malibus, Impalas, and Monte Carlos. Most of these vehicles are available for less than $15k - especially after the special deals and employee pricing - but clearly reflect their bargain basement prices. The interior build quality is usually poor, all these vehicles aren’t too moderate on the gas, the handling is weak, the driving feel is non-existent, and the steering is loose and imprecise. I’m hopeful the new generation of American cars will address these issues. After all, they are kind of at a do-or-die point now with GM and Ford getting downgraded to junk status by major banks and losses steadily rising.

Just now, I noticed that Toyota has discontinued the Celica… eh, whatevers. Bring back the Supra to replace it

While driving into work today I was listening to NPR (KQED) and Forum had a segment on military bloggers. The audio stream isn’t available yet (the show is still running I think), but they read some excerpts from 365 and a Wakeup and Armor Geddon which were simply amazing. The quality of writing and the stories these guys are simply gripping - makes us tech bloggers look totally weak and whiny (I’ll admit I’m whiny sometimes ).

There is also a Wired article this month about military blogs by John Hockenberry, one of the guests on the show today.

Anyways, I’m subscribed! These blogs are fascinating…

Yes I know, what a cheesy post title

I’m not entirely sure what my opinion of Google Talk is yet… is it genius because of it’s simplicity? or is it a weak offering in an already overcrowded industry?

Or perhaps it’s a just another place where Google can grab more of your personal information and use it to build up a profile about you. Check out this quick analysis of the privacy policy.

Specifically this part:

We may record information about your usage of Google Talk, such as […] the contacts you communicate with.

Ah, so Orkut wasn’t working so well eh? Google Talk may be just what they need to augment users’ profiles with social network info.

And of course, Google’s using Jabber, but they haven’t opened up server to server communication yet. Although John Battelle links to a Businessweek article where Georges Harik claims they are working on it.

So we’ll see what happens… maybe developers will latch onto Google Talk and build up a little ecosystem of cool IM apps. Maybe that will cause the other IM networks to stand up and take notice, open up their walled gardens, provide SDKs and web APIs, and finally federate their networks.

Or maybe Google Talk will fall flat just like some of these other services…

I’m hopeful that the more competition there is, the better it is for the end-users

By the way, where is Scoble or Zawodny on this?

What’s this now? Reuters is reporting that the LA Times said Google is planning an IM system.

This has been rumored for a while now… Om suggested a VoIP spin just yesterday and everyone’s been wonding when they were going to do something with Hello (part of the Picasa acquisition?).

They say it’s built on Jabber which means it should be able to talk to ICQ and iChat. What about the other major IM services (Y!, MSN, Skype, AIM)? Why can’t they all just get along

Supposedly it could be released as early as tomorrow… we’ll see

UPDATE: Apparently, their Jabber server is available right now. A few people here are already playing around with IMing through talk.google.com

And as Terry mentions in my comments, Jabber is an open protocol that should be able to interop with other IM networks. I can’t wait to see this client interface; I wonder if it’ll be sweet…

UPDATE 2: Dru thinks that Google will announce that they’ve bought Trillian tomorrow, but Derek in my comments points out that Download Squad has the screenshots now… thanks for the tip Derek!