June 30, 2006

Manageability says :

Truth is, although we collectively see the train wreck coming, we don't collectively witness the failure as a single unequivocal event.

Why is it, that although we can intuitively grasp that a standardization effort is wrong, we don't seem to be able to stop it?

Why are Standardization Train Wrecks always in Slow Motion?

Maybe there's some simple game theoretical model for this. It's some kind of interesting question about individuals and the commons. You win if you jump on the standard early and it "does" become a standard. You also win if you avoid the standard and it fails. However you lose if you jump on the standard and it fails, or you avoid the standard and it wins. This is basically the "majority game". You want to join the majority.

On the other hand, there's a timing issue. The later you join a standard, the less value you get from it. (Because early movers have an advantage in designing it to suit themselves, plus they get early expertise and recognition.) On the other hand, the earlier you are in a standard's life, the less knowledge you have about whether it will be succesful, so the more risky the investment.

Also, maybe the stakes for being involved in a standard varies with the size of the agent involved. It's worse to be seen as a multi-national who missed out on the next big thing, than a small-business who pragmatically waited until the signposts were clearly visible.

Could be an interesting model for someone to build.
ThoughtStorms: BloggersAsEndoSymbionts

June 27, 2006

Dave Winer :

I bet there's a law here, or a life lesson. Either you're going to be a platform vendor or not. If you choose to push a platform, don't go half way. Platforms that are picky usually don't gain traction. If you got a platform you must be open to all comers, enthusiastically, without reservation.

June 26, 2006

OpenLaszlo Project Blog

Here's something else I'm keeping an eye on.

Yesterday we announced a strategic partnership with the Dojo Foundation: OpenLaszlo will not only be licensing the Dojo Toolkit for use within our next major release, we will also make substantial code contributions back to Dojo for use by the entire Ajax/DHTML community.

OpenLaszlo getting into Ajax.

Here's a prediction : the two most important GUI standards in the next two years will be the Flash virtual machine and DHTML + Javascript + SVG.

OpenLaszlo is probably set to become the first major tool to develop for them both.

Macromedia have "owned" this area for a while, with Dreamweaver and the Flash tools. The question is how their expensive proprietory software model will compete against OpenLaszlo. Their advantage : good connections with the existing, mainstream web-development community. Their disadvantage, not so good connections with the rest of the developer community. Flash is still weird for programmers.

Microsoft are where? They've turned VB from a decent GUI scripting language into C# with a different syntax, and both are essentially following Java on the road to statically typed oblivion. (With the added disadvantage that these languages are being used to try to drive people onto .NET 2 and Vista)

Java Applets themselves are out of the running. Scripting languages like Python, Perl and Ruby have no real story for GUI development. Tk is from two generations ago. wxWindows is still an awkward amalgum of two third-party libraries. The Java community themselves are deserting Swing for the Eclipse library. So why would, say, a Python programmer choose Jython + Swing? (Although she might, I suppose, choose Jython + Eclipse.)

Mozilla's XUL is intriguing but has no significant support from a development environment.

Everyone as Co-Creators

I have to say, I agree with everything Dion Hinchcliffe says here.

Spreadsheets, Hypercard, Wiki ... three of the applications that have most succesfully created a gradient between "user" and "developer". New innovation will come from the "edge" experts in these, and other software which takes its lead from them.

When I wrote that innovation came from developers and couldn't come from users, what I was really thinking of was "passive users", who couldn't or wouldn't understand what the deep model of what was going on in their software. Those that can understand what it's really doing, and can use this knowledge to help pioneer new ways of use, may very well be innovators.

June 24, 2006

Users vs. Developers

Trolling Chris Pirillo wants some.

He he! Well someone’s got to stand up and school the boy!

I admire people like Winer (who’s link I followed over there). I like the notion of “users and developers partying together”. I think developers have a responsibility to give the users good things. And that developers can’t know what these good things are without paying attention to users.

But the idea that development is a commodity - which is essentially what he's saying here -, that it’s like the water supply which can be turned on or off or piped-around at the will of the user, is wrong, wrong, wrong.

The reason is, that good software creation, like any other creative activity, requires a deep knowledge of the nature and constraints of the medium. You can’t invent the transistor without a profound understanding of physics. Nor write a great novel without being a master of your own language. Nor a great painter without knowing paint. Nor invent radio or television or the computer without a background in the relevant science.

Without knowing your material you’ll never see the statue in the block of marble. Never have intuitions about the possibilities that the medium holds. Users can, at best, offer advice for incremental improvements : “I want something like X but with these annoyances fixed.”

But that’s never going to give you radical new things. There are no great innovations in software that have been driven by user demand. (Name one!) They all come from geeks who knew what software looks like from the inside, who saw something in the computer and said “hey! I could make it do that, too. Kewl!”

Users and developers have common enemies : dinosaur software companies, companies who like to specify and buy software which they hope will impose particular work-practices on their employees, market-forces which require cutting-corners to get the product out of the door today rather than get it right.

Let’s work (and party) together to fight that.

But the other … don’t fool yourself. Users have been dreaming of getting rid of developers, like, forever. They never succeed.

Every time the users abandon us, we go away and create yet more cool stuff, on our own, for ourselves.

And eventually the users come crawling back, because they want more of our pure, raw innovation, rather than more of the stale old fluff which is all that they and the marketing people (who "understand real users better than those disfunctional geeks") are able to come up with by themselves.

Update : MP3 of Pirillo's presentation at BloggerCon. Yawn! ;-)

June 22, 2006

Robot Wars

Microsoft release their first Robotics Product : Robotics Studio.

Will this be more than a niche? It seems that here's the place where hardware is currently far from commoditized. It will need to be, before the software layer becomes a serious platform.

I wonder how it plays with White Box.

Meanwhile iRobot is a platform, too

June 21, 2006

Ning Screencasts

Aha! Ning have started putting up screencasts like Create an App from Scratch
Seb’s Random Thoughts � Is It Finally Happening? Will Skype Soon Become What We feared?

Are we about to experience a wave of audio-spam on Skype?

This is particularly obnoxious because Skype is using our machines to route calls. We, reasonably happily, give up "spare" processor cycles in return for free phone-calls. But what happens if Skype is deluged in spam? The spammers will be consuming our resources.

Hwo will Skype handle this? Maybe an official blacklist of spammer accounts? But what else?

June 17, 2006

Some discussion over on ReadWrite Web on the new Netscape front-page.

My thoughts. Clearly it's not as bottom-up as Digg. But it's sure as hell more interesting than the previous incarnation of Netscape. Calacanis seems to be a smart guy who's pushing AOL in the right direction. Let's face it, when was the last time anyone talked about Netscape?

It's also worth comparing digg.com with something like slashdot.org, which has a fanatical community but is editor driven.

In fact, there seems to be a lot of cross-over between the Slashdot and Digg crowds, which suggest that the editorial-driven approach and the mob-driven are complementing each other rather than mutually exclusive rivals. If that's so, it's not hard to imagine that such a complementary mix might work within a particular site too. It's going to be worth experimenting with, at the very least.
Interesting Gillmor Gang.

Steve Gillmor is arguing with Mike Arrington about Google's Spreadsheet vs. the SocialText / WikiCalc deal. Arrington thinks WikiCalc, hosted inside enterprises, is going to be the bigger threat to Excel and a Microsoft Office oriented workflow than Google's Spreadsheet which business can't trust with its private data.

Gillmor thinks that Google are wiring their applications together to lure the user into their suite of applications. For example, an email received in Gmail that talks about a dinner-date, offers the user the chance to make an appointment via Google calander. This is turn leads to the option of looking up the venue on Google maps. Gillmor's assumption is that this will suck people into the Google-suite. Essentially each application recruits users for the next via a simple hyperlink.

Could Google eat Excel the same way? Automatically offer to import (and show via Google Spreadsheet) any Excel spreadsheet that passes through Gmail?

Finally there's an interesting discussion over the fact that these online spreadsheets can't really import the formulas from Excel. The Gang seem surprised by this, but given that Excel packs the whole of Visual Basic for Applications as scripting language, it's obvious that the more sophisticated applications built with it will be hard to port without reproducing the whole Excel / VBA engine.

June 16, 2006

If I ran Microsoft ...

We can all fantasize, right?

Here's what I'd do if I ran Microsoft.

  • My brand is "Hot"

    "Hot" not "Live" (which doesn't exist), not "Windows" (which is in trouble), not "MSN" (yeuch!).

    • Skin Hotmail.
      I'd tidy up Hotmail to make it CSS compliant. Then get a bunch of designers to churn out multiple Hotmail "themes" (ie. templates + CSS sheets). Get some input everywhere from the ZenGardeners to Design Outpost. Let users choose from a fifty different looks for their Hotmail page.

      The result would be cheap, cheerful and generally get Hotmail talked about for a while.

    • De-advertise
      Take most of the advertising off Hotmail. And talk about it very publicly. Maybe put some very discrete text-ads like Google's but nothing more. Tell people I'm willing to lose money on this in order to make Hotmail a more comfortable place for my users.

    • Spelling and grammar-checking in Hotmail
      'nuff said.

    • HotPage
      Whatever it is that MS do with Spaces, blogs, homepages, social networking etc. etc. It should be blended in and standard with your Hotmail account.

  • Open Source Security

    I'd admit, right now, that security is not a question of private advantage but of public good. It should be treated as such.

    Vista and IE 7 must have an architecture with fairly loosely coupled modules. (If not, MS are fucked anyway.) Take all the security related modules, and release them under a BSD-style license, without patent restrictions. This includes malware detectors, live-updating code and things that check for pirate copies of Windows etc.

    Ownership of all these modules should be given to a new, independent, non-profit institution, spun out of Microsoft with some reasonable financing and some good staff. (Think of the Mozilla Foundation.)

    This foundation should have a general remit to be responsible for protecting the Windows ecosystem from all threats. As some of its products will be folded back into the main (proprietory) Windows codebase, employees should have the right to see the rest of the Windows source-code.

  • Channel 9 and MSDN are the same thing

    Developer mindshare has always been a key part of Microsoft's success. Ever since MS were a company that made Basic, the whole edifice is balanced on their ownership of the development tools and leadership of a developer community. And, largely, MS have understood that and served that community pretty well. MS development tools are pretty good and not allowed to get rusty. Compared to Sun or Apple or IBM, MS have been relatively forward in opening up and allowing their geeks to take part in the global geek conversation. Channel 9 was a statement of intent and a recognition of the way things are going.

    Compared to the free-for-all of MS blogs and Channel 9 and the expectation of smart developers around the world, MSDN looks corporate and restricted and simply ugly. It's time for Channel 9 to take over. Time for MSDN 2.0.

    • Lose the subscription charge. Everyone who wants an account can get it for free.

    • Copy all the documentation into wikis and let people get in there and edit it. There'll need to be some policing by staff. But generally let the wiki stay up-to-date and reflect real problems people have.

    • Ultimately - and there's a question of "sequencing" here but, nevertheless, ultimately - the conversation with developers is more important to MS than the revenue from the tools. VisualStudio, the various compilers and .NET libraries should eventualy be released as free (open source) software.

    • MS should really embrace Python. (It could be another scripting language, but Python seems to be most likely.) VS for Python should be given equal billing with C#, ASP and VB.

  • We don't need a browser. We need Flash

    Seriously. What value is there in Internet Explorer? MS will never make a penny selling it. The browser is one of the most commoditized components of the web. There's no "rent" to be captured there. Pitifully little incentive for anyone to violate standards like HTML, XML or SVG.

    The idea that MS should try to "own" the browser is based on a flawed analogy with the early 90s competition between Windows, OS2 and Mac; an analogy which Netscape originally encouraged (as in "the browser will commoditize the OS") to their own destruction. In fact, while browser standards are commoditizing the OS, the specific flavour is not an issue. Almost unanimously web-service developers are highly wary of getting locked in to IE or Firefox specific code and are targetting all browser standards. Treating the browser as an "ownable" platform is absurd.

    There isn't even a public service value in competition between Firefox and IE. Because Firefox is free software, there are dozens of micro-projects pushing it in different innovative directions and building new things on top of it.

    IE is a waste of time and money and attention for Microsoft. I'd simply forget it. We can channel some of the team and resources into supporting Mozilla. (And maybe folding some of the good bits into the Mozilla code-base.)

    At the same time, there is a remarkably successful proprietory platform on the web. One which has the kind of domination that MS can only dream of. And which seems to have its status almost by accident : Flash.

    Seeking advantage for MS, and if I hadn't completely cured myself of the habit of wanting to own platforms, I'd put some serious thought into how I could get hold of, and embrace and extend Flash.

    I wouldn't want to buy Adobe / Macromedia. But I might want to buy Laszlo Systems and push Flash authoring into Visual Studio.

    I'd make Flash a first-class citizen of the XBox and of XBox Live Market.

    I'd put a player as standard into Windows (probably as part of Media Player).

    Like I say, if I was still not cured of the whole "embrace and extend" tactic, this is where I'd probably try it. To allow some of the Windows media file formats to be embedded in Flash.

    Oh, and as we're in the realms of pure fantasy here, why not buy Broadband Mechanics and put Marc Canter in charge of the MS assault on the home.

Update 2007 : I wrote some stuff about the opportunities of Excel.


Sorry I haven't been writing much here recently. A lot going on. Here's a quick resume of some things I've noticed.

Several of the usual suspects being insightful :

Umair Haque has a couple of great posts on Google, linkfarms and the effect that Google has on the media.

Robert X Cringely has had some really interesting thoughts on the different types of IT giant, with particular emphasis on IBM's woes. Not to mention some criticism of Google.

I'm naturally very excited by the news of strategic alliance between SocialText and Dan Bricklin's WikiCalc.

That's a very clever and sensible move. Bricklin has a lot of brand recognition as the inventor of the spreadsheet; and WikiCalc is a pretty good synthesis of spreadsheet and wiki : two genres which I've long thought have a lot in common. SocialText are already deeply wise about social software and seem to have mastered the art of selling wikis into companies. They complement each other beautifully.

(At this point, I guess I should disclose that I've recently applied for a job with SocialText, and I haven't (yet) had any sort of official rejection. So read this in that light. Nevertheless, even if they turn me down, I'm gonna have a hard time not cheering them on.)

Finally, Robert Scoble leaves Microsoft and Bill Gates says he'll stand-down in favour of Ray Ozzie. Nothing much to add here. Obviously MS will suffer from losing Scoble in the short-term, but I believe them to have been on the Cluetrain for a while, and I don't think they'll get off. Ozzie will continue to blog, and understand the internet. Others will blog from their corners with continued MS blessing.

Scoble has a big brand, but I'm not sure how it will survive away from MS. Scoble and MS were a great double act. The incongruousness was part of the allure. Scoble as good blogger in what's essentially a straight journalistic organization may have less creative friction to work with.

Meanwhile, as Cringely points out, MS's business model is still disappearing. Proprietory software, ultimately, doesn't scale. MS can't hire or co-ordinate enough coders to keep churning out 95% of the world's users' software every three years. And as the market fragments, the benefits MS get from network effects decline catastrophically.

The whole thing (ie. the Windows product line / world domination strategy) is a huge soggy mess, and it's not surprising Gates is bored with it. MS have got the talent and money to change direction. But they may be suffering the classic "Innovators Dilemma". The alternative business models available just aren't as profitable as they're accustomed to. They can't easily take Google's advertising market. And even if they could get in there they'd end up dividing it. My betting is that even Ozzie and a clued-up workforce won't be able to escape this problem.