R.I.P: Finder/Explorer AKA "the Desktop", 1984-2007

I'm calling it - time of death: June 2007. Its been a long time coming, but it seems clear that the Blackberry, the iPhone, Outlook, Picasa, and iTunes all herald the end of document-centric computing.

The iPhone really didn't create this trend, but I'll say that its certainly a very visible final nail, just as the Macintosh Finder was the "visible" start back in 1984. Today, you don't keep piles of stuff on your "desktop" and activate applications (or applets) against relevant document parts: the vast majority your content is organized against your applications, not the other way around - the iPhone is a computer that doesn't even have a desktop, in any traditional sense.

There was a brief resurgence of the idea that the document was the gateway to your applications in the early 90's with OpenDoc and OLE (Object Linking and Embedding).

Hah.

We're at the starting tip of an orgasmic diarrhea of content creation in the form of e-mail, blog posts, music, photos and videos. And every single one of those is organized against single media form computing - barely a compound document in sight... you go to custom applications to create, edit, organize, and consume all the vast amount of gigabytes and terabytes of data we all share.

Vista Search and Spotlight in OS X only demonstrate even further how increasingly irrelevant the Finder and Windows Explorer are for everyday users.

In a slightly related tangent: What's most shocking to me today is how right Unix got it in the 1970's. URL's and hierarchical file paths seemed like dinosaur concepts in the early 90's before roaring back.

Either that, or we just haven't had the imagination to organize our way out of a paper bag since 1977... there's a parallel here.

Another way of saying "cool idea - wrong problem".

MacWorld 2008: Software Industry R.I.P.

It's only fair, and certainly poetic, but the circle is now complete. As Apple, with the Apple II (and a little nudge from VisiCalc), heralded the beginning, so too did Steve Jobs (and Apple) herald the end (or at least, the end of the beginning) of the software industry.

No longer is a "what" (as in, "What do you?"), its now officially a "how". I'm calling it: Time of Death - January 15th, 2008, 9:48am PST (or thereabouts :))

What's am I going on and on about?

Steve Jobs made some cool announcements at the annual Apple-o-phile incest/love-fest: iPhone/iTouch upgrades, AppleTV stuff, MacBook Air, Time Capsule/Airport, and no mention of OS X...

(Tiger? Leopard? What's next? I dunno!! - some damn cat??)

Its not THAT wierd he didn't mention it - after all, Leopard just shipped, and I expect the OS enhancements will probably debut at the Apple Worldwide Developer's conference in June.

What was wierd was that nobody noticed... or cared.

It's a trend that's been developing for some time, but, these days, saying you do "software" is now as meaningful as saying you're in the "customer business" (or as insightful as having an "Audience business"? :P)...... just not that descriptive, dipsh!t.

And 2008 just made it official.

If you'll forgive the math mangling: It's only a hyperbole, if you can't see the asymptopes.

The Iowa Caucus

Updated: Saw this after posting... Our Voting System is a Loser

Wow... already commented to death, but the Iowa results were very surprising, especially on the democratic side. Few had predictated such a margin of victory for Obama, and certainly not with a third place finish for Clinton. Huckabee's victory over Romney was certainly not a foregone conclusion, but if it was an upset, it also certainly was not a surprise upset (if that's not too oxymoronically hair splitting of me to say).

Interestingly (and I use the term loosely :)), that Dem/Rep difference in expectation also plays out as you look back at the Iowa Caucuses. Excluding incumbent candidate years, since 1972, only once (1980) was winner in Iowa not the Republican nominee, yet on the Democratic side, only three times was the winner the eventual Democrat nominee (vs. 4 times not).

Why is that?

Certainly there's the much publicized Democratic caucus process itself. In a particular bit of difference, voters get the opportunity to "re-cast" their votes if their candidate is deemed to be not viable. This is kind of cool - it means that you don't have to feel like your vote gets "wasted" if your picking a "risky" candidate who best aligns with more of your views.

This is a brilliant innovation in Democracy that I think better captures the "will of the majority", as opposed to the usual "will of solidarity" embodied by special interest groups. I get the idea - people stick together even if they don't agree on everything just to make sure their voice and vote matters. But I think its this blind allegiance to collective relevance over individual desire that's removed the shades of grey from our hyper-partisan brand of modern politics.

That said, the core problem with the Iowa Democratic process, I believe, is that the viability adjustment (the vote re-cast) requires that your vote is not by secret ballot - that you sacrifice anonymity (think about it). And that means that peer pressure and public perception play a much larger role in the process. The secret ballot provides protective anonymity, and is a vital and important cornerstone of modern democracy. It means people can vote their mind withou fear of reprisal in defeat (as part of the possible minority).

And this delta - that of anonymity - is why I think the Iowa Democratic Caususes are not a very strong predictor of future performance. I'm not saying Obama won't win his party's nomination, but I am saying that he's a candidate that its publically easy to align with him.

I think the real solve for this is to let people pick alternatives (perhaps multiple) up front, in some rank order - a complete overhaul of the existing system.

There's an opportunity for the overhaul as we introduce digital election process - which can provide much greater turnout, as well as enable a whole new class of election "services". The security problem will get solved, though not necessarily in our parents lifetimes (sorry, old folks).

And when it does, (not to sound hokey, but) I think it creates an opportunity to better serve the spirit of the Constitution, and of Democracy, than our current process does.

We'll see what tomorrow brings.... as always :)

Toddlers bond with Robot

Pretty cool story in the New Scientist on Monday: Robot becomes one of the kids. Basically, researches found that, with the correct behaviour emulations, an advanced robot was able to integrate into a toddler group as a peer (as opposed to "as a toy", or "as a pet", even, based on touching clues and other interations).

Video below (link):


This has implications for group behaviour theory and social development (it hints at a lot about how we develop thinking about "us"). But mostly it speaks to the evolution of robotics assistance in the classroom, especially for young children.

There is clear evidence that the "uncanny valley" gets substantially farther and wider as we age - the question that this study ponders, but only partially answers, is: Why would we ever use/need *humanoid* robots?

Still: cool :)

Steady Search Improvements

I'm not sure when it started happening, but both Google and Live now display a simple "table of contents" (link) for the first entry (Yahoo doesn't currently, and Ask has its own variant)....

Its really quite handy.

Back when, I had proposed a similar feature for AOL Explorer, based on the idea that an auto-discoverable RSS feed (a simple feature all browsers now have) is more than just an "update feed" - its really also an alternative editorially managed table of contents for the site. The idea was that you'd visit the site (like ESPN or CNN), and perhaps a small toaster would appear in the lower right presenting the feed as a mini TOC for quick navigation.

In any case, no idea if that's how the search guys are doing it - but the idea that the RSS feeds are actively programmed means they present an alternative, intelligent view into the site clearly must feed relevancy (which is really another word for "recommended" if you think about it...).

And little features like this showing are going to be increasingly important - more content means more specialization (witness: search engine fatigue)- so nicely done.

Software the Ultimate

If you're a language geek of any sort, you should really be regularly reading Lambda the Ultimate (er - "language geek" of the digital variety, that is - you freak-ish polyglots can go elsewhere). Of course, if you are one, you're likely already there, so the reference here is redundant.

That said, there was an excellent post from a few weeks ago that any futurist should embrace as a foundational principle (uh - IMHO :)). That principle, emodied in an argument about the base nature of programmable circuity, is that "Everything is software - the rest is just wiring". At some point, this will seem like a obvious thing - we'll wonder how anyone could have ever imagined it differently (and I'll bet some already do :)).

The most dramatic public demonstration of this is the iPhone, of course, where the "hardware" interface (input keys, etc.) is configured on-demand, programmatically (i.e. in "software"). But, the rise of the Programmable Processing Unit (the "PPU", whether called the CPU, GPU, Embedded processor, or whatever) has been underway for a long time - implicitly masked in the rise of "Edge Processing Capacity": smart devices of all sorts (phones, fridges, routers, blah blah ...), presaged by the Personal Computer itself.

Not there haven't been some promising mis-steps (Transmeta Crusoe comes to mind) - but it is the path. Most industries today, including Video, are full of single purpose, limited function ASICs, and that will change.

A few trends contribute directly to the idea, and value, of "Software-as-Hardware":

• Specialization of function delivering a higher quality experience: think IMDb v. Yahoo Finance v. Google. And if you don't believe even Google believes this, ask yourselves why they have CodeSearch, Google Finance, and the like.
• Aggregation of access points: think PDA/laptop/cell phone, or TV/Internet and Game console convergence - I don't mean "connectivity" here, but your physical access point to digital services.
  This is driven by what I think of as "the Lazy IT" principle required for mass commoditization: You just don't want to manage - that is, administer, install, update, and (in the case of portable access points) carry - all these access points.
  Sidenote: Mobility of access as a proxy for personalization will be an interesting trend to watch here.
• The rise of what I call "Content Engineering": data driven design systems (think HTML, Flash, and to a lesser extent Java, .NET, etc.) enabling richer and more dynamically flexible relationships between content and services providers and their end users. The essence of this practice, on-demand delivery, is at the heart of what drives the move to progammability.
  

Another future wrinkle will be as our devices (access points in this context) become even more configurable. We see a *tiny* bit of this with some fun phone form factors (say THAT 5 times fast), but I think piezo-electric stuff (and/or some karmically related technologies like digital ink, or the like) will drive some dramatic application innovations that create significant behaviour shifts in next 8 years.

Imagine what it'll be like if your apps or content can change not just the surface, but the shape of your terminal.

Power consumption impedances (in the "laws of physics" sense), I think, are the only unknown blocker, versus greater programmability. Though perhaps there are creative ways to solve even that...

Of course, its not entirely impossible I'd feel differently if my title were "Chief Hardware Architect" :)

Filling the Gap (AKA Ideas are cheap)

So betwixt my last work environment and my current one, I had begun going down the path of (again) building my own software. I had (have) a number of ideas - but that's really the easy part. And I'd say "good ideas", but, in my humble opinion, the difference between a "good idea" and a "bad idea" is in the execution.

Which is to say, its value you accrue in arrears, not a judgement you can really make a priori.

I suppose - best case - you hope its a good idea before you start doing it.

In any case, I had 3 things I did actual work on to a real degree (in terms of execution): some middleware, a web app-y thing, and a tool. They had some overlapping technology components but with reasonably distinct market segment targets. Given where I am now, I'm thinking I'll release them each in some form over the next few months - no sense letting them rot on my harddrive.

So - check back; more info soon.

GPL v3: A mess of your own making...

The GPL v3 was ratified and released a few weeks ago by the Free Software Foundation - which owns, and upgraded to GPL v3, the copyrights to an important swath of Open Source Software (OSS).

The intent of the GPL (GNU Public License) all along has been to empower people who were using software released under GPL ("copyleft" software) to make modifications to said software, and to use and redistribute modifications and improvements to others, for collective benefit. The primary legal intent of this latest revision, as I had mentioned previously, was to close the server and computing devices "holes" in previous versions - whereby companies were:

(a) using GPL software on their servers, but NOT releasing changes or improvements (because they were not distributing the software - which was the previous trigger for obligating a company to share its improvements to any GPL software) to anyone, per se, but rather letting users consume software services remotely, and/or

(b) preventing (effectively) modification and re-distribution of the software because their device was in some way preventing any modified software being loaded or running on the device (so called "TiVo-ization").

There are other OSS licenses which were always intended to be less radical: BSD, MIT, ZLIB, MPL, etc. These licenses have always been of the "do-what-you-want-but-its-not-our-fault" variety, and have arguably been as successful with their projects as the GPL - and the GPL has often been described as "viral" in that any GPL code may not be linked with non-GPL code.

But that's not the point.

The point is that they revised the license to close the loophole. And from there, much hilarity ensued:

Microsoft statement about GPLv3
Groklaw on Microsoft and GPLv3
GPL: fear is the key
Linux creator calls GPLv3 authors "Hypocrites"
Apple and the GPL

So, to summarize: group that OWNS the copyright to some software closes a loophole in how they intended that software to be consumed - and everybody's pissed (note that this is NOT retroactive - nor could it be; that's part of the intent of the philosophy of the license!). The big problem, of course, is all the folks who mistook the GPLv2 as (a) giving them rights they weren't supposed to have (e.g. using it to build closed source server software) and/or (b) took the license as contract for future commitments.

Oops. Sorry.

(btw, Linus Torvalds, of Linux fame, expressly does NOT seem to be confused. He just doesn't like it - and as copyright holder to the Linux kernel, he's free to do as he chooses - and I suppose to make overwrought analogies.)

Presence is not an Identity feature

I realized in my "Unique Visitors" post that I skipped some background - didn't "show my work" as every math teacher I had always said.

When Google released GTalk, I remember feeling like they had missed an opportunity to correct the Presence, Identity, Messaging and Accounts model - that they should have made "location" more than just metadata, but I don't think that's right, actually.

On the PC side of the Instant Messaging (IM) universe, Presence is exerted by an Identity, and (nowadays) often qualified by Location (for Location Based Services) and/or Network of Origin (for interoperation messaging systems), but I think Presence, especially for Comcast, is really a device feature, and Identities are transiently bound to those devices, generally with an (essentially) eternal TTL (time-to-live), though in the case of PCs, that TTL amounts to IM session length.

Identity is a kind of user virtualization construct (usually reasonably synonymous with "Credentials": username and password - at least for our purposes), while Presence advertises availability for messaging, and Messaging is generally a point-to-point session negotiation and communication pipe, though it may be one-to-many (broadcast), or many-to-many (chat). Note that Messaging and Presence need not necessarily be interdependent. Finally, Accounts are a billing relationship construct.

A typical use case in a househould is likely to involve multiple STBs (set top box or other consumer electronics device on the Comcast networks), multiple .net IDs (e-mail addresses), and multiple PCs in the house, where Presence, Identity, Message and Accounts overlap each. And of course, wireless, school/office, and multiple residences complicate the picture.

Usually, scalability for Presence is scale limited by subscription/notfication events, and by that I mean the "m X n" implied by the "Buddy List" construct, where any user may be subscribed to presence notification of any n users (effectively O(n^2) ). Messaging is bound by active simultaneous communication sessions (and type - multimedia for example) and is usually O(n).

In any case, I'm just rambling now - all of that's a roundabout way of saying that messaging is a service feature, not a product :)

Oh the humanity...

This post will probably read like spam, but....

GREAT website... click here to learn more!

In all seriousness, the website is a nice resource - it posts information on how to get to a live, actual person when voice connected to call centers of a variety of different companies. That's especially useful when you know your problem just isn't a problem automation can help....

...or you get cut off (either talking or on hold)...
...or don't find the right person to help you...
...or don't have the time to wade through their voice menus...
... or... well, you get the idea...

Anyway check it out: http://www.gethuman.com/us/

Automation systems are fine, in principle, and voice is great for high throughput (and especially that embodied by abstract thought, or requiring dialog), but generally poor for latency (compared to say, text, or the like) - so its not so great for user interfaces.

Futurists have been extolling the coming input revolution for a long time - but clearly speech recognition is/will be only a tiny, tiny part of solving the problem.

CPU Evolution

Hardware.info has a nice article outlining the evolution of computing hardware from the CPU/motherboard/bus model to the Co-processor/Streaming model. Its probably long overdue, and will require a substantial re-working of software to fully leverage, but it only makes sense: massively parallel tasks are ideal for computation, but we've been increasingly (and arguably, at long last) gated by the linear HW/SW programming model.

This isn't just the next pin configuration standard, or motherboard communication protocol (ala PCI 2.0 Express)- its a paradigm shift, though in practice the results will be phased and somewhat gradual; there's significant infrastructure in place that'll all have to change to broadly see the benefits.

That said, there are two other "industry" factors I think the article ignores...

First, the increasing proliferation of general purpose computing interface and access points (phones, settop boxes, consoles, PDAs, iPods, PDAs and the like), i.e. NOT the desktop - and secondly, the concurrent and related emergence of the web software model; the network's going to become (is?) the bottleneck for most in the real world, and the PC will be only one of many entrypoints.

Additionally, although AMD (and therefore ATI, whom they recently acquired) and Intel have a shared vision of the "new PC", I'd imagine NVidia imagines it shaping differently...

Apple iPhone: Wow

Everything is software - the rest is just wiring.

I've mentioned the idea of the specific trumping the general for UI, and the value of general purpose computing environments for providing more specialization of application and eliminating distribution barriers (and mentioned Jeff Han's multi-touch work before).

And Apple's built something real - much sooner than I expected. Lots and lots of details to follow I'm sure, but the key innovations are:

• Ubiquituous connectivity,
• Portable form factor, and
  
• "Hardware as software" (most importantly multi-touch)
  

I think this is disruptive for reasons we don't even know yet. I'm sure many will predict failure, but they're wrong because the killer app won't be your cel phone replacement - though that'll be the excuse to buy it.

Read about Steve's Keynote here (and the iPhone here).

There's plenty of Jobs-ian hyperbole ("the first fully usable browser on a cellphone?" the Opera guys, or the Nokia S60 browser team, or the mobile Internet Explorer team might disagree - for very good reason) but its a doozy of an event and signals a significant transformation: Apple wants to become the last mile (the UI) for everything you do. Let everyone fight over the increasing commoditization of infrastructure and networks; Apple just wants the users.

That doesn't mean they're going to get them, but I do think this is a real effort... its no iWork (like that's a real replacement for Microsoft Office) or (*cough*) Zune... in fact...

...this is what the Origami should have been...