ISO: Engineering Lead & Co-founder

We've begun circulating the following job description...

Your mission is to co-architect and build a peer-to-peer system for editing and circulating XML documents with diverse content types, for a market of millions of business teams. You possess remarkable collaboration skills and software engineering talents; a passion for elegant design and quality code; a sense of urgency and attention to detail. You are courageous enough to fly in the face of conventional industry wisdom and navigate the inevitable right-angle turns that every start-up encounters. You are adept at creating connections and developing them into relationships. You are excited to join a pre-seed start-up, help drive it to the next stage, and become a co-founder with a substantial equity share. Prior experience in a high-intensity start-up is a plus.

We are a software startup applying a novel architecture and web UI to a vexing, long-standing problem afflicting millions of business teams. Our goal is to build an irresistible product that spreads virally, becoming part of everyday life for the vast majority of knowledge workers.

Always-on-you Web Blog Paused

We're taking a break from blogging to help a friend get a business started, and do some more detailed prototyping of an always-on-you web app. More to follow late this year...

[AoyW] Why Email Persists: It's P2P

"Anne 2.0" writes that email persists despite numerous attepts to replace it over many years because it is "good enough". Benefits she points to are: 1) Interoperability 2) Personalized Organization 3) Easy access control 4) Single point of information access. All good points, as are those that she enumerates where email fails.

However, there is a fundamental difference between email and all the mechanisms that have been proposed to obsolete it: Email is peer-to-peer. Everything else is server-centered. Email uses servers, true, but they merely provide relay service or secondary storage. The email application sees a field of peers, identified by unique addresses, and reachable via the nearest SMTP service.

What hasn't been tried is applying the user-centric P2P approach of email to writing and content-creation apps, which are the fundamental tools of the knowledge worker. Isn't P2P messaging really a feature of these apps, rather than a disconnected app itself?

Were a P2P wiki framework to emerge, enabling private wikis distributed among teams of collaborators, with commentary or chat embedded on any page, would the email client still persist as the primary—and frustrating—tool that it is today?

[AoyW] UWB Wireless Finally Working?

A Japanese outfit I've never heard of has announced a Wireless USB hub. Is ultra-wideband finally going to escape from the demo hall into the real world?

This device is essentially a wireless docking station for your PC. You plug a flash-drive-sized UWB radio into a USB port on your laptop, and plug USB peripherals into their small hub unit. Voila, you're connected to your peripherals wirelessly.

If this gizmo actually works, terrific. Even if it does, it's a bit of a yawn. UWB is supposed to empower ultra-mobile devices, like media-pods, cameras, smartphones, and flash drives. Imagine a flash drive that you never have to plug in to access; walk up to any PC, and have access to all your data (and always-on-you web apps :-) at gigabit speeds. Here's hoping this is a first commercial step towards that future.

[AoyW] The 'Web Office' Is a Tiny Niche

The web has solved a lot of problems in the past decade, but has had little impact on personal and team computing. We're still banging away in MS Office and shoving docs around via email, or piling them up on file servers. Any number of web-based online services and intranet software products have been offered to change this. They have seen relatively low adoption. As Exhibit A, I offer this Alexa graph of well-known web office (aka Office 2.0) sites, including—for contrast—the popular sendspace.com email-attachment manager:

click for full size

The Alexa numbers may well be inaccurate; what I find interesting are the trends of this graph. Why do they look so flat for all but email attachment distribution? I believe these services can't spread virally, unlike consumer web services, because business technology requires approval by management and IT staff. Management is leery of shipping confidential data out to web services run by a third party, and even if IT depts had ample time to evaluate new tools, they are uneasy about systems they can't control.

Certain "on-demand" apps like salesforce.com have made some headway among businesses which don't have the IT resources to deploy such solutions in-house. Management can approve outsourcing for capability which they couldn't otherwise access, e.g. CRM. But email is the only personal app which has made substantial headway as a service for businesses (perhaps because incoming email originates offsite anyway?).

Intranets.com, the biggest success of the Web 1.0 team sites, was acquired last year for $45M, a little more than they'd raised in venture cap. That was after six years of chasing users. Can an AJAX UI change the adoption curve of such sites? Check the trend lines on the graph for jot.com and projectpath.com (the most popular of five Basecamp domains), and draw your own conclusion.

The online web just isn't reaching a lot of users in this space, and server-based intranet solutions don't spread virally either, as end-users can't set up servers. The always-on-you web can change that.

The always-on-you web is any web app that runs on a user device (laptop, Wi-Fi smartphone, flash drive) and shares data with other users on a p2p basis. It's a web app that doesn't require an intranet or online server. It's a web app that the user carries, so it's always-on-you, as opposed to always-on(-if-you're-online). And it's a mechanism which can spread virally in a business environment, because users can deploy it themselves.

[AoyW] The Office 2.0 Conference

I just came across the web site for the upcoming Office 2.0 Conference which has an interesting list of speakers from web 2.0 startups.

As you might expect, most or all of these outfits offer hosted or intranet solutions. As such, they have to run the gantlet of the corporate IT dept for vetting/approval before they can solve any user problems. In other words, they can't spread virally among most enterprise users, a problem which consumer web services don't have.

Nonetheless, the goals of these startups are much the same as those of the always-on-you web, even if their solution architecture is Web 1.1 :-)

AOL Releases Search Logs. Who's Next?

Several hundred thousand paying AOL members have just been deliberately stripped of their privacy to the entire world.

Oh. My. God.

Now, folks, is there any question whatsoever that personal and business use of online apps like Writely, Basecamp, and Google Spreadsheets may be unwise?

As a member of such services, you don't own your data, and therefore you can't protect it.

Most ordinary web surfers already get this, which is why web-based personal computing apps have faired so poorly over the past decade. I wonder when the webtopian pundits are going to wake up.

Update: AOL's apology isn't remotely satisfactory. What are they doing to put the genie back in the bottle? They need to launch their legal eagles at those sites now redistributing this data, even if the license they attached to it allows redistribution.

[AoyW] Coming Attractions

I've been unable to blog much in the past month and topics for blog posts are piling up. Here's a preview of the posts coming in the next couple weeks...

David Beers, in an essay on his blog, proposes a mobile phone approach to the need addressed by the One Laptop Per Child (OLPC) project. He favors the "always-on-you web" model with a phone as the personal server device, rather than a flash drive. Wi-Fi smartphones would enable this, although Bluetooth or UWB (aka Wireless USB) are preferable for their point-to-point networking, i.e. no access point. (Wi-Fi can do that, but it has to be reconfigured by the user, who then loses Internet connectivity.) Unfortunately, Bluetooth has had little success outside of the wireless headset, and UWB continues to be next year's big thing.

* * *

Just to point out a flash-drive-based app system that is NOT the always-on-you web... Lexar is promoting the Power to Go software, which enables a variety of Windows apps to run directly from a flash drive. You could call this the "always-on-you office". This is cool, but if it's a web-oriented future we're rushing into, this is a backward-going time machine.

* * *

A reader wrote in to point out the Bouillon Project, which enables a peer-to-peer, worldwide wiki, wherein pages are accessible only when they've been recommended by your "friends" in the network. This sounds pretty neat, but it's not obvious to me what the mass-market application is, outside of the social networking game. I'll try to keep track of this to see what application ideas they propose.

[AoyW] The WebOS Bandwagon

Paul Boutin ought to jump off the WebOS bandwagon and have a look at the road! Like so many others for so many years now, he trumpets the vision of the WebOS, whereby PC apps and data move online... into the clutches of a posse of middlemen: the app service, the network service, and the billing service.

I've written at some length on this blog about why that model hasn't caught on (and won't), and about an alternative model, the "always-on-you web", that brings web benefits to personal and team software without forcing your apps and data online. Have a look at the "Key Posts" sidebar group on my main page.

But the bandwagoneers play on, content to speculate on the road ahead, rather than contemplate the road underfoot, which is wholly impassable.

Dear Tim O'Reilly

Dear Tim O'Reilly,

I chose the name Web 2.5 for my blog (subtitled "the always-on-you web" and the first result at Google for "Web 2.5") because I believed the term "Web 2.0" to be generic. I would not have chosen this name had I known that CMP's trademark was in the works, because in that case a conference named the "Web 2.5 Summit" could be said to confuse the market, even if a Web 2.5 event sought a completely different audience than your Web 2.0 Conference.

I accept your assertion that your conference defined the term Web 2.0 as it is now broadly used. I agree that you have established commercial ownership of the term for conferences. And I know your motivation is pure when you assure the web community that Web 2.0 can continue to be used without restriction outside the conference context.

The trouble is that much of the web community is not comfortable with a term that is restricted so. We find ourselves in the bind of having embraced your term enthusiastically, widely, but now admonished not to use it, nor similar terms like mine, for conferences. We feel blindsided by this turn of events; we have a great deal invested in this meme; we are not sure how to respond.

We still earnestly hope, despite your signals to the contrary, that your partnership will cede the term Web 2.0 to the community, and rebrand your conference ever so slightly (a move that would surely generate tremendous coverage at this juncture). That feels to us like the kind of magnanimous gesture we might expect from your organization, at least as we understand it from its track record in the community.

We will be set adrift, for a time, if you refuse us this gesture. What new label should we apply to our movement? What thought leader should we turn to? What cost will this transition exact? Our creativity, I'm sure you will agree, is better spent in other endeavors.

Most Sincerely,

Liam Breck
Network Improv
5th of June, 2006

[AoyW] Eliminate E-meetings, Collaborate Better

I just came across this fascinating post on collaboration. It posits that group discussion fails when soliciting ideas and feedback. The first few replies in a discussion may be authentic, but every other reply is tainted by them.

It proposes one-on-one email between the facilitator and each participant, e.g. BCC all participants with the request, then summarize their replies, without attribution, to all for feedback/vote via BCC again. This restricts "conversation", which makes sense; everyone knows what a waste of time meetings are.

In the context of peer-to-peer wikis, a dicussion/chat mechanism could provide a "deferred reveal" feature, i.e. all responses are posted to the shared space, but if the topic is marked "no-reveal", only posts by the facilitator are visible to everyone, until the facilitator unmarks it, or enough time elapses.

Going further into wiki editing, the facilitator would put up a page and request enhancements, and participants would change the initial page, without seeing others' changes. The facilitator would then merge the best stuff. How big a pain is merging vs. losing great ideas that don't emerge, and arguing about lesser ideas?

Reminds me of "management by walking around".

P.S. don't read the other comments before leaving yours :-)

[AoyW] Nokia Releases Mobile Web Server

Nokia's open source Mobile Web Server software turns your Symbian phone into an Apache server. (They should have called it the Pocket Web Server.) Browsers reach your phone via the internet through a gateway component, which routes traffic to the wireless service. It makes the phone in your pocket internet-addressable, provided you have a data plan on your phone.

How would you use this? Personal publishing and file-sharing come to mind, e.g. offering photos and recordings of the event you're attending to an audience of friends. Everything you capture is immediately accessible to them, without a separate step to post it.

This is pretty cool, but it's not what I mean by the always-on-you web. The reason to carry a web server, or more importantly a web application server, is to enable productivity and team apps with a web UI. That is, a UI where interlinked pages, with hypermedia content, are the focus, instead of files/folders, applications, and a stack of windows. A web UI also allows you multiple local screens, each of which may view different pages in your webs/wikis. Users would naturally want to share some pages; those would be sync'd peer-to-peer, whenever net connectivity is available.

The portable web app server gives you a framework for online web apps like Writely and Basecamp, without driving your data onto the grounds of a third party, nor forcing you online whenever you need to edit. (That's key; most knowledge workers can't simply sign up at an online app service and start posting company data to it; you have to get approval from IT mgmt. That's not just a pain, it isn't very gosh-darn likely!) This personal web service could run on a wi-fi smartphone, an Origami slate, or from a flash drive on any available PC.

Update: Turns out that this post is a decent response to Gabor Cselle's recent musing, What's Missing in Web 2.0?.

Thanks to Oliver at MobileCrunch for alerting me to the news.

O'Reilly, Get Real

There was an O'Reilly-style way to handle the trademark issue: 1) Announce that CMP was about to receive the servicemark. 2) Note that, since the filing, many web 2.0 events have occurred. 3) Indicate their desire to strengthen the Web 2.0 Conference brand. 4) Ask the web community for feedback on how to proceed.

Monday Update:
During this debacle, many bloggers have noted that they never liked the term, and please can we try a different one now. As you've gathered from the title of this blog, I think it's a fine moniker. The web has evolved; blogs and wikis are novel and genuinely useful, and advertisers are crawling all over it. Social networking is probably a fad, as are other aspects, like Writely and Basecamp, but it's all worth trying. And the web will evolve further, inventing tools that give ordinary users a web context for everyday work without forcing them into the clutches of third parties—that's Web 2.5, the always-on-you web.

The web community attracted unrelenting cynicism and bitterness after the dot-com bubble burst, and Web 2.0 is a way of telling the world that we're back, we're here to stay, we're here to change everything.

Saturday Update:
As I noted in a comment to John Battelle's defense of his partners' actions, the servicemark 'Web 2.0 <event>' (where event is a generic term like conference or workshop) was the wrong thing to trademark. Having done so, not foreseeing how widespread the use of 'Web 2.0' would become, it was the wrong thing to defend. Changing the name of your shindig to 'Web 2.0 Confab' or somesuch would have caused your partnership little trouble, and would have been defensible, both legally and morally.

Yesterday, assuming that O'Reilly couldn't be this clueless (or this), I came out in defense of them. Today, I'm joining the mass demonstration.

Seriously, would anyone be willing to help organize and/or speak at a Web 2.X Conference in the Bay Area for this fall?

Get O'Real... O'Reilly Not to Blame for CMP Misstep

Update: As it turns out, I'm dead wrong. O'Reilly IS to blame for this blunder. (Rick Segal offers a pithy analysis.) I couldn't imagine they could be this foolish. May I suggest you not attend the O'Reilly Web 2.0 conference this year. I won't.

An Irish non-profit has received a cease-and-desist from CMP Media, claiming to own a service mark on the term 'Web 2.0'.

Uh-oh, I hope there isn't a mass migration to 'Web 2.5' if CMP sticks to its (rather low-caliber) guns. That term has been coined, folks!

Tim O'Reilly, a fellow with probably the best public image in all of tech, is taking heat (lots) for a legal move by another company. The blogosphere has taken the original post on the story at face value, and not bothered to read the text of the C&D letter, an image of which is included in the post. Yet another example of how bloggers are not journalists.

Update: Intriguingly, there has been a fight going on at the Wikipedia Web 2.0 page about whether to mention the fact that CMP has claimed a trademark on the term. So far, the editorial consensus is to exclude this detail.

Apologies for the off-topic post, readers. We will return to the decidedly less hype-driven Web 2.5 story shortly.

53651 Users Can't Be Wrong...

But they often are. As pointed out by Josh Kopelman, this is the readership of TechCrunch (current count at right), without counting readers who go direct to the website (a lot). This crowd has a very different relationship to the net than the mass market. Many of them believe that the internet will someday become a global brain, which all logic & data will move into. Near-term, they believe the internet is the PC.

That is not a vision that the mass market will ever embrace. Ordinary users have a healthy fear of centralized control, and a rational aversion to organizations that would rent property that individuals would benefit from owning, like their data. If you sign up the entire TechCrunch readership to your shiny new Web 2.0 app service, you might not sign up anyone else.

Croquet Proposes Web X.0: P2P, R/W, 3D

Nothing like a little alphabet soup to start the day... Croquet, a research platform that has escaped from the lab, sprung free by a startup called Qwaq, proposes to be "an operating system for the post-browser Internet". Nothing ambitious, mind you.

The focus of their post-browser net vision is not publishing, but collaboration, purportedly on a large scale. Intriguingly, its architecture is peer-to-peer, so end-users can build the Croquet web one node at a time. That is a Good Idea. Today's server-centric net tends to serve organizations well, and individuals rarely. MySpace and Blogspot are about the extent of the user-defined net, and they don't support much more than blather.

The Croquet user experience is a 3D universe of interlinked worlds; or perhaps interlinked apartments, as each world is more likely to be a set of rooms than landscapes. Clearly the stack of overlapping "windows" pioneered by PARC and first commercialized by Apple is a terrible way to organize or present information. The browser, with its hyperlinks and history deck, is far more sensible, and akin to the ubiquitous spiral-bound notebook. Croquet takes this idea into the third dimension.

But 3D conveys the feel of wandering around with a patch over one eye and a cheap scuba mask on your face; you can't experience realistic 3D without stereoscopic display and peripheral vision. I've not seen any research showing that 3D UI dramatically improves on 2D UI for ordinary tasks. 3D is hugely popular for gaming, so the technology works, but it hasn't migrated to more productive uses. What does 3D add to mostly-textual content? Think about a bookshelf: it's a 2D array of titles; grab one, open it, and you see a pair of 2D pages, in a stack.

Another potential stumbling block for Croquet is its apparent complexity. Perhaps this is simply due to the way the web site describes it, but it sounds like a bear to master as a programmer. You have to learn smalltalk, for starters, and then a mountain of APIs and paradigms. The wonderful thing about the web is how little you need to know to do useful things. Given time, the Croquet team may hide some of its complexity. But given its academic origins, that time could be a long one.

Robert Scoble got a demo recently. The Wikipedia article describes the grand vision and project history.

Personally, I'd like to see a 2D, P2P, Read/Write web for personal and small-team applications; based on SVG, and incorporating PC documents/apps. Hmm, that sounds familiar... Oh right, I'm writing it! It's called airWRX.

AoyW vs. In-Office Spam

In-house spammers, i.e. most people you work with, have become a leading headache at the office. These well-intentioned miscreants habitually send messages to a group of people, often with documents attached, when only one or two are involved in the issue in question. They repeatedly send a document after each edit which they feel is significant. It's gotten so bad that knowledge-workers are now spending their days in Outlook, trapped! Outlook, or any email client for that matter, is hardly a project-oriented workspace tool, as you may have noticed.

This situation highlights three gaps in the desktop environment, when compared with the web: 1) The PC has no project workspace, whereas an editable web site is a project workspace. 2) Email is a protocol, not an application; email should be delivered to project workspaces, not in-boxes. 3) The desktop has no content-sharing mechanism other than email.

Creating a workgroup web site is an obvious approach to the problem, either on the intranet or at an online service. However, web workgroup tools have seen little adoption, largely because they are server-based. As such, they cannot be deployed by end-users; they require IT Mgmt approval and/or support. (IT Mgmt is especially skeptical of online services which pull company data offsite.) Also, centralized tools force users to think twice about everything they write into them, as all of it will be accessible to colleagues and managers, for all time.

Enter the Always-on-you Web: a peer-to-peer web (which runs for each user from a flash drive or other mobile device) with both shared and private workspaces. In these shared webs, an electronic discussion is simply an object on a page with other content, e.g. a spreadsheet. New messages in that discussion, or changes to the spreadsheet, are distributed only to those who are sharing that web. They either choose to be alerted to changes as they arrive, or peruse them as time permits, depending on their role in the effort. In-office email is virtually eliminated.