|Russ Lipton Documents Radio
simplex veri sigillum
Bootstrapping or Beta?
I was irritated last week when a poster to the Radio discussion forum commented smarmily on Userland's chutzpah for selling a beta product like ... Radio .... for forty bucks.
This not only misunderstands the Userland support process but the difference between bootstrapping a technology and shipping a product. Even more fundamentally, it mischaracterizes the collaborative nature of the Internet and of Radio itself.
Every shipping product blesses us with bugs along with features. This includes your car, your pacemaker (ouch) and your computer software.
Products like Microsoft Word sell us thousands of bugs. Big deal. Only a tiny percentage of those bugs are 'show-stoppers' - they crash the software or otherwise make a user's life hell. It is indeed a vendor's responsibility to see that such bugs are squashed during the beta phase of a product's development before it ships to paying customers.
A product's remaining faults range from extremely minor issues that affect a few users on an occasional basis to more serious bugs that demand correction on an as-revealed basis.
In most cases, vendors require customers to report bugs anonymously, either over private email or the phone. After all, why embarrass sales, marketing or development with the underbelly of a product's deficiencies? Check that. Why tempt customers or prospects to doubt the quality of a product?
(Wait. Here's an idea! Why don't we charge users fifty bucks a month or a buck-a-minute on the phone for the privilege of allowing us to fix our own mistakes? And, no, I have nothing against this. Let the buyer choose and beware. Just keep it in your mind here for comparison).
No company handles its beta product process perfectly. Still, I have watched (and, at times, participated in) Userland beta programs. The developers work maniacally to find and solve the bugs uncovered by the testers. I can't remember a product release that left show-stoppers in place.
The same commitment marks Userland's correction of faults after the software is released. The discussion groups provide a trail of evidence to this effect.
The discussion groups illustrate another dimension of the support equation as common to Microsoft as to Userland. We users frequently 'see bugs' where none exist. Either we didn't read the instructions provided or we demand that the software do something for which it wasn't designed. No harm, no foul on our part but this isn't the fault of the vendor either.
This does not mean that Userland support covers all the bases. The documentation supplied is fine but it is somewhat haphazard and ill-organized. Though the company does a swell job of matching supplied features to a vision of what those features can be used to build, a multi-hundred page documentation gap splits those two dimensions.
Now, keep in mind that this forty-buck product incorporates a full-blown web server, scripting language, multiple databases, an outliner and its own rich user interface and end-user feature set. If Userland waited on its production of a thousand-page user manual, you and I would be complaining that they moved far too slowly to satisfy our lust for the cool features they deliver on a regular basis.
You can't lose for winning.
Using Those Bootstraps ...
While Userland's vision of a collaborative writing and publishing environment has remained constant over many years, the company calibrates that mission continually to two vectors - what its users actually want and what the broader infrastructure (e.g., the Internet) can support reasonably at a given moment in its own emergence.
Userland's products are visible to developers and users down to the source code that drives them. Userland's developers derive real pleasure from other developers taking that code and extending it in their own directions. Royalties are not solicited, though the intellectual assets encapsulated in Frontier and Radio are enormous.
In many respects, Userland is one ongoing, floating, virtual seminar on the direction of the Internet and the Net's ability to sustain a collaborative development and use community. (And, psst, to sustain a collaraborative development and user environment .... can anyone say Frontier-Radio?)
Those who misconceive bootstrapping as beta fail to realize the integral connection between this process and Userland's shipping products. How better to enhance the products (which work perfectly well right now, thank you) than by encouraging developers and users to kick the crap out of them in the context of right-now debates among the world's smartest programmers over scripting, RSS, XML, SOAP, WSDL and more?
To users like you and me, these debates may seem silly or merely hopelessly arcane. But when we click the "News" item at the top of our Radio Command Menu, we are reaping the benefits of five years of debates over RSS as they have shaped real features in a shipping software product.
With its founder still at the helm, Userland has escaped the curse of full-blown open software development (the lack of a single, controlling hand). Steve Jobs, Bill Gates, Ray Ozzie and a host of other real-world examples illustrate how crucial this direction is for computing. See Henry Ford, Walt Disney and Winston Churchill for related illustrations in other fields. Even Linux would have been dead-in-the-water without the early direction of its Linus.
(Memo to Dave: take a vacation, don't forget to exercise and please put a succession plan in place that includes the next Dave Winer ....).
This said, Userland puts its ear very close to the ground to hear what its developers and users want right now from the company. As with support, Userland conducts this in-the-raw. It isn't restricted to partners with big bucks or developers who have ponied up the requisite loyalty - though loyalty is recognized and rewarded by Userand as it darn well should be. The community is reasonably polite but it withholds few punches. Dave Winer has come away with some bloody noses himself.
When Userland recognizes the need to bootstrap a protocol (XML-RPC, SOAP or OPML), the company gets started on its own. As emails, discussion forum brainstorming, face-to-face meetings and industry contacts bubble up, the company has demonstrated a willingness to cooperate with the efforts of others - even a willingness to support competitiors (cf Blogger).
Of course, this coincides with a determination to defend its own mission and go its own way when Userland believes it vital. It doesn't genuflect to committees. The sometimes rancorous debate over the RSS protocol illustrated this. When all is said and done, Userland is a business, not an open source vendor. This infuriates some but the company presses on stolidly through the firestorms as it must.
The foregoing was not written to canonize Userland or Dave Winer. Both have fallen off the high-wire and been barely saved by the Net underneath. Pun intended. Apologies have sometimes followed rants, both inside- and outside of Userland. Got it?
Let us overlook the hilarious fact that Radio is worth ten times its forty buck price.
Alas, the market is barely ready to pony up forty bucks these days. Userland shrewdly recognizes that. Since value is no abstract notion, forty bucks is just about right. At that price, Radio may become this decade's Pascal - a bone thrown to those of you with graying hair who remember Philippe's then-notorious strategy and the Byte ad that launched it into legend.
To the point: I know of no show-stopping bugs in Radio. This ain't beta software by a long shot.
The foundation on which it rests (Frontier) is a decade-old product that is amazingly robust considering the enormous scope of its underlying services.
(I will let pass the remarkable ability of Userland to deliver product simultaneously to PC and Macintosh users. I will even skip over its leading-edge delivery of near-daily fixes and updates to us over the Internet. Well, I won't entirely skip it: this dimension of support is a nightmare for the computer industry. Most of us Userlanders have grown so accustomed to our daily 'fix' that we take it for granted.)
Of course, Userland's core mission - bootstrapping the world's best writing-and-thinking environment on the Internet - leads to some false (product) starts as well as brilliant innovations.
Some Userland features end up orphaned; others proliferate in unexpected directions. This delights some customers; infuriates a few. Nothing new here: have you used Windows recently or, for that matter, OS/X? Heck, have you used a computer?
Nonetheless, boostrapping must be distinguished not only from beta software but - finally - even from the notion of the "bleeding edge". Bleeding-edge products are not truly usable even if they avoid show-stopping bugs. Ted Nelson's Xanadu has been bleeding-edge for decades.
Leading-edge products are usable but often entail significant capital and business risks to those who adopt them. While Userland certainly occupies the leading-edge at many specific points, I am not comfortable with characterizing its products that way. It connotes a degree of adoption risk that is inaccurate.
I would rather say that Userland grafts leading edges onto a sound foundation of classically-proven components and services.
(Frontier-Radio's web server, scripting language and databases are fully modern but, happily, not bleeding- or even leading-edge).
Userland bootstraps new protocols, layers, services and features on top of this long-since-bootstrapped foundation.
Forty bucks. Sheesh. If this is beta software, then may Userland sign up a few million users as forty-buck beta testers. Without apology. With five hundred million on the Internet and double that expected in five years, why should that be an unrealistic goal?
If I can earn some bucks supporting a few of them someday, all the better. No apology there either.