Posted by Antonio
1 year, 6 months ago (Jan. 6, 2007)
Last year I was obsessed with 2006 being the year of the "casual publisher" which I guess it might have been if you buy the load of hoopla Time magazine heaped on us with the Person of the Year but I'm not sure we got as far as I would have liked. Sure YouTube is the absolute bee's knees and they really do deserve almost $2b of Google's money they got, but it's not because of the explosion in casual publishing they've enabled (it's because they did something really amazing in giving the bits of television that serve to hold our culture together permalinks that we could all email, IM, and post to our blogs).
And no, wikipedia— though even more awesome than YouTube— does not make for a universe of casual publishers (for instance, it's been 18 months since I last contributed to an entry). What Wikipedia did prove in 2006 is that the dynamic from the Professor and the Madman can be amplified 1000-fold with the Internet and broadened way beyond the scope of the OED.
We spent the year at Tabblo focused on the casual publisher. Our particular twist was that the distance between "photo sharer" and "casual publisher" was short and easy to cross if you could pre-select the type of person coming to you by giving them an app with the right attributes. Our angle was (and still is) collaborative story-telling with pictures and words, helped by a personal art director in software and the ability to go from online to print with a few clicks of the mouse.
The good news was that more than a million people seemed interested enough to at least drive by and take a look with a good number of them staying around to make some really outstanding stuff. The more interesting news however, is that we've still got a bit of work to do to truly make it casual and that this looks like a great year for that to happen.
By casual we don't just mean easy (which is where we'll be focused for the next few months) but also compelling, relevant, and right-sized. This last bit is the big lesson from all of personal publishing in 2006— and the hardest one to get right. For instance, blogs are really interesting and compelling but for most people they are the hammer of personal publishing; that is, sitting down to compose an entry like this one takes forethought, time, and sweat. I am afraid that is is still the same for tabbloing— giving people that much creative freedom amidst a community of artistes who live to look at each other's work necessarily implies that authors have to really make an effort to tabblo. We attempted to mitigate some of this by building a publishing platform where distribution could be varied with access controls (from 1 person to the world), and as the folks at Six Apart have learned, this does help a bit— but it doesn't solve the whole problem.
We also attacked it with different formats— defining some as quick and others as more involved and elaborate. We did this specifically with the introduction of books and other printable products that could also be shared virtually. This gives you two axes— intended audience size and type of output which does provide some freedom to do all sorts of different kinds of things depending on your mood and intent.
But there is still one big problem in getting the experience to the right size: the ease of publishing when the moment motivates you to do so. The mobloggers got this from the very beginning— blogging from your cellphone wasn't just about "hey cool I can blog from my cellphone!" but also about being able to casually publish during the holes in your day (waiting in line at the supermarket, riding the subway, driving). Unfortunately the experience is just too cumbersome for most device/platform combinations. That said, the product that did the best job of bringing the potential of mobile-based publishing this year was the phoenix that rose from the ashes of Odeo, Twitter. Sure, they're only aiming to help you answer the question: "What are you doing?" but everyone of those SMSes has just picked up a thread, an audience, and most importantly, a permalink (no surprise that this product is from Ev and some of the other ex-Blogger folks).
Even if Twitter solves it for the words that come out of SMSes though, I want to see it solved for photos, audio, video, even the data that comes out of my Nike+/iPod running partner. We'll be giving it our own unique twist for photos at Tabblo this year, but as 2007 bring all sorts of new device/data combinations, the opportunities are almost limitless.
To close: why the hell should we care? I would argue that we should care because it turns out that casual publishing is a lot more like voicemail and a lot less like commercial radio— it's an opportunity to make connections with the fellow tribesfolk you most care about in new and interesting ways. And as Barry Schwartz argues in his book, it is these relationships (and the constraints they bring with them) that truly make us happy and fulfilled.
And even if you don't buy that, Chris Anderson (of Long Tail fame) had another possible reason in the best answer in this year's Edge question with his "Metcalfe's law of minds." Essentially, it this type of activity that can make us smarter as a species more quickly than anything that came before it. This too should be plenty of reason why you too should continue rooting for the casual publisher in 2007.
Posted by Antonio
1 year, 2 months ago (May 5, 2007)
We just finished our first week as HP employees. Part of the transition meant moving from Cambridge to Marlborough (or Malboro as some folks spell it) to a gigantic building called MR01 which was originally one of DEC's buildings (before Compaq acquired DEC which HP then acquired in giant game of Russian dolls).
For fans of 24, the building reminds me of the bunker that Victor Drazen gets transported to during Season 1— gigantic concrete structure, built into a hill, and surrounded by grassy fields on all sides. It is quite the change from being right in the heart of Cambridge to say the least.
That said, the commute and the food— the two things that I was sure would annoy me the most in the first week— have been quite decent. Instead I've come to realize why Gmail, Meebo, and other communication/collaboration/productivity tools that are deployed as webapps are going to win big in the enterprise: because most of the IT impositions inside of large companies generally make life really really suck for regular employees. Two ways to get on the VPN, all web traffic proxied through slow caching servers, and more ports blocked than in Cuba are just a few of the things that make getting productive quickly really hard. Add to that the fact that all support comes from disembodied voices on monster conference calls and you've got a sure-fire recipe for insanity.
On the plus side, one of my absolute favorite things about MR01 is the deep sense of computer history lore that permeates throughout the entire place. This is the case with all of HP— as not only the biggest but also the oldest company in Silicon Valley, it is just fabulous to visit the Palo Alto campus if you're into the birth of this industry. Digital Equipment (DEC) is like that but for the east coast; despite the fact that DEC totally missed the shift to PCs, its pioneering work in minicomputers laid some serious pipes for those to come later.
For example, on Monday night Steve Wozniak spoke at an event at the Museum of Science in Boston and mentioned how the old DEC PDP machine architecture taught him a lot about computer design and influenced his own work. Similarly, Bill Gates and Paul Allen wrote the first MS BASIC with heavy influences from DEC BASIC and on an 8080 emulator running on a PDP machine.
Then yesterday I saw Dan Bricklin at a Business 2.0 roundtable at MIT. When I told him about working in MR01 his eyes lit up because he'd worked there back in the early 1970s for the typesetting group. He told me that DEC back then was like Apple today and that the typesetting group was like the iTunes music store.
Tabblo is now located in a part of MR01 that looks out on the old helipads that were used by all of the DEC employees to fly between the various different facilities throughout New England (and as Dan told me, for the occasional boondoggle). Ned found a really old printout from the 1970s that had been typed up with all of the driving distances between the plants behind a filing cabinet. It seems that everywhere you look in our new home, you can see a piece of the old DEC— even the folks from HP REWS (real estate services) who go all the way back to the original DEC days and remember when Ken Olsen would buy wood at the lumberyard to make cubes.
Despite that few physical artifacts remain of the old Digital Equipment Corporation, its legacy stretches far and wide, right to this very MacBook that I am typing on now. As an inspiration for the end of our first week at MR01, you could do worse.
Posted by Antonio
1 year, 5 months ago (Jan. 29, 2007)
Well it seems like after the suspension of disbelief that was 2006 as far as making money from social sites, web 2.0 mashups, etc., we may see 2007 as the year when people start asking themselves that question in earnest. All of those eyeballs have got to turn into dollars after all, and thus far Rupert Murdoch seems to be the only guy who's been able to do it right.
At the end of the day though where all of these social sites are concerned, you either have an audience that advertisers want to pay for, or a base of potential customers that can be incented to spend money. Google, Yahoo, and all of the rest of Web 2.0 fall into the former camp, and innovative new companies like Smugmug, Cafe Press, and Tabblo fit into the latter.
For our part, after having spent six months building an initial member base that was attracted to the Tabblo service (either because of the editor, the online/offline combination, or the quality of the products we make), we're now starting to experiment in earnest with products that will begin to motivate members to buy at greater frequency than most other output vendors.
Tabblo Valentines, which we are launching today, is an interesting first experiment in this regard for a few reasons. First, at $1 including shipping, it's a relatively low commitment to try. Second, Tabblo Valentines are an offering that builds on all of the core parts of the free service that we've built over the last year. The same editor, the same mix of photos and words, and most importantly the same focus on collaborative/social content creation. So it will be really interesting to see whether these elements when combined get us to something more interesting than "yet another site offering some Valentine card gimmick." Finally, the inclusion of a relatively small feature, auto-addressing makes the Valentine product something that is really interesting in the context of an online community.
Most people who meet online don't know each other's physical addresses— usually not a problem (and sometimes an advantage). However there are times when it really is nice to interact in meatspace which is why people still read real books and like to get real cards. This is what auto-addressing is meant to fix. By letting Tabblo be a blind intermediary in handling the addressing and shipping of the package, it becomes possible for members to strengthen their online ties offline without doing too much work or feeling awkward about it. And in doing so, we're also lowering the friction usually involved in sending real mail (a nice thing to do despite living online).
This is the first step in what is going to be one long experiment with our business model— focusing on the types of physical products that make sense for our online community— however this seems like a very timely experiment now that 2007 is here to stay.
Posted by Antonio
1 year, 5 months ago (Feb. 4, 2007)
All of this hoopla over the new releases of PC operating systems has been bothering me lately, mostly because it's made me think back 10 years to when Windows 95 was released, or even 5 to when Mac OS X first shipped, and think about why today the releases feel like such non-events. Just the other day I was struck by an eager teenager who excitedly tabbloed his Vista unboxing, and at the office they have been relentless about some Vista-related quote I gave an AP writer (mostly because I just really don't care about Vista or its impact on our "lab").
Even this year's MacWorld— usually replete with pronouncements about how OSX is going to change the face of personal computing by introducing features which are long overdue on most client operating systems— was bereft of even a single OSX-related announcement. Craig Syverson over at Venture Cast had an interesting thought on this that got me thinking along a different line: he basically argued that the future of the Mac OS ("Mac OS XI" as he called it) was really iTunes in that more and more of what is relevant to users (media, iPods, mobile phones) was going to start to flow through iTunes, whether on the Mac or on Windows. In this case then, for media hungry consumers, iTunes is the new runtime and everything underneath it really is just a bunch of buggy device drivers.
The geek in me bristles at that, first and foremost because of how closed the whole iTunes ecosystem (including the forthcoming iPhone) is relative to POSIX, Win32, and even OSX. Even the relatively closed cellphone runtimes appear open as a commune by comparison. That said, plenty of valiant startups are finding ways to hack into iTunes to extract metadata and insert themselves into the flow of media— so if the desire is there and the resulting user experience doesn't end up being too encumbered, I'm sure that enterprising and curious geeks will find a way to make Apple's media-based OS XI a more open system.
A more relevant reason as to why OS releases are no irrelevant (hello Marc Andreesen?) is that all of the interesting things you can do with computers now (outside of specialized content creation like programs, video, music, etc.) has little to do with the OS itself unless you count the bits required to run a fast (and standards-compliant) web browser. I first heard Tim O'Reilly articulate this 4 years ago in a talk on the web as platform, but it has become dramatically more true with the development of Web 2.0.
Just this morning, I ran across a YouTube video that really captures the underlying sense of wonder in the possibility of it all when we start thinking of the network as the relevant runtime for innovation. After watching it, I'm sure you will come away thinking that the real innovation these days is as far from a shrink-wrapped box of device drivers as internal combustion is from a barrel of feed.
I highly recommend this video to any of the new web naysayers who think that the biggest things in the web over the last 5 years have been the return of pastels, rounded rects., and XmlHttpRequest replacing hidden iframes. Though it is gimmicky at points, it does a wonderful job of communicating why the folks on the leading edge get so excited about what's been happening in Silicon Valley over the last couple of years.
Update: A sadder but potentially more impactful of example of why the web-as-platform is so much more compelling than 3-D shading on windows and buttons is the search for Jim Gray via the use of a lot of the work/collaboration patterns that have really caught on in the last couple of of years.
Posted by Antonio
1 year, 5 months ago (Feb. 23, 2007)
Though I am really happy to see Google experimenting with revenue models outside of advertising, yesterday's launch of their premium business applications left me thinking that they are either a) gleeful about the possibility of putting a stick in Microsoft's eye or b) not yet clued in to what makes things work on the web if they don't have a search box in them.
Software as a service (SaaS) delivered in the form of dynamic and rich web applications is a really great thing— in fact probably the most significant thing to happen on the Internet since the coming of the web itself. However this does not mean that every possible thing someone wants to do with a PC can be crammed down the HTTP-pipe and decorated with pretty Javascript effects. Below are some of the standard reasons:
FIOS-enabled homes and offices notwithstanding, there is still just too much latency on the Internet, plain and simple. This is the main reason why taking any old desktop application and webifying it usually doesn't work. Word is faster than Writely, Excel is speedier that Google spreadsheets, and no matter how hard we may try, iPhoto or Picasa will always be faster than Tabblo for manipulating photos. Therefore, if you've got an application which sits open on your desktop all day ready for you to edit, compose, and otherwise author 20-50 times a day as is the case with Word, Excel, and Outlook for most office workers, taking that online is a big bet to make.
There are things you just need calls into the OS for which most modern browsers do not give you access to yet. For instance, office applications often produce documents that people want to print, and yet most browser-based printing is still stuck in the stone ages (how many times have your printed Mapquest directions only to get that hanging chad of a last turn on the last page?) Ditto for really rich APIs into other sources of data that live on the local filesystem (Outlook pst files, OSX Spotlight index), or even for rich access to the local filesystem itself. Try as we might, if you have things that you need to either source from the local disk or output to the local peripherals (printers, CD burners), the browser is not your friend. Interestingly enough, Adobe's Apollo projectmay solve this (a big "may" though as many have tried).
Where the data lives is a pretty important thing to think through, especially if you are a business. Sure no one backs things up properly, and for sure your office is more likely to be burglarized, burned down, etc. than a world-class datacenter, but for most people knowing that the data is under their control is more important at times than the data itself. Sure, a service provider can encrypt the data and partition it accordingly, but in much the same way that big financial institutions with liability exposure don't allow their employees to use standard IM, I suspect most big companies will be leery of trusting Google with sensitive, mission critical data. On the consumer side this is also a problem, albeit one that is mitigated by the fact that most consumer data that makes is to the web is not as sensitive (and the fact remains that most savvy consumers see uploading the data to a service as a poor man's backup).
These reasons by themselves should serve as powerful motivators for thinking carefully about trying to move any major desktop application to the web. If an application is going to take on the encumbrances mentioned above, the benefit of living on the network have got to be very high— in terms of ubiquitous access, collaboration, or best of all, the symbiotic relationship that the application can have with its users as it grows up.
Ephraim Schwartz interviewed me about Tabblo recently and wrote a nice piece on this co-development process where I make all sorts of bold claims about the productivity improvements that come from having an online app that lives and grows directly based on the use patterns of the community around it (which is itself growing). That same thin HTTP pipe I mentioned as a problem above becomes your best friend in this regard— by serializing every action into a very digestible set of log entries for subsequent analysis, it gives one a much better real-time view of what is working and what isn't. And because the user-facing application runtime is the web browser, UI flows can be changed near instantly without having to redeploy to an installed base.
The overall mental image I have of the process is that of an acelerated time-lapse movie of coral growing according to its environment and the actions of its co-habitants.
The main problem with trying to grow a coral from a word processor, a spreadsheet, or an email client is that the specs for these products are fairly known and have been pretty fixed in users' minds for about a decade now. Witness Gmail— every time I hear complaints about it, it is from someone who comes from the Outlook worldview and is almost certainly around the most innovative (and web-native) parts of the interface. The patterns are already well-worn so it becomes difficult for the application/community to take on a life of its own.
Paul Graham wrote an essay a while ago about Web 2.0 that argued that Google was succeeding because it was aligned with the "grain of the web." I completely agree which is why it is strange that they would use the office suite as their first beachhead into for-pay SaaS.
Posted by Antonio
1 year, 4 months ago (March 1, 2007)
In further helping to middle the line between desktop and web, Adobe held an event this week to showcase the promise of the Apollo platform. For those that don't know, Apollo seems to be the Flash runtime bundled with an HTML renderer and Javascript VM in a desktop application harness. It's supposed to allow developers to write hybrid web/desktop applications that can run offline and take advantage of local services.
I continue to scratch my head over the Apollo platform mostly because it seems so incredibly ambitious and because I really do have a hard time understanding whether it is going to do anything meaningful for adoption (the real problem that rich client software which does not ship on the platform has these days).
My guess would be that the devil will be in the details— after all, the JVM could have owned this layer on the client 10 years ago if a) it wasn't a 10MB download, b) it didn't take 45 seconds to boot, and c) it didn't run and look like a pig. Instead Flash beat it to the punch because it was less that 1MB and much less ambitious. Less is more— which is what is also something that is weird about Apollo. Why dial the ambition back up when they could do so much by just continuing to extend Flash?
As I think about a site like Tabblo, I realize that we would just love to see Apollo deployed on every computer— our users constantly have to struggle with the pain of crossing the desktop-web membrane with their big photos, our engineering team is well-trained in the ways of the browser frame and ECMAscript/Javascript, and there is tons we'd love to see done with the local CPU that we could completely offload from our image front-ends. At the same time I have a hard time believing that if this was going to such a huge win for us, we wouldn't find other clever ways of putting a footprint down on the desktop (see this clever OSX-specific way to run a browser frame as a full flown application).
And while I am on the devil in the details, one huge win from Adobe betting on Apollo internally could be a properly instrumented version of Online Photoshop that sites like Tabblo could easily embed into our experience. We've had tons of inbound interest to partner with the myriad of "online photo editors," that are out and while the promise is always there, for the most part, startups are like mosquitoes looking to sting: not all that likely to work for all when done in groups. On the other hand, easily being able to integrate with the 800-lb gorilla is a tide that would truly lift all boats.
Posted by Antonio
1 year, 4 months ago (March 3, 2007)
I've been cynical about the One Laptop Per Child project ever since I first saw Nicholas Negroponte talk about it at the D conference two years ago. That said, I'm really happy to be wrong in that it seems like it's way more than a science fair project and that millions of kids may actually reap enormous benefit from this little $100 laptop that could.
This morning I was reading an article in Businessweek about the OLPC's innovative UI (which the guys tell me was apparently shown on a real X0 at Pycon this year [it's mostly written in Python]) which talks about how innovative the UI is because it abandons a lot of the WIMP metaphor we've been stuck with for the past 30 years.
While I am all for innovative interfaces, I wonder if starting with the world's poor is the best approach to take. It's not that I don't think that an OLPC alum couldn't pick up a WIMP interface very quickly (after all my first "computer" was a Milton-Bradley Bigtrac and that helped me understand programming probably more than any other single toy I had growing up)— it's just that building up a whole new set of interaction primitives seems like a big risk to take on an audience which is already taking on a lot in using the OLPC in the first place. It's taken 30 years to polish windows, menus, and gestures (which of course the OLPC team will no doubt leverage experience from) and it seems a little strange to dump it in favor of something radically different.
Plus, I'm not sure that fancy New York graphic design firms (the article talks about how Pentagram is doing design for the OLPC) are best positioned to take on this job. After all, even the folks at Apple cribbed heavily from Xerox PARC back in the day.
However, I've learned my lesson with the OLPC and my pragmatic cynicism, and am thus optimistically awaiting the release of Sugar and the X0.
Posted by Antonio
1 year, 4 months ago (March 3, 2007)
In reading a great piece on the future of social networking in today's NYTimes (via Pete Cashmore's great Mashable blog), I learned that Cisco has been snapping up "social networking" assets— a move that is being universally ridiculed in the blogosphere because Cisco is only apparently supposed to know about routers and switches.
This is indeed a bogus move but not because "social networking" is a perfect example of a feature and not a product. After all, Cisco is the king of buying features and making products out of them. The real elephant in the room on this one is that there is absolutely zero business model overlap here— and this is what will ultimately lead whatever assets they acquire to get marginalized, ignored, and finally written off as a hangover during that crazy "Web 2.0 bubble."
So far as I can tell, the only business model that works with social network sites, once they are at scale, is an ad-based media model (this of course begs the question of why you'd buy a social network site that is not at scale if you don't have the tools to bring it to scale). And I may be wrong here— but I'm not sure that I've ever seen Cisco's ad salesforce pounding the pavement on Madison avenue.
I remember that people laughed when Cisco bought Linksys and commented that the company didn't know the first thing about marketing and selling consumer networking equipment. But at least in that case, the business models (sell chips with RJ-45 ports, make money) was pretty complementary. In this case, I'm afraid of what the ultimate fate of these social networking folks is going to be 12 months from now. Unless of course, then can get good at writing router firmware.
Posted by Antonio
1 year, 4 months ago (March 9, 2007)
Jeremy Liew has a very nice piece for anyone interested in online media business models or even for anyone interested in the "build a web property that is popular and figure out how to monetize it later" approach to starting companies. It is a really short summary of the kind of traffic you should expect/need to get to build a big business based on ads depending on the type of content/audience you are hoping to aggregate.
I like the piece for two reasons. First, most consumer Internet entrepreneurs I meet these days do too little thinking about business model, preferring instead to think that either Google or Yahoo will acquire them or that they will just do the "ad" thing. This is the west coast's "build now, monetize later" investment hypothesis taken to the extreme. To that end, the piece does a great job of showing that banking on this approach has the inbuilt assumption that your web property is basically going to be struck by lightning. It happens— but you shouldn't count on it.
I also really dig the post because in looking at the third model for success— vertical content sites that are best positioned through content for certain types of advertisers— I realized what it is that I dislike about the advertising bet in startup business models: as an entrepreneur, it reduces your optionality (is this even a word?).
Having options when the initial plan doesn't work out is a great thing, especially for small companies in quickly developing spaces. But as I looked at the third, and most viable of the ad models, I realized that there is little option if the vertical content you create fails to attract advertisers or a significantly attractive audience. When that happens, you don't have the scale to just pump generic ads (model #1), and you don't have a model that will continue to itself scale without continued investment. You can try a roll-up across verticals but something like this only gives you scale on the sales side of the business and not on the core value propostion.
If you read the comments (and some of the very interesting posts linked therein), you will see people discussing alternative business models that may work better at maximizing optionality in the face of failing to reach a certain scale. My two favorites are product sales (preferably mass customized like Tabblo's because these avoid pricing pressure best) and subscriptions (but very high value like SmugMug's or eHarmony's), but there are others including "freemium" (though you need a little bit of scale to make this one start to work) and marketplace (where you need a very narrow niche or even more scale).
All of these other models seem to offer more levers to tweak when things don't quite go according to plan. Which is almost always— and a good thing to keep in mind when looking at some of the high hurdles that Jeremy lays out.
Posted by Antonio
1 year, 4 months ago (March 11, 2007)
I recently got a chance to read Scott Rosenberg's "Dreaming in Code," a story about the ups and downs (mostly downs in this case) of the software development process of Chandler at OSAF. The book is a great introduction to challenges of software engineering, especially for outsiders, though it is nowhere near as good as Tracy Kidder's "The Soul of A New Machine" despite the fact that most reviewers seem intent on comparing the two.
One of the central themes that Rosenberg tackles is that of the "software crisis," or simply put, software engineering's inability to scale the production process in the same way that other engineering disciplines have— most notably, hardware engineering but also mechanical engineering, chemical engineering, etc. In all of these other fields "scale" refers both to being able to linearly apply more manpower while achieving linear or super-linear boosts in productivity, as well as producing less error prone and more predictable output.
This question is of course near and dear to any software person who has ever missed a schedule or over-scoped a feature, so naturally I was drawn to the bibliography where I found an incredibly interesting, albeit dated, book by Brad Cox (the creator of Objective-C) called "Superdistribution" on exactly this topic of scaling software engineering. The book was published in 1996 and is now out of print but I would highly recommend it to anyone interested in software engineering as a fledgeling discipline.
The core of Cox's argument is that the notion of "Software ICs" (software integrated components similar to chips in hardware) has not taken off not because of technical reasons (as he learned after developing Objective-C for just this purpose) but because of economic ones. In essence, he argues that there is no economic advantage to the designer/developer of a software component that can easily be re-used (through good interfaces, cross-language bindings, good documentation, etc.) because the licensed-library model does not provide adequate compensation to create a thriving ecosystem, and because other models (such as per-unit royalties) have not been popular in the software industry for various reasons. Hence capable programming teams are left much more motivated to re-invent the wheel in the hopes of building the next MS Office and are less likely to make a good living at some lower layer (for instance, the word-wrap object, the pagination object, etc.) in the same way that their hardware counterparts have been able to.
His solution turns on a clever attribute of software but is nonetheless completely impractical. Software has no control over its being copied, Cox argues, but it has very tight control over its being executed, and as such, it could establish a payment mechanism based on "useright" rather than "copyright," metering its own use and charging accordingly. This seems counter to the grain of personal computing to me, and in fact the only examples I have ever seen of this model have been with really expensive software ($100K+) where the vendor leaves no other alternative. And of course this is to say nothing of the administrative nightmare that a system like this would be. Current DRM systems would look like "Hello World" by comparison.
So I don't like his solution but I do agree with his conclusion. To scale we do need more modularization, and the reasons for not being there are most likely economic and social at this point— and not because we're still looking for the one runtime to rule them all.
Coincidentally, it is interesting to see how in the intervening decade two trends have helped combat the coming software crisis that Cox predicts in his book. The first, the mainstreaming of open source, has created an alternative economic model for encouraging software re-use, though I suspect Cox would argue that this alternative economic model is not robust enough to survive or scale to what he is imaging. Still, you'd be a moron to try to write an RDBMS or a webserver at this point, and this productivity boost comes directly as a result of the open source production process. The second trend, which seems much more in line with what Cox argues for, is that of metered web services. Amazon S3, EC2, and their brethren are great examples of a pay-per-use model at work, and the fact that the bits are not being pushed down to the client seems to be a great way to solve the language/runtime interop problems that he discusses in the book.
Economic arguments aside, there is tons of good stuff in Superdistribution (if you can get past its datedness and its wonky illustrations). For instance, Cox covers how time-sharing systems have left us with a legacy of independent processes with individual memory address spaces and are thus difficult to integrate in a component-based way while the futuristic systems of the 1970s (Smalltalk in particular) tried to eliminate the distinction between the machine's runtime and the processes running in part to solve this very problem. This was sort of mind-bending to me as I've always seen this as one of UNIX's greatest features (at least when compared to DOS/Windows/Mac OS-pre-X). He also makes a wonderful attempt at building a vocabulary for layers in software and often uses rich analogies from other engineering disciplines to make his arguments.
But best of all, it is clear that Cox is an in-the-trenches engineer who has deep knowledge of languages, platforms, and building big systems. Thus the book avoids that "architecture astronaut" feel that you sometimes get from this kind of effort (sorry but the much beloved Go4 book still feels like that to me). If you are interested in this sort of thing, it's well worth reading.
Posted by Antonio
1 year, 4 months ago (March 5, 2007)
I really loved this post on O'Reilly Radar from Marc Hedlund, especially when he says:
One of my favorite business model suggestions for entrepreneurs is, find an old UNIX command that hasn't yet been implemented on the web, and fix that. talk and finger became ICQ, LISTSERV became Yahoo! Groups, ls became (the original) Yahoo!, find and grep became Google, rn became Bloglines, pine became Gmail, mount is becoming S3, and bash is becoming Yahoo! Pipes.
It makes sense if you think about it: UNIX was after all the first network-native multi-user (social) operating system. All of those users back then in the 1970s were living William Gibson's future: present, just not widely distributed, so they developed patterns for work and play that most of us are still rediscovering today.
That said, I still really don't get Yahoo Pipes. When Yahoo launched this mashup application a few weeks ago, the entire blogosphere lit up like a Christmas tree with raves about how this was bringing the ability to program the web to the rest of the non-programming world. Just like OO and software components were meant to 20 years ago and just like Applescript and Automator were supposed to over the last few years.
I tried two Pipes. The first one (which took me the bulk of the 40 minutes I spent playing around) was meant to test a theory that Jon Udell told me about recently regarding common identity across various web services: Jon said that one way to tell who was who was to look for common usernames like fetching or supergeek57. So I tried to wire a Pipe that would take one of the recent tabblos RSS feeds, strip out the username, and do searches for same-author Flickr photos:
No dice though. Despite having a Fetch RSS component and a Flickr-specific component, I just could not wire the canned components in the right way. In Python this is a 7 line script but in Pipes, it's a whole lot of cool AJAX dragging and no satisfaction.
Then I figured I would do something simpler and take the Huffington Report's heinously monster RSS feed and filter out Chris Kelly (my favorite blogger from there):
This was much easier to do but I've got to admit that there have to be easier ways to get this done than by building a Pipe.
Unix worked where Yahoo Pipes doesn't because people wrote a large ecosystem of small programs and because the real pipe (|) was more loosely typed and flexible than the Yahoo Pipes inputs/outputs. The first problem could be solved with lots of programmers writing Pipes for regular folk but I have a hard time seeing that so long as it is possible for these same folks to whip off quick Perl scripts that do the same. And without the rich ecosystem of Pipes, I'm not sure the inter-connects will ever get flexible enough.
So while I agree with Marc on the overall insight, I think that Pipes is not the web's replacement for bash. I do think that replacement exists though, and its arrived in a really weird first incarnation: MySpace.com.
Sure it is the uber-daddy of social network, but to me to most compelling part of MySpace was the way that it allowed regular folks to copy and paste bits of HTML and Javascript into profile pages in an interactive way that UNIX folks from the 1970s would have been proud of. And by the look of some of these profile pages, it is unlikely these folks stitching random bits of execution together are particularly technical. Just particularly motivated.
So Marc is right— UNIX is getting re-invented on the web, mostly because the problem domain is the same and the social urge is still there. It's just not the geeks anymore which means that the form it takes is not always a clear and obvious mapping of the UNIX of old.
Posted by Antonio
1 year, 4 months ago (March 13, 2007)
I was nearly in tears during parts of this cartoon on Steve Jobs and Bill Gates:
My favorite part is Bill Gates on the music prevalent in most of today's PC/Mac ads: "What's up with that music?!? It's like baby music..."
I feel pretty lucky to have grown up during the phase of computing when the titans were Bill and Steve, given that they are both such colorful characters and so distinct from each other. 20-30 years earlier it would have been the Watsons from IBM or the wacky cowboys from the minicomputer revolution. Worse yet, if you were coming out of school today, the titans would be the Yahoo guys, the Google guys, or maybe even the guys from Facebook. Bright and driven, but not terribly distinguishable from each other. I can't really see them inspiring a cartoon like this one— can you?
Update: In digging around on the source of the cartoon, I discovered that it is a show called "Current News," by Josh Faure-Brac, which seems to be cartoon shorts of popular culture/politics hosted on Al Gore's Current TV network. It's too bad they don't serve the shorts as a video podcast feed, but some of the other ones are well worth checking out, especially the Oval Office one.
Posted by Antonio
1 year, 4 months ago (March 14, 2007)
Josh Kopelman (of First Round capital) has a nice summary of the argument around how Web 2.0 allows investors to place cheaper bets including some data on how much cheaper it is to launch a web startup today. This means that failure comes faster which he argues is a good thing both for the investor and for the entrepreneur.
Largely I agree with argument he makes, but— as is the case with most hypotheses when they pass into the land of "conventional wisdom—" it turns out to be a little more complicated than just: open source, commodity boxes, off-shore development equals a $200K lottery ticket to riches. I think that this model works very well if what you are doing is taking a social media bet whose main goal is to amass users and flip to a big portal before getting to either the "mass market," or more importantly, the monetization.
Flickr, Delicious, even MySpace— all of these guys fall right into this model as do the 50-100 companies that have been micro-funded like by Josh and others. But what happens when Yahoo, Google, IAC, etc. stop buying? How much of the critical risk of building a business is really removed by initial geek traction? Josh himself in an earlier blog post has a brilliant write-up of why it is so hard to get Internet consumers to part with money. Shouldn't this be the key risk to remove before getting to this key valuation inflection point?
I am biased because at Tabblo, I decided to take a hybrid approach: raise an A-round that is reminiscent of the 1990s ($5MM from Matrix partners) but operate the business as though we really had only about a fifth of that. There were two main reasons for this:
I knew we were going to have to build a bit of technology and felt that we needed a pretty unencumbered 18 months to do so with a decently-sized engineering team
I wanted enough cash to get through proving the monetization phase with what we considered a "mainstream" consumer.
Neither of these goals lend themselves to a "failing fast" test (if by fast you mean 3-4 months) so it's pretty hard to do things the microfunded way. And as far as the Tabblo approach is concerned, the jury is still out, so stay tuned.
Posted by Antonio
1 year, 4 months ago (March 22, 2007)
Neil Gershenfeld has this really neat book called Fab about the coming era of "personal fabricators" which are sort of like printers for 3-dimensional objects. In it he argues that soon we will be using simplified CAD programs to design our own objects which we will then just print at home for use.
I think a big part of me loves the idea at the "Star Trek gut level" much in the same way that people loved the original Motorola Startac because it subconsciously reminded them of Captain Kirk's communicator. In the case of the personal fabricators, we're bringing to life the replicator— Gene Roddenberry's utopian answer to renewable resources poverty.
In the meantime, we've got the Tabblo PhotoCube :), the latest product in the Tabblo family. It's a personal favorite for two reasons:
First, it's the first product we are asking folks to build (print, cut, assemble) at home. Despite being outside to our original business model, I think that the PhotoCube captures a lot of the immediacy of the web in a much nicer way (see pictures I like, make them instantly into a 3-D physical world object). It does have that "some assembly required" quality, but in the world of makers, this is a feature and not a bug.
Second, the PhotoCube is great because in being 3-D, it does give one the illusion of being a sort of "matter transporter" where you can literally recreate the thing you've photographed as an object that sits on a desk. Obviously it's not the same dog/car/beach that was photographed, but it is fun to see 2-d facsimiles built back into three dimensions.
Posted by Antonio
1 year, 4 months ago (March 22, 2007)
The Acela, Amtrak's high-speed train that runs the length of the eastern seaboard (and my favorite train ride in the world), makes an interesting stop in New Haven Connecticut. As you are heading north on the train, you will spend about 5 minutes stopped there during which the train will jiggle this way and that while something is being done to it. And then, zoom! the rest of the way to Boston, the train speeds at what feels like 150% of the speed that you had coming up to New Haven.
I've never know exactly what happens at that stop but in my cartoon imagination, I've always thought that the Amtrak guys must strap some sort of rockets to the top of the train.
Five minutes ago, a press release went out that said that the Hewlett-Packard company is acquiring us to execute on their web-to-print strategy. Details of the deal were not disclosed, and I am not going to elaborate here, save for three facts:
We (Tabblo and our investors) are very happy with the outcome
We (Tabblo and the Tabblo community) are going to stay on board at HP executing on our combined strategy and building great products
We (Tabblo and HP) are incredibly excited because it feels like someone is strapping that rocket to the top of the car
I think #3 is the reason why I am not simply feeling the "pheew, we're done" emotion that entrepreneurs talk about when they sell their companies. On the contrary, I'm itching to see just how fast we can make this train go now that we've got a rocket to boost the "print engine of the web" strategy. So stay tuned there.
On a more personal note, I could not be prouder to be joining HP. Growing up in Venezuela, HP and Apple were the two companies that represented everything that was great about the American entrepreneurial ethic both in terms of ingenuity and hard work. I still have the HP-29C my father gave me when we first got to the US (its programability got me into college), and have very fond memories of my struggles at Terman (the engineering building at Stanford that Hewlett and Packard gave in name of their mentor).
In short, I am really happy to be merging Tabblo's future plans with such an incredibly rich and long-lasting institution, and I'm really looking forward to what the future holds for us.
On that note, if you are from HP, please don't hesitate to drop me a note to introduce yourself at antonio -at- tabblo -dot- com.
Posted by Antonio
1 year, 3 months ago (April 8, 2007)
In what is a great example of how Digg/Reddit have made blog post titles irrelevant, Paul Graham has a piece called "Microsoft is dead" which is really much more about how they are now irrelevant. The thrust of his argument is that Google and AJAX make their platform irrelevant for new hackers.
Fun stuff, appropriately lambasted by wizened veterans like Dave Winer and Don Dodge and put properly in the context of "Rich Internet Applications" by Ryan Stewart. So rather than plumb the depths of the argument, I wanted to keep the plane at 30,000 feet:
For a platform strategy to be displaced (and this is what Microsoft has done brilliantly up until now), you need to have a new platform strategy and a viable economic model.
Are AJAX-enabled web applications enough? Not by themselves. It's cool to have a much richer experience in the browser but I doubt that this alone is discontinuous enough. Outlook web access was after all, the first AJAX-enabled web application, made by the very folks at Microsoft who this new platform is supposed to make irrelevant. Despite the portability advantages of data-in-the-cloud, turning the world of PCs into dumb terminals doesn't even necessarily feel like it is altogether a step forward in the evolution of personal computing.
However, there is something else that makes Web 2.0 a much better candidate as a full on platform: people. Tim O'Reilly has been writing about it for years, and plenty of others have jumped in; collective intelligence, smart mobs, folksonomies— however you label it, the idea that people can come together and trade, write encyclopedias, etc. is really powerful. I don't think we even know how to understand what this new people platform is really defined by, but we can see its outlines in examples emerging from the wild. For instance, Twittervision is a great example of an app built on the people platform.
That said, we still need the economic model to give the platform enough time to grow its ecosystem. Microsoft has had wonderful one here in software licensing. Open source has a pretty good one in the gift economy and ego/reputation (albeit not as good as Microsoft). Web 2.0 most definitely does not have a good economic model in "selling the feature to Google/Yahoo." Furthermore, it's not clear that the ad thing is going to work, especially as we see increasing returns to scale in the online ad networks, and try as we might, no non-content provider has yet to crack the subscription model on the consumer side.
Without a viable economic model, it will only be a matter of time before Microsoft and Google incorporate the features that make the people platform into their respective businesses and leave a vast wasteland of cynical folks capable of using pastels and rounding corners on rectangles. This stuff is really hard to get right; you need the platform shift (timing), the economic model (business), and the emergent ecosystem (developers) to work together for a long stretch of years. And no matter what the cool kids at Y-Combinator are worrying about these days, I'm not sure we've got all of the key ingredients in place.
As an aside, if you want to see a great example of someone who got this right on the money the last time around, go and listen to this long-ish talk by Bill Gates from 1989 when he talks about the PC ecosystem. He must have been in his mid-thirties at the time, and it is just amazing to hear how well he can navigate the level of detail up and down to paint a really cohesive vision of where personal computing was going back then.
Posted by Antonio
1 year, 3 months ago (April 9, 2007)
There was a great article about our new owners in today's New York Times which is a great read if you are interested in the internals of HP and the future of printing. Of course it elicited a bunch of emails from my joker friends wishing me well in what seems to be an epic battle as people consume less paper and more screen.
What is most interesting to me is a point which VJ makes in the piece about how most companies don't innovate when they are on top, but rather when they are struggling at the bottom of the heap. One of the most exciting things about Tabblo being part of the new strategy is that we get to take part in the strategy from this rather unique vantage point.
Today Apple shipped its 100 millionth iPod. This is a wild and crazy success and yet it's interesting to think that HP's installed base of home printers is significantly larger than this. Is Apple thinking about what comes after the iPod and are they willing to swing hard to get there?(1) Is Google making acquisitions around what might come after AdWords? Does Microsoft really want to make a product bet on web-based software?
Christensen's Innovator's Dilemma is supposed to be business-101 for anyone working in technology. It's good to see that my new bosses have actually read it.
Posted by Antonio
1 year, 3 months ago (April 25, 2007)
Having spent a bit of time thinking about developing APIs since the HP acquisition took place a few weeks ago, I was very amused by an ACM Queue piece, "APIs with an Appetite," which considers how APIs can get too fat too quickly.
In essence the piece argues that a well thought-out API will have just enough method calls to provide the intended functionality plus an "escape hatch" call which may be ugly but which plays a critical role. The example given is the UNIX file-access API with its 5 basic types of calls: open, close, write, read, and ioctl, the last being the escape hatch call which lets one pass almost any data structure down in the guts of the operating system and get almost anything back.
I've always thought of ioctl as the ugly stepchild which somehow broke with the elegance of the "stream of bites" abstraction of its four siblings. And yet, as the article argues, it is precisely ioctl's ugliness that allowed the rest of the API to remain so pure while still actually being useful.
It's interesting to consider this design approach in the newly emerging set of web service APIs. The Flickr API is the posterchild for the new web and yet, while the API is quite useful, the author of the ACM piece may consider it "fat." After all, there are a bunch of repeated methods for dealing with collections of things: photos, people, sets, etc. And yet, each of the method families belongs to a different namespace and returns slightly different data structures.
This short piece is definitely worth reading for anyone working on web service APIs. The final point about monitoring the use of the escape hatch call and learning from it how to evolve the API is perhaps most applicable in the web environment (where everything is centrally logged for later study). As the coolness of just having an API fades and is replaced with an explosion of services and APIs, we'd be well served to take this piece to heart.
Posted by Antonio
1 year, 2 months ago (April 28, 2007)
The other night I was up with a heinous toothache and decided to consume some content to forget about the pain. Because the television was borked (HD TV is some sort of early adopter hell these days), I started cruising old episodes of Ze Frank that I had missed and discovered Jonathan Coulton, a kindred spirit of Ze's who set out to do a song a week for an entire year in his "Thing a week" project. Or I should say, I rediscovered Jonathan, as I had first been exposed to him through his hilarious (and great) song, Code Monkey.
I am in awe of guys like Ze and Jonathan not because of how talented they are but because of how hard they are willing to work exploring new means of connecting with their audiences. They truly are the new media entrepreneurs— except that in their case there is no clear path to riches as is the case with most of the folks starting the Flickrs and YouTubes of today. Instead they work between two elephants dancing: 1. the big media companies with their constant threats of "sizing down" to match the new forms on the Internet and 2. the peer production orgy that drives sites like Wikipedia.
In fact, in all this talk of the everyman getting access to the means of production and distribution— after all, anyone can be a record producer now— what is often lost is that most of this user-generated content actually sucks and wouldn't entertain a late-night toothacher for more than 2 minutes. It is an interesting social phenomenon for sure, but ABC has nothing to worry when it comes to the entertainment value of the content created by the average Joe videographer.
The Zes and the Johns? Maybe. Probably if someone figures out how create a viable revenue model that doesn't require them to sign the big record deal or the tv production show contract. And certainly if features like podcasting and products like the Apple TV gain mass market traction. And I for one, can not wait!
One final note: I noticed that Ze went to Brown and John went to Yale which got me to thinking about all the folks that I knew in undergrad who were equally talented and creative but who had not taken to the net with quite the same zeal. It's not for lack of talent— I think the real barrier is sticking your neck out creatively like these two have done. Fortunately as they do so, Ze and John are really paving the way for the torrent of talent that will undoubtedly come behind them.
[Note: to see them both together, go and watch this episode of The Show— a fabulous inside new media joke]
Posted by Antonio
1 year, 2 months ago (April 29, 2007)
I was really looking forward to getting into the Economist's latest survey on wireless ("A world of connections") this weekend. But about halfway through it I realized something that just soured the rest of the experience. Though it was written in 2007 with the central hypothesis being that every "thing" is soon to be connected wirelessly to the Internet, tis very same survey could have been written in 2002. Bluetooth is just around the corner in terms of gaining mainstream adoption (and more importantly, mainstream usability), cool whizbang companies like Ember are just about to deploy millions of wireless sensors (just like they were at the end of 02), and before you know it, your refrigerator will be ordering its own water filter over the Internet from GEappliances.com.
It all reminds me of the old adage, "Don't confuse a long view for a short distance." The Jetsons world of wireless is still not here and rather than drink the kool-aid, we're all well off to remember how little progress has been made in the last five years— possibly because there may just be something quite fundamental about the every appliance Internet-connected model that the industry has been preaching for a while now. COGS that are too high? Too much configuration complexity? No real compelling usecases?
I was surprised to see though, that the absolute best non-WIFI wireless device that I have seen and experienced over the last 5 years got barely one mention in entire piece— the Apple/Nike+ sports kit which effectively combines a wireless transmitter in the sole of your shoe with a receiver docked to the bottom of your iPod so that you can track your run telemetrics closely. I've written about it before, but I just think the world of this general architecture for deploying cheap, semi-Internet connected devices that capture really useful (and fun) metadata about our interactions with the physical world.
The device is cheap ($29) because all of the heavy processing (including the connection to the cloud) runs on what is effectively subsidized infrastructure. First, you've got the 200 MIPS of the iPod to capture the telemetry, and then you've got the harness of iTunes to get it up on the network.
But best of all, the effort of docking the device to the network is subsidized by another activity as well— the act of getting the latest music/podcasts on your iPod before the workouts. What a great way to parasite an already popular activity. Social engineering at its very best!
It bears mentioning that Nike has done a brilliant job of making the application a lot of fun by leveraging competitive tendencies (for instance, virtual challenges with friends are a huge source of entertainment and motivation). However, at the end of the day, the architecture is what bears close study, and copying where appropriate.
Posted by Antonio
1 year, 2 months ago (April 30, 2007)
Today was our last day as Tabblo employees and my last day as CEO. To mark the passing of an era in the history of this group, we continued pushing fresh versions of the site (as we do every day) up to production even as the movers were taking our office out to Marlborough one box at a time. In fact, I had the distinct pleasure of pushing out the last version of the site (including the soon to be announced "Caption of the Day" minisite) to go live from 810 Memorial Drive. And to commemorate the moment, I managed to clobber the private keys we use for storing credit cards on all of the application servers thereby rendering the shopping cart completely useless.
I mention it not because it was a fairly idiotic thing to do (it was) but because it was a bittersweet moment for me. One of the best things about working within a group of a dozen people is that every single person counts— a lot— and the fact that right up until the last day of Tabblo, the CEO had the keys to push builds out the door speaks volumes to the amount of cooperation and trust that we have all built with each other over the last two years.
As folks scrambled to retrieve keys from the offsite backup and fix the site and I walked on to the bar where we were scheduled to have our last Cambridge beers, I thought of how there are jobs you have where it is crystal clear why you get paid, and then those rare ones where you think you've pulled one over on the world because someone was silly enough to pay you for showing up every day. Running Tabblo has been from the latter category for me— mostly because of the folks involved. After all, at the end of the day, the only thing that is more rewarding than knowing that you are making a difference is the feeling that you've helped create a place where everyone can.
Posted by Antonio
1 year, 2 months ago (May 7, 2007)
A few years ago when I was talking to him about commercial opportunities in open source, Bob Franskton said something that has stuck with me ever since: he likened the open source geeks to the CB radio fanatics of the 70s— as soon as they finish investing a ton of time getting their Linux machines all set up and connected, all they can do is spend days on end on the Internet talking about their respective setups. It reminded me of growing up with my ham radio dad who'd scramble all over the roof to get the 50 foot antenna perfectly aligned... just to talk to some Australian about aligning antennas.
I think this applies to social media properties as a predictor of mass market appeal much more than it does to open source. Let's call it the CB radio index and define it as the amount of communication/activity/etc. on a given social platform that is about how cool it is to be using that platform relative to total communication/activity/etc. As this number approaches 1, you get an increasingly more niche product that is less likely to appeal to your mom, "Emily," or even your lawyer college roommate.
Here is a great example of a high CB radio index web property: Twitter. Every time I dial up the volume on my Twitter feed I am amazed at how even veteran twitterers can spend an incredible amount of bandwidth writing stuff like "cool, twittering from dentist!"
And an example of the opposite: YouTube. I have never once been sent a link to a YouTube video that is about YouTube (with the painful exception of the acquisition video the founders made).
Now Twitter and YouTube are very different in many ways— but as social media platforms— as examples of the People Platform, I believe they are similar enough to be good tests for the CB radio index. It would be interesting for some grad student somewhere to try this analysis on most of the Web 2.0 sites out there today to see if it indeed has any predictive power.
Posted by Antonio
1 year, 2 months ago (May 9, 2007)
You know I need more original blog titles for my posts when the New York Times (the paragon of the trailing edge) is using the same Star Trek jokes to talk about the coming of 3D printing in this recent piece by Saul Hansell. The article essentially covers the startups that are bringing cheap 3D printers to market (where cheap is $5k-$10K, but remember the first laserprinters) and makes a few other points about this technology's broader applicability.
Neil Gershenfeld covers what this is going to mean for regular folks well in his great book Fab— personal fabricators in everyone's home. I'm guessing though that before we get there, we are going to see all sorts of booming businesses that adopt a service provider model much in the same way that the first laserprinters gave birth to a whole slew of Kinko's copycats that basically rented the printers to users on a per-page basis. This is likely to be the case for two reasons: first, even if people shelled out $5-10K I doubt anyone is going to want to wear/use/display they "gray objects with a rather sandy finish" that the low end personal fabricators will be outputting for a while. More significantly though, the web is the ultimate demand aggregation channel for a service bureau business that can make you a custom version of whatever you want.
Sites like Threadless have already proven that user-led design powered by service bureau can be a great opportunity. Now imagine what that can become once you step outside of t-shirts and into just about any physical product you currently buy at the store. Forget about replacing your cellphone cover, and think for a moment about designing custom toys for your kids, custom wardrobes, or just about anything else that currently exists in your life and is made out of simple materials (it's going to be a while before we're printing robots). Or better yet, imagine remixing other people's design in the way that we can now easily remix web content.
What is going to be even more interesting is when we start to explore the possibilities around creating objects that were not born of the physical world in the first place. Hansell mentions Second Life in the piece— so imagine for a moment if we could start seeing the wacky objects that people build for themselves in-game but as part of the wardrobes we see in the physical world.
Posted by Antonio
1 year, 1 month ago (June 7, 2007)
I spent this past week in Europe visiting HP sites, partners, and vendors and was taken aback by how much connectivity we now take for granted in our regular work here in the United States. The last time I was in Europe for work was 9 years ago (when we'd be amazed just to be able to dial up from the hotel room), but I guess with all the talk of broadband being much better overseas and the 3G cellphone networks being ubiquitous, I was sort of expecting at least parity with the setup I've got here. Boy was I surprised. Here are some of the things you can expect to go wrong:
Power (this one was on me for being unprepared). Between Spain and England, I discovered 3 different plug types. What is more, if you travel with a laptop and a phone (more than one device to plug in) and check in late, good luck getting the hotels to have anything to lend you to plug your American appliances in.
Consistent SMS/data on your cellphone. Having just switched to a GSM network, I was really excited by the prospect of 3G networks and zippy-fast mobile data. While voice worked everywhere, SMS and data did not. In fact, SMS was the flakiest of all of the services that I've come to rely on— I could receive messages almost everywhere, but I had at best 50% odds of being able to send them.
EVDO and Wi-fi. I knew my EVDO card wouldn't work at all; I guess I just hadn't realized how much I rely on it to stay in touch with the team back in Boston while waiting at airports, in cabs, or in dead times between meetings. And as far as the wi-fi is concerned, it does seem to be fairly ubiquitous, but in 100% of the cases it was expensive and encumbered by either its billing mechanism or by some lame proxy server setup that blocked most of the useful Internet services you'd want to get access to.
Overall Internet speed. Finally, the speed of "broadband" connections (especially in Spain) is painful. In this new world of rich Internet applications, it's easy to forget that we've only just been able to get to the point where we can use them in the US and that this is far from a given for other parts of the world. For instance, in Spain Tabblo.com was completely unusable, and even Gmail was severely hobbled by the dearth of bandwidth.
I'm not just writing this to gripe about my technical defficulties with the "old world," or even to warn other innocent travelers, but because I realized for the first time this week how much I've gotten used to filling the small cracks of free time when I am traveling with little bits of collaborative work and communication. It made me think that the macro effect productivity boost must finally be something that economists can measure as a positive impact of PCs and mobiles— always-on connections to the data and people that you need to get stuff done.
And yet at the same time, it was sort of nice to have a few cab rides and airport stays to do some disconnected work. Or even better, to just sit and think.
Posted by Antonio
1 year, 1 month ago (June 15, 2007)
Roger von Oech has a nice short post comparing his behavior when he reads articles online versus in print that makes the basic point that his relationship the content differs depending on whether he is flipping the page or scrolling with the mouse. Expectations, time spent, even conclusions drawn can vary, despite the fact that it's the same exact content.
The post underscores the premise that presentation matters, the fundamental underpinning of the graphics design profession. However, Roger also gets to something more: the commitment that a reader makes to a particular bit of content when he decides to consume it on paper. It is almost as though the act of stepping outside of the natural flow of screen life implies that a different mode of thinking is going to be used a different part of the brain engaged.
When we all became employees of HP, we were given this really sweet All-in-one printer to help us start thinking about how a printer is used in the home. In my case, this was the first time I've had a consistently reliable printer outside of work since I left my college dorm so I've been overconsuming paper and ink quite a bit. In fact, I've got this new pattern of consuming content where I star all of the items in Google Reader that I want to go back and consume on paper at a later point (as a general rule they have to be about 500 words or more and particularly thought-provoking for me though in my "skim to star" mode, I often get the latter wrong). Then, when I'm ready, I'll print off a whole bunch of posts and go through them as I might have a newspaper 10 years ago or the latest issue of the Economist on a Saturday morning. Right now this mode of consumption has two big limitations:
There is no backchannel for communicating feedback on the really good stuff back into my electronic life (I can't add it to del.icio.us for example or email it to a friend)
but outside of these two limitations, I find that I'm doing more careful consideration of the arguments presented in the pieces when I consume them on paper. This whole Facebook/F8 meme for instance, has been something that I've read about mostly on paper where I am more likely to focus on the longer more thoughtful pieces and less likely (frankly because I can't) to jump around from soundbite blog post to soundbite blog post.
Along a somewhat related vein: in anticipation of the most hyped product launch ever, I've been reading "Designing for Small Screens," a book about user interface/interaction design for display-constrained devices. The book is well worth reading if you are interested in this vein of different mediums/different modes of relating to content. It starts from the premise that the visual interface is the most high bandwidth way to consume content and then walks through a number of interface patterns that are useful in maximizing limited real estate. In light of the paper versus screen modalities, you can not help but wonder what mode of relating to content might emerge as these smartphone-like interfaces really do gain mass market traction over the next couple of years.
Posted by Antonio
1 year, 1 month ago (June 18, 2007)
I remember Seth Godin from the first Internet boom as the guy who wrote the smallish marketing pieces about colored cows that I never really paid attention to. But then I had his good friend, Lisa Gansky, on my board at Tabblo and heard all about how Seth was really one of the few Internet marketing folks who "gets it."
Which was why reading his recent blog post about John Heilemann's piece on Steve Jobs and the iPhone was such a pleasure. Better yet, in the blog post Seth points to this old Fast Company piece of his where he writes about Walt Disney as the ultimate three-time rifter:
What's a rift? It's a big tear in the fabric of the rules that we live by. It's a fundamental change in the game, one that creates a bunch of new losers -- and a handful of new winners.
I love this notion— that some people are just hard-wired to spot the real rifts— because it explains a lot about the best entrepreneurs that I know. Some people call it "pattern recognition," but for me it has always seemed more like good taste around what really constitutes new opportunities and not just hype and line noise.
As Seth tells it, most rifters wander into their first rift almost by accident and then go on to think that they are gifted enough to spot rifts even when they aren't there (I guess this would be the business equivalent of second system syndrome), but it is those rare few that can keep on spotting the rifts again and again despite position, age, or stature.
Posted by Antonio
1 year, 1 month ago (June 24, 2007)
A month ago I switched to Google Reader from NetNewsWire because I was tired of the cross computer synchronization problem. Reader is also one of those nice applications that doesn't just ape its desktop predecessors— perhaps because there isn't a real strong legacy of desktop RSS readers— but that instead tries to make the whole experience of consuming feeds more web native by weaving in social features around the sharing of posted items. I'll have a lot more to say about mixing together feed management and social features over the next few days (as I am on vacation and will have finally have a chance to write about friends, feed management, and the secret of Facebook's success) so for now I will just say that Reader is a good application that anyone interested in going beyond very basic feed consumption should definitely check out.
This post is not about Reader though, but about Google Gears, a new browser plug-in that was released recently around Google's first Developer Day to make it possible to take web applications offline. Basically Gears combines 3 things together in a local runtime: an RDBMS (SQLite), a local web server, and a multi-threading primitive in the form of a worker pool. The plug-in (which currently runs on IE and Firefox and sort of supports WebKit/Safari) provides these three services and leaves the rest to you, the developer.
I have to admit that when I first read about Gears, I was a bit disappointed that Google had chosen to bite off so little in order to solve what is probably the Achilles heel of all web applications, offline use. If solving this problem had fallen to me (which it did in a small way 7 years ago when we were building a bandwidth-enhancing service for the Memora Personal Server), I would also have used a local webserver but made it capable of running server code (in the form of Java/PHP/Python/Ruby plug-ins) and capable of transparently rewriting URLs so that the enduser was unaware of which server was actually handling the request while the developer had less work to do in building out hybrid applications. We pulled this trick off at Memora to serve big video files from a local webserver instead of across the network when the assets existed in both places with the best part being that outside of a bit of middleware which the developer had to be conscious of in certain cases, the rest of was completely transparent.
That said yesterday I got my first chance to test Gears via a synced version of Reader that went "offline" with 755 of my most recent posts for a long airplane ride. Holy smokes, that thing is fast! Though it could have quite a bit to do with the way that Reader itself is architected, the best part of using Reader offline is that you don't have to wait for the server loads as a new bunch of posts is loaded in. It just screams. By comparison, I've got a single user instance of the entire Tabblo website running on my machine for development purposes (MySQL, Apache C modules, Python, Django) and it is much much slower despite having no real server load, so much so, that I'm beginning to think that there may be quite a bit of connection setup/teardown overhead between a web browser and even just a local server, and that the Google guys may have been conscious of this in the design of Gears.
Gears is not without some enduser annoyances that are imposed on it by the browser, the biggest one being that you have to remember to leave the browser window open before you go offline or you're stuck despite the plug-in having all of the data anyway. There is also of course the nasty problem of distribution; despite the fact that it's provided as a smallish .xpi or ActiveX control, it would be a whole lot better if it could come pre-bundled with the browser, or at the very least with the Google Toolbar. There is zero doubt in my mind that Gmail is going to get a Gears life any day now, something which will likely remove the last obstacle for my going to it full time for all of my email needs, so perhaps there are plans to address both of these issues.
In the meanwhile, I can't wait to try using Gears for a project as it's clear that the Gears folks have done a very nice job of doing a UNIX (little small tools that can provide big leverage) to a key piece the "webapps all the time and everywhere" challenge we're all dealing with these days.
Ever since I read Neil Stephenson's The Diamond Age, I've been waiting for someone to crack the code on moving books into the digital age. Despite the false starts of local companies like E-Ink, it's clear that paper (or some form of it) is here to stay. The form factor is just perfect for some use cases and, more importantly, the emotional attachment that all of us nerds have to the page-flipping format is way too strong to be broken by even the most over-hyped and anticipated of gadgets.
Moreover, the history of print design is rich with a lot of smart people thinking very hard about the best way to present information to be consumed on the printed page, and it's not likely to be eclipsed by the world of web designers anytime soon. Even the guys at 37 signals are all about how online design may now be making a U-turn and heading back to the realm of print.
For me, the biggest challenge in understanding how the book will transform into a digital object is in creating a proper back channel so that users can continue interacting with content in the way that the web has enabled. Take Digg or Reddit as cannonical examples of how we are used to injecting feedback into the content consumption process: if I read something really awesome, I want people to know about it. I might even want to leave a comment if I think that I can help extend the argument.
Until the book form factor solves the back channel problem, we're stuck with their being static artifacts of the information explosion going on around us. Which is again, why I was sad to miss Manolis's presentation. Of all the work I've seen around e-books, his comes the closest to expressing what could be a very viable product roadmap for the venerable book.
The only thing I'd ever known about Marc Andreessen was that he was the Netscape guy that brought a ton of shit down on himself by claiming that Microsoft was soon to be relegated to a bunch of buggy device drivers back in the day when Netscape itself was having trouble shipping good browsers. These days though he's writing a stellar blog about all sorts of interesting things and his latest post on startups and success is probably the best single piece I have ever read on the question of success factors specifically, team, product, and market.
His argument boils down to the fact that a killer market trumps all else. My favorite line is that an exploding market will pull the product out of the startup even if both the team and the product vision are less than good. This is readily apparent to anyone who has been on the east coast for a while and had the chance to meet some of the PC and Internet 1.0 entrepreneurs, arguably the two largest exploding horizontal markets.
Well worth reading, especially because it is a bit more nuanced than similar posts that get to the same message by usually arguing for "making something that people want."
The first major reviewers of the iPhone have cast their votes and on the whole, the response is overwhelmingly positive. Walt and Katie had a great time with the phone calling it a "whole new experience and a pleasure to use," and thus far David Pouge (New York Times), Steven Levy (Newsweek), and Ed Baig (USA Today) all agree. To be sure there are quibbles about the network and about the number of steps required for certain operations, but that a device this over-hyped has its early reviewers so ga-ga is an incredible achievement for the folks from Cupertino.
I'm really happy to see it, not because I'm thrilled about the iPhone (which as a two-time owner of smartphones, I am) but because I love to see folks rewarding Apple for taking such a big product risk. Founded when I was just three years old, I feel as though I have grown up with Apple; from the Apple ]['s evolution as the first mainstream PC, to the Macintosh's shift to the Intel processor, I can remember all of the major product announcements the way one remembers big family events. With every major product Apple (under Steve) has been bold in crafting a unique product that often times defines a market by disrupting what heretofore has been nothing more than engineers grasping in the dark.
Which is not to say that Apple has not been without its big misses, most of which can be attributed to Steve Jobs: the Lisa, the lack of HD in the original Mac, the absence of licensing for the Mac OS, the Apple Cube, and even the company's deaf ear to the early days of the Web are a few examples of times when Apple was just plain wrong and often stuck with their mistake far longer than what we might refer to as "nimbler" companies.
Overall though, a small price to pay for the company that has given us a great UNIX on our desktops, Wi-Fi in every broadband connected home, the iPod ecosystem (remember when pundits called it an acronym for "Idiots Price Our Devices"), and hopefully the new computing platform that the iPhone is about to usher in.
Few companies are that bold about the things that they put out into the world— and none have survived so many misses at bat as Apple has while at the same time going on to have so many hits. It is precisely because of this that I think the "Think Different" ads best captured what I think are at the core of both Apple and Steve's soul. And as we get ready to line up for the iPhone, I'm just glad that they have kept on swinging hard.
Update: David reminded of this great piece a few weeks ago in the Economist about Apple and innovation that just needs to be added to this post.
Back in 1999 there was this dot-com company called Third Voice that was the first Internet 1.0 attempt at providing "co-presence" as you surfed on Internet pages. The idea was to use the content of the web as a substrate for building real-time communities where people could interact by "chatting" on top of pages. Needless to say, the concept did not catch on in any meaningful way, and to this day, many Firefox extensions and widgets later (even Flickr 0.1 was just a rehash for photos of this co-presence idea), we're still not there. I'm almost ready to call it and assert that the notion of a layer of people interacting in real time on top of the more slowly moving content of the web is just not likely to be a mode of interaction that will ever catch on. IM is for realtime, Second Life is for realtime— but the venerable and scaleable web— even with its rapidly accelerating update cycle (fueled by wikis, blogs, and microblogs) is just fine without it.
Except of course that every once in a while I experience a content consumption craving that leaves me thinking about the promise of the Third Voices of the world. For instance, today, on the most momentous product launch of 2007, I am sitting thousands of miles away on vacation in Barbados, feeling very far from it all. I'm not sure that I'd be standing in one of the lengthy queues encircling each and every Apple store in the US (though I might be), but I would definitely love to get a better feel than what I can get from the blog coverage, Twitter updates, and occasional big-time journalist dispatch.
A few brave folks are trying video steaming but it isn't working well (because of the lame reason that the multiplexers don't seem to be able to take the load), and the Engadgets of the world are doing a nice job of the constant update to keep feeding us photos.
But overall, cobbling these bits and pieces together is a pretty dissatisfying experience. What I want sort of like Techmeme (who has the right starting approach to this) but faster moving, more tunable to my network of people, and way richer in terms of the multimedia being shared. Discoverable through Google— in fact maybe even promotable through Google, and capable of taking user comments, photos, and videos all one one page and with adequate peak-load capacity planning.
It's not clear to me why this hasn't emerged yet as a sort of web platform for events. It could be used extensively at conferences and then brought out for product launches, political events, even disasters. What is more, it is clear that various folks are bumping around its edges: from the Facebook news feed to the Twitter update, all of the leading-edge social platform experiments are feeling various parts of the elephant in the dark. And eventually I suppose, something might emerge from all of this experimentation that will finally deliver the live web we need to cover these types of things.
In the meanwhile, I'll be clicking refresh on apple.com...
So I've been on vacation in Barbados this last week where I used to come when I was a kid. One of my favorite things about this island are its people— incredibly friendly and outgoing, they are just a sheer delight (and given Barbados's public beach policy, this is not constrained to just the fancy resort employees). In fact, on the beach most days you are likely to experience Bajan (Barbadian) entrepreneurship at its best with everything from jetski rentals to foot massages being offered by a gaggle of passing merchants*. Everything is negotiable, and surprisingly, the process of being sold is almost as good as the product or service being offered.
My absolute favorite Bajan entrepreneur this year has got to be Stephen, the man who made palm leaf hats for the three of us. The hats are made from a single palm tree branch, curled and cut in just the right way and then woven together from the brim on to the top of the hat. Stephen has been making these hats for 29 years which you can tell from just looking at the way he works with nothing other than an armfull of palm, a sharp knife, and one short 1-inch wire to hold the brim together. From start to finish, he can make a hat in about eight minutes.
As we watched him work (see the tabblo for more), I got to thinking about how nice it is to have mastered a craft in the world, to have a skill making stuff, where you can really get into a groove. This is especially nice when you can make your living from it. In fact, this was the third time in two weeks that I thought of this: the first being 10 days ago when I was in the chair of the most skilled root-canalist I have ever seen. This particular endodontist had all his own tools and the way of a master craftsman about him as he patiently worked his way through all of the canals in the root of my tooth. After he was done I felt compelled to tell him: "I wish that someday I could be as good at anything as you are at doing root canals" to which he replied dryly and without apparent pride, "I wish I was as good at anything else as I am at doing root canals."
The third time that I thought about craftsmen recently was in reading the fabulous The Amazing Adventures of Kavalier and Clay. In it Joe Kavalier, a brilliant magician/escapist/artist refugee from Prague, comes to the US during the comic book boom in the late 1930s and succeeds as a comic book artist. The novel is a great riff on the American Dream, and all of the best passages in it are all about Joe and his single-mindedness when it comes to practicing his craft.
For me programming a computer comes the closest to practicing a craft. I like to write, and I love to build companies, but at the end of the day, if you dropped me into a strange city with no history, credentials, or reputation to ride on (one of my favorite job interview questions), this is the main thing I know I could do to earn a living. More importantly, I love to be around great software engineers, to read beautiful code, and to talk shop in all of its annoying and seemingly self-centered ways: chatting about tools, approaches, and war stories for hours on end. As I've gotten older, I've dabbled in more and more things but for most of them, it's just that— experimentation until the novelty wears off. With software, time stops and things just seem to fall into a very nice groove for me.
I'm sure that I'll never be as good at writing software or designing big systems as Stephen is at making his hats. But at least I'm glad to know that this is what I love to do, and that in some weird joke-on-the-universe way, I get paid for it.
Incidentally, it feels to me like it is the best time to be alive for all of those people still looking for crafts to love. Kavalier and Clay's comic boom with its creation of a new medium pales in comparison to what we're seeing as the tools of production become ubiquitous and the net binds us ever close together.
*Between writing and posting this, I even had someone come by and ask if I wanted to have my hair braided— my chest hair that is.
Sorry to everyone that comes here for the "big thoughts" about the following shop talk. Feel free to skip— for those interested in web frameworks, read on...
Being the conservative guy that I am, I was willing to bet millions of dollars of investors' money on an "unproven" open-source web framework, but not my own puny personal blog. But a few days of vacation have finally cleared the head and have caused me to abandon the whole Ruby-on-Rails powered fad (more specifically the twice-abandoned Typo project) for a blog powered by Django.
I can't say enough good things about Django. Professionally, it was one of the best technical decisions that I got to make early on at Tabblo. I had always assumed that we would use it for the alpha, maybe the beta, and then dump it for a custom system that could perform. Instead we rode it from the first alpha until the day we got acquired by HP. In the meanwhile we ran 5 million photos through it, a million users, and over 150,000 content publishers. We had some ugly moments (the shift from mod_python to lighttpd/FastCgi and the move to magic removal stick out in my mind), but overall it performed incredibly well. A lot of props go to Adrian and gang for putting together such a flexible framework. They used it as a lightweight CMS, but we were able to take it and build out a full UG content publishing platform (with a lot of help along the way from memcached and solid engineering on the part of the Tabblo folk).
In fact, the HP folks doing diligence on us started by asking why we had not chosen Ruby on Rails and came out the other end incredibly impressed with our Django choice. We survived our security audit (coming out much more unscathed than a lot of other HP acquisitions) thanks mainly to Django, and have continued to develop on top of it even as we get ready to truly become the "print engine for the web."
For sure there are things, I'd love to see improved, namely: mutliple DB support (sharding), a better testing framework, and general speed improvements. But hey, now that we're HP maybe we can find a way to help out with those things!
In the meantime, theonda.org (or an.ton.io) is now officially "Django Powered." I haven't had this much fun programming since at least 6 months before we got acquired. In fact, I've decided that Django— when it comes to web stuff— is perfect for "vacation programming." What do I mean by this? It's the type of programming where you get to focus on the fun stuff and ignore all the crap you find yourself having to do again and again on every web project. A few examples:
* Generic views: if I had 5 dollars for every time I've parsed URL fragments to present a list of objects by date/type/whatever, I'd be long retired.
* RSS/ATOM framework: this one is a bit messy (tightly coupled) but man is it great not to have to deal with XML encoding/decoding. Wowee, the funnest 30 minutes of the project.
* All sorts of nice html-ifying functions: want to wrap your paragraphs in p tags? Format columns of text? All built in with Django.
Finally, Django is just well architected. I use a venerable Mac program called MarsEdit to write all of my blog posts. Unfortunately MarsEdit seems to support mostly crappy old XMLR-RPC-based protocols for publishing (Blogger1, metaweblog). Fortunately for me, between Python's support for xml-rpc and Django's URL dispatch mechanism, supporting metaweblog was a "nap time" exercise (it would have been less time if there was clean and current metaweblog docs on the web). Fun stuff.
Now that I've moved to Django, I'm really excited to start playing again. The thing about Typo was that it was hard to make any changes to it without having some of that RoR "magic" explode in my face. With the Django there is no magic— just clean, explicit, and simple design.
For our part, after having spent six months building an initial member base that was attracted to the Tabblo service (either because of the editor, the online/offline combination, or the quality of the products we make), we're now starting to experiment in earnest with products that will begin to motivate members to buy at greater frequency than most other output vendors. 
The third time that I thought about craftsmen recently was in reading the fabulous 