Electronic Consumer Hardware

If you have a stereo or a set of speakers in your house, you’re dealing with technology that hasn’t really changed that much in almost a century. In the same way, TV sets have seen little evolution in the way of being smart over the last 50 years: when you think about it, the biggest steps in TV have moved from black & while to color (in the 70s), cathode ray tube to plasma and LCD (late 90s), and analog to digital (early 2000s). In each case, the focus has been on the picture and not so much on the logic to receive and display information.

While new models of television of radio are introduced on a yearly basis, the fundamentals behind them are essentially the same and the features that are introduced are, for the most part, marginal improvements.

However, over the last 4–5 years, a quiet revolution has been underway in the hardware business with the addition of two apparently innocuous components making their way through into a lot of the more recent version of new devices: USB drives and internet connectivity (either wired or wireless). This has made it possible for the core software components of those devices to be upgraded, either by downloading the upgrade on a USB drive and plugging it into the consumer device or automatically in the background via the internet, opening a whole new world of possibilities.

The remote experience

A radio station in the 1920s used a knob to find the right signal and today it’s a button. The clarity improved by moving from AM to FM and HD radio but the user experience is basically the same. For TV, the experience has moved from having to turn a knob on the TV screen to turning a knob on a device that was connected via a cable to the TV to pressing a button on a wireless remote control. The interface has not changed much, except for adding more buttons to allow access to more complexity.

Meanwhile, an increasing amount of touchscreen devices are starting to populate home, whether they are mobile phones or tablets and there is now a slew of software that allows for operating some of the more complex technologies out there. Using simple infrared adaptors, those devices are now allowed to share their own smarts with TV sets and radios, pushing the creation of simpler and more adaptable interfaces for remote controls to increasingly becoming software based solutions. As software evolves, it can be upgraded easily and include new functionality without getting rid of the devices that consume it.

Hardware is a platform

In this instance, we are seeing software trumping hardware. Short of the dependence on certain physical components being natively included in hardware devices (for example, the ability to support certain forms of communications like infrared or short range radio), the hardware does not really need to get upgraded unless its primary function (eg. showing a picture or playing some sounds) is itself in need of an upgrade. And one could see a time when the receiving hardware would receive software upgrades that allow for this interoperability to be a smoother experience because hardware is a platform.

A group of technologies around interoperability have made it easier for this phenomenon to happen. While few people care about such standards at HTML5, HTTP, TCP/IP, WiFi, and DLNA, they can serve as the building block of the future.

Imagine a television, for example, where every channel would be served through an internal web server that rendered everything on the screen via a web page with an HTML5 embedded video player. All of a sudden, the TV screen would become a giant web browser in full screen mode, allowing to not only access any content on cable or broadcast TV but also any content available on the internet.

Add a layer that would allow for throwing HTML5 applets on top of that screen and you would have a standard compliant approach to developing things for television. Throw a webGL interpreter in there and you have something that is drastically more advanced than what any TV in 2011 can currently do. Now make this layer addressable as a web service standard and you could not only see other companies incorporate it but also see an explosion of support from the development community.

Standardization is they key to TV’s future

Today, developing for television (or send streaming audio to a stereo) is an often frustrating experience, forcing developers to encode content so that it meets the requirement that each TV or set-top box manufacturer has set forward. This often complicated landscape has left most developers avoiding it because the return on investing in a single platform just isn’t there. A way to battle this is through standardization of the underlying interfaces. Since the 1990s many people, myself included, have pushed for a standard language to merge television and the web.

These efforts need the support of TV manufacturers and the understanding from said manufacturers that open standards will not only lift their industry but potentially fuel another area of growth for their offerings. As more and more applications get developed on top of a standard compliant deck, there will be increasing demands on the part of the developers to give access to other parts of the hardware, which could be completed via paid software updates. Hardware manufacturers would then find themselves in a world where they could make money on the initial hardware they sell but also add extra revenue by turning on extra functions through software sales.

Lack of standardization creates a winner-takes-all market

Apple has an early lead in the audio market, with Airplay (and its support from many hardware vendors). For example, in our house, we have equipped several rooms with Airport Express adaptors connected to powered speakers or stereos. Apple had initially locked Airplay to Apple only devices, but an ingenious company called Rogue Amoeba put out a piece of software called Airfoil that allows us to play from any source we can find on a computer. It would be nice to get such functionality on our mobile phones and tablets but, because Apple is locking up its system, it is unlikely that such thing could happen unless allowed by Apple.

An open standard that reproduces such functionality could not only hamper Apple’s ability to lock people into its ecosystem but could also help proliferate the rise of inexpensive devices that can be addressed from the internet.

In this case, as in the case of TV, we could see a single closed proprietary standard emerge and create an early lock-in for the owner of that standard. However, if there is one lesson the internet has taught everyone in the computing industry, it is that open always trumps proprietary in the long run. Early platforms may get early wins, but eventually, the open standard disturbs their marketplace and destroys the proprietary aspect.

Some people may consider the the receiving devices for new content as dumb devices but I would argue that they would fall in the category of smart devices: after all, isn’t the smartest person in a room generally the one that first listens and only speaks their mind after receiving appropriate input. Maybe we should create a new category, calling those wise devices instead of smart ones.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Interop: the future of hardware. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

Core Components

One of my favorite quotes about the future comes from William Gibson: “The future is here, it’s just not evenly distributed yet.” Based on this, it is relatively easy to make future predictions by looking at some of the core things that are happening today.

Some of the trends I’m starting to look at in terms of defining how the next generation will work include

• The rise of always-on high-speed internet connections
• The IPzation of everything
• The drop in the price of real-world sensors
• The rise of participatory applications

Looking at each of the components individually, it is hard to form a picture of where the world is going. However, looking at them together and how they interact, a picture starts to form.

A seismic generational shift

We are now starting to reach the tipping point in what will be a major generational shift. Over the next few years, people who were born in the Internet era (ie. after the Internet became commonplace) will start entering the workforce. In a way, my generation (I’m 35) was too old to be part of this shift. Younger people, however, are used to a word where IM, email, SMS, and always-on communication is an inalienable right.

I believe this will force companies to start dropping restrictions on such uses of the network. Already, we are seeing more corporations adopt Instant Messaging (IM) as an approved form of communication (primarily for corporations with widely distributed workforce) to tag along with email, which has now become a de-facto in business. IM has changed some of the factors inherent to communication in the past: traditionally, communication between members of a distributed workforce was slow and relatively inefficient (yes, even Email had some inefficiencies in terms of productivity.) As things like IM, video IM, and shared applications over IM, take hold, workers can now have meetings with faraway places in the same fashion as they used to with their co-workers in a similar physical setting.

As those physical boundaries start to drop, what’s happening is also a rethinking of how teams are organized (with a growing emphasis on distributed virtual teams) and where people have to be to work (it doesn’t matter as much any more). Some smaller start-ups even forego the idea of having much of a physical location (for example, WeblogInc, which my friend Jason Calacanis sold to AOL last year, had no physical office prior to the sale).

RSS, while still nascent, is also starting to take hold in companies. As it becomes more and more of the way data is exchanged, RSS will start replacing other methods of giving updates. With the pervasiness of RSS as a delivery envelope (and, since it supports enclosures, RSS can deliver any type of data), the way sites are authored in the future will be not through a centralized approach (as is the current model, with a web server being connected in a one to one relationship with a data storage piece like a database) but through a decentralized model where specialized sites will offer narrowly focused types of services that will then be aggregated on a page. The early indications of this shift can be seen in the “mashups” of sites like Google Map and craigslist, for example and I believe the next generation of people entering the workforce will demand such freedom to recombine systems in the future. Along with the recombination of applications, they will also demand that such flexibility exist in the way they work and we will see the rise of a more modular and flexible workforce, with virtual teams replacing the more rigid structures that currently exist in corporations.

This is the first article in a 6 part series. You can read the following parts here:

• Part 1: Intro
• Part 2: Always on
• Part 3: IPzation
• Part 4: Sensors
• Part 5: Participatory Applications
• Part 6: Conclusion

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Future Tense — Intro. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

Yes, absolutely free. How much for that domain name in the window? Nada, zilch, nothing.

But for a lot of people, it seemed too good to be true. And for a lot of people, it just didn’t happen. At about 9:01pm, the site started to experience VERY sluggish response times. First, they claimed the NSI registry stopped answering.… Then their nameserver crashed… Then the web server crashed. Then, hopeful registrants around the Internet started talking.. and talking… and talking…

It’s Scalability, Stupid!

In the end, the issue came down to one simple overlooked issue: Yes, you guessed it.

It’s give me an S…
give me a C…
give me an A…
give me an L…
give me another A…
tired yet?
give me a B…
give me an I…
and another L…
and another I…
and a T…
and finally a Y!

What does that spell.… SCALABILITY!

While they were expecting under 500,000 unique visitors, they logged over 2 million attempts to enter the site and in the end under 10000 names were registered.

As a result, RegisterFree went from being the new kid going against Network Solutions to joining the ranks of Victoria’s Secrets and ABC in the annals of sites that failed to scale to demand.

By 12:10am tonight, RegisterFree issued the following statement: Ladies and Gentlemen:
This was truly awesome. We never expected ANYTHING like the AMAZING response we
received to our RegisterFREE Free Hour Promotion. At certain times over the course of the evening, the NSI registry (the universal database which provides domain name availability checks) was unable to process requests for those domain name availability checks. As a result, traffic was significantly slowed at times, and some people were unable to register their domain name. We hope to receive better support from NSI Registry during our next promotion, which will be coming up very shortly. We can safely say that hundreds of thousands of people came to RegisterFREE.com tonight, and thousands were able to register their domain name for free. At this point, we ask all those who were unable to register a domain name to send an email to cs@registerfree.com and let us know, so we can deal with each query personally. Congratulations to all of those customers who did get through and were able to submit their registration request. RegisterFREE.com STRONGLY believes that Domain Name Registration should be 100% free, and we will have this promotion again very soon. Again, thank you for your continued patience and support as we try to make all domain names 100% free. The RegisterFREE team.

Now I can hear many of you scoff at the fact that they couldn’t scale.

Yes, it’s true, they mis-planned but how many of us plan for such traffic surges? And how do you plan for something like that. Do you overbuild and hope that people will come? What is the factor by which you have to multiply your expectations and get an appropriate number. Last summer, drugstore.com went down on its first day, deluged by over 1 million requests. Should we use 1 million as the magic number?

Many large web sites have been designed with little attention paid to scalability and every time one of them fails, we all look bad. Why? Simply because the Internet is considered as unreliable when that happens and THAT scares away potential customers.

Don’t believe me? Try asking my 82-years-old neighbor! When Yahoo! and Ebay were taken down, she was so scared by the negative press that she told me she wouldn’t shop online because hackers would steal her credit card and overcharge it. After explaining to her that she was only liable for $50 maximum and that incidents like this were rare on the Internet, I think I managed to rescue one more customer on the Internet.

But enough negative publicity along the lines of what I expect to see in tomorrow morning’s newspaper (especially considering the great coverage RegisterFree managed to get BEFORE the event) is undermining consumer confidence in the Internet and that is what I worry about.

But Wait! It Gets Better!

More troubling was the fact that they were blaming it onNSI. While I don’t particularly like NSI, I have to say that their whois was working tonight. I spent all evening checking domain names in the NSI whois while I was trying to get into registerfree.com. It worked fine for me, actually working faster than it usually does. I even ended up registering domain names from there. RegisterFree.com pointed out that someone at NSI decided to take the “registry processing systems offline for an indeterminate amount of time” but other registrars were fine. After the promotion ended, RegisterFree was still the subject of some discussion regarding ethical issues. It may be just a technicality but other registrars like Bulk Register and Enom (just to take a couple of names from the ICANN accredited list) are cheaper.

So can registerfree be trusted? I just provide the links. You make the call.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Trouble in RegistrarLand. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

Unfortunately for them, it wasn’t quickly enough and tens of thousands of people got their hands on the software and started redistributing it. While it will most certainly be used for stealing copyrighted material, this category of software interested me at first because of the indexing technology that was built in them. One of the great thing they do is index the files on your system and make them available to everybody else who’s connected to the same server as you are (in Napster’s case) or to the network in general. This could be significant if you were to build a search engine.

Imagine search engine software that would be installed on every web server out there. Not only would it index the pages for the server administrator, but it could also report back to a mainstream search engine. Many studies have now come out about most of the large search engines (Inktomi, alltheweb, Google, etc…) only managing to index a fraction of the web. With a service a la Gnutella, you could have every web site call back the search engine directory to post the changes they had. This could happen throughout the day and might work better than the spider technology currently being used by most search engines.

The other potential use for tools like Gnutella is as a mass corporate cooperation tools. Right now, when you fire off Gnutella, it connects to Gnutellanet, or pretty much anyone who’s using Gnutella at the time. As I just checked on it, there are 700 people connected. There is no central source or server, which means that a tool like Gnutella could be used to share files without having to worry about a central server go down.

For years Sun has been claiming that the network is the computer. With a tool like GNUtella my hard drive can be become a portion of a larger hard drive. I could have a marketing hard drive, a finance hard drive, a HR hard drive of which only a portion would be sitting on my computer. Compare this to current corporate client-server systems where you have to deliberately save a file to the corporate server as well as to your hard drive if you are traveling… forget to save it to one or the other, and you’ll be stuck without your work or somebody else’s later revision. With a GNUtella like system, YOU would continually have the most updated versions of the files YOU author, without having to remember to separately save them.

At the same time, however, it seems to be lacking in a couple of critical areas: first of all, Gnutella could do with some sort of an authentication mechanism. That way, I would be able to create profiles and give access to certain files to certain people. For example, I would be able to mark a spreadsheet as accessible to the finance department while I would have a powerpoint presentation accessible to the marketing folks. The potentials are endless.

While Gnutella is considered a major threat to the music and movie industry, it is those corporate uses that interest me. I believe that, in the long run, those tools will make their way in corporate America and not just because someone wants to download the latest version of Santana’s new album or Julia Roberts’ new movie.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title AOL’s dark little secret. You can follow him on twitter here or receive his weekly newsletter by subscribing here.