45 posts tagged “manifesto”

It's been a rough week. I was able to ride on Sunday, do some rack-and-riding, and the Ride of Silence. But, because of other commitments, tonight was my first opportunity to do a "fast land mammal" ride that i wam want to do on the Habby.

Unfortunately, work has been hard lately. Too many projects with a tight deadline. To many parts that were out of my group's control. Too many bad decisions inherited. We're having a meeting tomorrow to see if we have any hope of salvaging it.

I got home over an hour later than I intended. At that point, I was pretty much burned out. There was a large part of my brain just wanted to plop down on the couch with a remote, a laptop, and a bowl of popcorn. I simply wanted to shut down for a bit.

But, with a three-day weekend ahead, I wanted to get some miles in so I'd be ready for some good riding. I changed clothes, pulled out the Habby, and set off.

By the first climb, I was no longer thinking about work.

By the first decent, I was focusing on the ride. Was I in the right gear? What was my heart rate? Was there a car behind me?

When I got home, I felt silly for thinking I wouldn't go ride. I was feeling good about the ride and how I did. I had shed the problems of the day, and relaxed.

In the spirit of Chuck Norris Facts, I give you

1. Virtualized servers can complete an infinite loop in about an hour.
2. Virtualized servers made the Y2K bug drys for it's mommy, the Y2C bug (that was discussed a lot in 199 AD)
3. If Werner Von Braun had a virtualied server, the Saturn V would have sent men to Saturn
4. Skynet externinates humans; virtualized servers exterminated Skynet.
5. The only heat from a virtualized server is the warm glow of love.
6. If WOPR had been a virtualized server, it would have won tic-tac-toe, and launched the missles.
7. There is much debate about if a virtualized server can create such strong encryption that another virtualized server can't crack it.
8. Within five years, Pixar will be put out of business by a virtualized server that not only renders the animation, but also writes the story. Given that Pixar only has one or two plots, that is only a fraction of the virtualized server's capabilities.
9. Virtualized servers have their own iPhones
10. Virtualized servers don't create LOLCats; they create LOLHumans.
11. Once, when computing the values of Pi, a virtualized server found the answer to all of life's questions, as well as what the hell is going on on "Lost."
12. There is no such thing as "random numbers" when a virtualized server is around. It sees all the pseudo-random patterns.
13. A virtualized server could digitize a human and force it to ride motorbike simulations. However, the virtualized server would then have bugs.
14. A virtulaized server could run a complicated simulation of Los Angelos, and use this virtual reality to enslave humans. But then its processors would only run at 3%.
15. A team of scientists and linguists are coming up with a new term to describe how many FLOPS a virtualzied server runs at.
16. There is no Illuminati; only a virtualized server controlling the presidency (through Obama's BlackBerry).
17. With a virtualized server, you don't get it fast, cheap or right (pick any two). You get it fast, cheap, and AWESOME!
18. A virtualized server can beat a grand master at a game of chess while headlining in a Broadway performance of "Chess."
19. "Duke Nukem Forever" was finished by a virtualized server. It's incredible.
20. A virtualzied server pounds the Fonz to get him working.

One of the bigger buzzwords right now is "Cloud Computing," the notion is that you would have your data out on the Internet ("the cloud"), and leverage browser-based tools. A good example of this notion is Google Documents. Your documents are hosted on Google's servers, and you can edit them with browser-based spreadsheets and word processors. The advocates of it suggest that one day, we'd all "live in the cloud," carrying around very low-powered devices, and all of the heavy lifting done on the back end.

While I appreciate the concept, and they even have a niche in my life, I'm not entirely convinced that this will ever replace the notion of PCs. The classic objection is "what happens if you get on an airplane." That is to say, what happens if you are unable to access the cloud. Google addressed this with Google Gears, to allow some bits of the web data, along with your documents, to sync back and forth. Recently, they enabled GMail (perhaps the best web-based client for mail) to use Google Gears. There have been folks who feel this is the ultimate Outlook killer. For a fair chunk of functionality, I could see this, though that may be damning GMail with faint praise.

The overall notion comes up from time to time--that a thin device will leverage applications and computing resources on a central server. Historically, this has been killed at least three times in the last fifteen years. It used to be that, if you wanted to use a computer, you leveraged a "dumb terminal," and connect back to a shared system. In college, for instance, we all shared a VAX. Networked PCs replaced that in the early nineties (though it started in the eighties). Everyone would have their own system, with their own application. Data might be shared on a common server, but, by-and-large, the systems could function independently.

As we moved into twenty-first century, the notion of a "thin client" came into being. Simple systems based on PCs but with more limited resouces would use tools like Citrix to leverage a common pool of applications on a central server. The server would do the heavy lifting. HP even offers few systems.  With the possible exception of a help desk, I've never seen these used.

The final example of a thin client using centralized processing is a fairly recent one: the iPhone. Those of us with first generation devices will recall that Apple did not have an API or App Store. A special set of style sheets and techniques gave web sites an iPhone feel. Anyone who used a twitter client client before the API and after will tell you how much better the new ones are. This example most closely maps to the common description of cloud computing.

Why is this? Each platform--be it an OS type (Windows, OS X, Linux, etc.) or device (laptop/desktop or phone)--has its own strengths and weaknesses. Since switching to the Mac, I can definitely tell when an application was written for the Mac, as opposed to a relatively straight port of a Windows or UNIX application. The "cloud computing" notion, built around a browser, is a lowest common denominator. It can't play to a system's strengths or avoid its weaknesses.

With respect to GMail replacing Outlook, Neven Mrgan put it well, "Well, I’m convinced. I guess I’ll just switch to an email client that doesn’t allow me to drag a goddamn file into the message to attach it." It seems like a simple thing, but it could make a big difference to someone who does that dozens of time a day. Likewise, I love Flickr for sharing photos, but for archiving and basic manipulation, I much prefer to browse with iPhoto.

Another disadvantage of the applications being in the cloud is that we are limited to only what the provider of the application allows. Case in point: if I want to use PGP to encrypt my mail, it is not a function native to GMail. Sure, I can copy the note, encrypt it, then paste it into the GMail window. However, there are simple integration points for Mail.app--click a button, and my mail is encrypted on the way out.

There are several things that may shake out of this recent effort. First, I think we might find some hybrid: an easy way to store files on the cloud, that we edit with our own tools. This might be augmented by some web-based tools. For instance, I might create a document in Microsoft Word that I keep on a "Google Dive." If I need to make a quick edit on, say, a hotel's kiosk, I could do so with Google Docs. In a way, since they introduced IMAP, we can do this with GMail.

Second, I think we'll see a rise of things like Twitter clients: applications that use data from a web site, but the heavy lifting for the interface is done by a local application. Twitter's API I think is a great example of this. There are a number of applications, designed for a number of platforms, allusing the same data.

The alternate view of my thesis is that, with computing, it often takes several passes before something really takes hold. PCs themselves first entered homes in the early eighties, but it wasn't until the mid-nineties that it took hold. LIkewise, this "fourth coming" of cloud computing could be the one that sticks. However, I think that, until a cloud application is indistinguishable to one on my desktop, I don't think its time has come.

There is a paradox I'm sure anyone who has lived with a small household animal (SHA) has encountered. By SHA, I'm talking about ones who have run of the house (and even get to sleep in the bed), rather than ones in cages. Think in terms of cats and cat-sized dogs (I have experience with shih tzus, though I'm sure it's true of similarlly-sized breeds)--say this tops out around 15-20 pounds. Simply put: even though they are small, they are able to limit the human's use of the bed in a manner grossly disproportiate to their size.

(Yes, limit the human's use of the bed. The human, who owns the bed, feeds and cares for the SHA, and works all day while the SHA makes complete use of the bed.)

Small Animal Paradox

How do they do this? By curling up very close to the human. Sure, it is cute, and, during the winter, quite warm. The human may roll a bit closer to the edge, or bump some of the covers off.  The SHA adjusts, moving closer to the human, who may move a bit.

Before you know it, you have scenario one (see illustration above). Here, the human as gradually migrated to the very edge of the bed, with the SHA the boundry of the rest of the bed. Or, perhaps it's scenario two, where the SHA lies atop a pile of blankets and sheets the human pushed off. The human, when he or she needs some blankets, the SHA acts as a paperweight, preventing full coverage.

The human may try to nudge the animal. I find this to be ineffective at best. More likely, the human loses ground, either in the form of covers or space.

In the sober light of day, the solution seems obvious: move the SHA. This is easier said than done. If it's two AM, the human may be half-asleep. To move the SHA, the human likely has to sit up, which, in turn, means going from half-asleep to fully-awake. Denial sets in: "If I lay here, I'll go back to sleep more quickly than recovering from moving Fluffy." Or, there is the concern that by moving the SHA, the human may upset him or her, making them leave the bed entirely. The human would prefer the company (and warmth) rather than upset the SHA.

Thus we have the paradox: a SHA may take up only a square foot or two of the bed, they are able to control the whole bed. I propose we fund a team of engineers to devise a bed that would continuously migrate the human/SHA set to the middle of the bed, preventing these problems. Who's with me?

By now, I suspect that most folks, not just cycling fans, have heard that Lance Armstorng is coming out of retirement. Honestly, I have mixed emotions about it. He is an incredible athlete, and I'm sure his return will bring some excitement to the sport, and get more people excited about cycling in general. However, I keep thinking about another athlete who kept "unretiring." Ijust want to scream, "Don't be like Mike!"

A 37-year-old coming back and competing in what is arguably the most physically demanding sporting event in the world is impressive. Not only would he extend his number of wins to eight, he would also better Firmin Lambot, the oldest winner of the Tour de France (36 years old, in 1922).

However, I think it would be sad for him to merely eek by. The 2004 Tour de France was an incredible run for Armstrong. Not only did he win the Tour, he dominated. He won five stages, some in outright sprints. It was an exclamation point on being the first man to win more than five Tours. In 2005, he plotted as he usually did, and was mostly on autopilot. One stage win--the time trial on the second to last day--with the rest being riden fairly conservatively. While I understand the heights of 2004 might not be repeated, I would hope for better than 2005.

Some speculate that he wants to show that he rides clean. He was the most tested athlete in the world when he retired--why would even the transparency he describes silence his critics?

I do, however, think he has an opportunity to silence other critics. Towards the end of his career, he was fixated on the Tour de France. While he would go to other races, he used them more for training. He competed in neither of the other grand tours (the Giro d'Italia and the Vuelta a España). He was a master of combining his training, his team, and his equipment providers for one goal. His fixation on the Tour was critisized, and the modern fixation on one race that he started was bemoaned.

On the other hand, Eddy Merckx (a certain cat's namesake) won not only five Tour de Frances, but five Giros, a Vuelta, three world championships, and set an hour record. Plus, he won each of the spring classics (except one) at least once. In other words, Merckx was a strong, all-around rider.

If he insists on being like Mike, I suggest that Armstrong be like Eddy: go for a Giro or a Vuelta. Have a strong spring campaign. A truly amazing thing would be to forgo the Tour de France in order to do so.

I doubt he'll do it. One of the reasons behind his comeback is to bring attention back to his foundation and fight against cancer. In the United States, there really isn't awareness of any race but the Tour. However, I think it would be something for him to come back and do that which he has not done.

It's hard to imagine I've been on Twitter about a year and a half. It has become so integrated into my life. I've met friends, kept my family up-to-date on what I'm doing (in the are-you-home-from-the-ride sense), and have found it useful to timestamp my life ("When did I go to lunch? Oh yeah..."). I'm glad I can use it as a messaging tool and a status display. 

There are two web-centric applications I have started to use the last several months that I've come to the accept that I won't be able to exploit to their fullest potential. The first is Zamzar. Upload a file in one format, and it will convert it to an alternate format for you. It can do YouTube to mp4, PDF to Word, and several others. The conversion is done on one of their servers, and they give you a URL to get the final product.

The second is Evernote. This allows for notes to be stored and searched. It will even perform OCR on images, so text within images can be referenced. I took a picture of a whiteboard, and, to some degree, I can search for the text that I wrote. Really cool. The notes are stored on a server, and accessed through either dedicated clients, or a web page.

These are both incredibly useful. However, I have to execute a thought process with each use: how confidential is this? If the data is something sensitive to me personally, I have to take a moment to consider what happens if the company in question goes evil on me. We've all probably had this consideration when using web mail. Where it gets especially sticky, and where the utility of the sites has the most potential but presents the greatest challenges, is using it for work settings.

Some of what I deal with is confidential to my company. It could be deals we are working on, or intellectual property, or other functions. Uploading a document constitutes transmission to a third party, regardless of the terms of service. As often happens, security wins out, and I don't use it.

However, I think this represents a strong opportunity for these Web 2.0 companies to monetize their offering. If they create a version of the server that can be put on common, commodity IT infrastructure, a version for just my company could exist. I can get all the utility, but the added knowledge that it won't go outside of the scope of control of those who need it.

There would be a few challenges. First, the creators of this software would probably need to tweak it--at a minimum, to support integration into existing security infrastructures. It may, however, require replatforming it to things that are closer to corporate standards (this may mean Windows).

Second, they would have to show its value when other groupware is already installed. How does Evernote's utility exceed or compliment SharePoint (it's obvious to me).

I think Google is already on track--the can sell e-mail services and their office suite for your company. However, I think security and legal considerations will always keep applications that live "in the cloud" from ever reaching their full potential.

I sometimes hate Caller ID.

I appreciate it on my cell phone, when I can pick out my wife (especially on the bike, when a distinctive ring helps me decide if it is worth stopping). However, it gets abused too readily.

For starters, I don't appreciate folks who request that I call them (via a pager or e-mail), then capture my number. Just because I call you from a particular phone doesn't mean that is the preferred number. It just means that is the phone I happen to use that time. If I've given you a number separately, use that--you are much more likely to reach me.

More annoying is people who call back based solely on the caller ID. It's one thing if you just missed picking up the phone by a moment. I've done that (but I think I'm going to resist the urge). But if more than a few minutes have passed, and I didn't leave a message, don't call back. I sorted it out some other way.

Here is the worst scenario--the one that inspired this rant. I was calling a colleague, who had his cell phone right under his work phone in his e-mail signature. I somehow mix up parts of the two numbers, resulting in a wrong number. As soon as I got the voice mail greeting, I realize my mistake, and hang up.

Twenty minutes later, this person calls back, wanting to know who I am, why I was calling, etc. I explained it was a wrong number, and she didn't need to take up both our time.

Please, use caller ID for good, not evil.

In 1994, the second year of Wired magazine, Nicholas Negroponte wrote an article claiming that fax machines were a regressive step. Basically, the United States was tracking towards having more and more data in machine-readable formats--ASCII, e-mail, etc. This would be able to move information quickly, and would be machine readable. Archiving, searching, and retrieving this information would be quick. An infrastructure to make it business-friendly, such as digital signatures, would be developed and become accepted.

However, as the fax machine became ubiquitous, using older business methods, such as physical signatures on paper, were allowed to remain. By scanning an image and sending this data (slowly), paper could remain the medium of choice.

Granted, at the time of the article, computers were $3,000 investments. Relatively few people have heard of the internet--a common interconnect between businesses to allow for e-mail between anyone. In contrast, everyone knew how a phone line worked, and a fax machine would cost less than a quarter of a full computer. This was less a barrier for large corporations, but rather for common, everyday small businesses.

Fourteen years later, almost to the month, paper documents still rule, and I blame the fax machine. Granted, e-mail has become an important tool in time-stamping communications. With iPhones and Blackberries, it is more ubiquitous than fax machines. However, when it comes to officially authorizing and authenticating documents, pulling out a pen, signing a document, and faxing it remains the only option.

There are some minor advances--fax-to-email gateways allow faxes to come straight to my desk. I can use a scanner to create an image of the document and e-mail it. However, at the end of the day, it is an image file that is no better than paper for anything other than looking at it. As far as a computer cares, it is no different than a LOLcat picture. It can't be searched, and is relatively large to archive. It still takes time to deal with.

What is really surprising to me is that I work for a fairly high-tech, as-close-to-paperless-as-they-come company, and I'm dealing with other Information Technology companies who are requiring the signature. You would think someone would invest in software that would allow the creation and distribution of digital signatures. It would marry the authentication that people look for with a physical signature with the advantages of true digital formats. This is the type of innovation that would move us forward.

Until then I'll keep my pen inked and my fax machine plugged in.