67.202.27.129: The oldest EC2 Instance in History?

Back in September 2007 I began to play with an emerging service from Amazon.com called the “Elastic Compute Cloud.” I pulled out my credit card, signed up, and lit a new Xen server in about 5 minutes. That server lived for years at 67.202.27.129. I used it for my personal blog, scottpenberthy.com, as well as several sites that erupted from nighttime dreams of entrepreneurial hackery.

Yesterday I ran rsync to copy the local files to my computer, making sure I had the latest and greatest. Uh oh. Near the end of rsync the disk activity started to spike, CPU spiked, and then rsync died. I tried to resuscitate the patient several times, cycling through several reboots. No joy. SSH stopped responding. I officially declared it dead at 11pm this morning.

Who knows? Maybe Amazon finally sunset the server hosting my virtual instance. It was nearly 3 years old, after all. Still, I’m impressed. Who would’ve thought that a virtual instance would live so long? I’m going to miss you, good ol’ 67.202.27.129.

North Highland Partners

Peter Drucker said it best. Every business starts with a customer.

Earlier this year I helped a customer build a few apps. Apps to me are the next wave of widgets. I’m pretty excited about the prospects of the iPhone platform, Flash 10, and cloud computing. Apps are seemingly everywhere, on the phone, the TV, my car’s dashboard, blogs, Twitter.

The customer’s CFO emailed me. “Where do I write the check?” I quickly incorporated, naming the company after the street where we built our first house: North Highland Partners. Today I put a placeholder at our domain northhighlandpartners.com.

In the last tech wave, we’d build the Powerpoint, craft a Web 2.0 slik sight naim, look for investors, attend conferences, send out fliers, get a logo, design t-shirts. We’d buy traffic, hook up AdWords and more. We hoped to be the next Four Square, Twitter, or Facebook.

In this economy Drucker rules. Start with what you do best, get a customer to pay for real value. Deliver quality. Cash the check. Repeat. Referrals will build when you deliver. Thirty days in, we crossed six figures in qualified opportunities. Last week we crossed seven. I was stunned.

The App wave is growing. Fast. Now, back to coding. We’ve got this really cool iPhone app for a movie that… well.. can’t say yet. Its under NDA.

Tab bar Icons

Today I spent several hours pixel twiddling UITabBar icons. The iPhone UITabBar class requires icons to be in an unusual format, initially created for the Safari browser. Blogs were replete with recommendations. Use Gimp! No, use Photoshop and mess with channels! Image matte, alpha transparency, JPG, PNG, grayscale, etc. etc.

My head hurt.

I finally came across some neat utilities from my old friend ImageMagick. After some experimenting I created a script that would convert any image into a 32×32 icon. I was so happy when it finally worked! I kept getting square blobs, regardless of what I tried (gradients, multiple colors, image masks).

The script is now available. I call it Tabulicious. The page uploads an image, converts it to the iPhone format, and stores both in Amazon S3 for immediate download. Here are some hints:

• Use images with a clean background, preferably white
• The main character of your image should have darker outlines, and lighter colors inside.
• Vector art works better than photographs.

The Journal writes that “the iPod growth story is over, tough times are ahead.” I’m not so sure about that. The iPod touch and iPhone are full fledged computers. For $199. Several mothers I’ve talked to have purchased iPod Touches for just that purpose — to provide Apple computers to their kids for a couple hundred bucks. Apple has moved to another playing field altogether. They make the netbook, ThinkPad, and all the other Windows boxes look antiquated. The next wave of computing is in your hand, in the cloud.

I’ve been writing software on the iPhone. Objective C screams on the little processor and the local libraries and facilities are simply amazing. It blows away all the other mobile operating systems I’ve tried, and I’ve tried a bunch (Brew, J2ME, Symbian, Nokia, Blackberry).

Alex Welch is a terrific startup CEO. While I worked for him at Photobucket, the sales team would call frequently, begging and pleading for us to build a whitelabel version of the site. The revenue was real, meaty, and immediate.

Alex stayed his ground. We focused on free services for the consumer, responded quickly to feedback, and never wavered. The team, strategy, and focus would be too diluted by trying to satisfy multiple masters. Alex was right. His focus and dedication are rare.

Last week I met the founders of drop.io. It struck me that Sam is building exactly what Alex avoided — a whitelabel version of Photobucket. They’re focusing on one master with a solid revenue model. The need is real and they’re off to a terrific start. Check it out! I’m already using it to share pictures and documents with coworkers and family, away from the prying eyes of the Internet. Very slick.

Simon Wardley is amazing. He turns 185 slides into a thoroughly entertaining, unforgettable presentation. His transitions remind me of Steve Jobs pitching at Apple, his humor reminds me of Monty Python. Whack the rabbit. I hope you enjoy his presentation on the future of web applications; he nailed it! Its well worth the 20 minutes.

GoDaddy provides rudimentary email services that appear to work. Naturally, one of her first clients tried to send her an email… and GoDaddy blocked it. So embarrassing. Why? The email contained a URL. I turned off spam filters. GoDaddy still blocked it, giving us the following error:

Delivery to the following recipient failed permanently:

xxxx@theofficetogo.com

Technical details of permanent failure:
Google tried to deliver your message, but it was rejected by the recipient domain. We recommend contacting the other email provider for further information about the cause of this error. The error that the other server returned was: 554 554 The message was rejected because it contains prohibited virus or spam content (state 18).

Frustrated, I started looking for cheap email accounts. Then I remembered someone Michael Eisenberg introduced me to way back in 2003, David Koretz, CEO of Blue Tie. Michael couldn’t say enough good things about David and his company. Blue Tie just cut a deal with IAC to handle a lot of email for the various web properties of Barry Diller. Very promising!

Within a few hours I had a deluxe email service for the low price of $4.99 a month. Here’s what it took:

1. Sign up for the enterprise email service, for 1 user, for $4.99
2. Change the MX Records at GoDaddy so they point to the BlueTie servers instead of GoDaddy’s secureserver.net. You can find this form by first clicking on “manage your domains” then click on “full DNS control.”
3. Request that BlueTie handle MX records for your domain using their online form.
4. Wait a few hours for DNS servers to catch up. Have a cup of coffee, go for a walk, see some friends.
5. Set up your email addresses so that BlueTie correctly routes username.enterprise to username@enterprise.com
6. Update your POP3 email client so that it uses pop.bluetie.com, port 110 for POP3, and smtp.bluetie.com, port 25 for SMTP

Eric Schmidt. Dr. Half IBM.

This evening I was reading the post mortem from 3Q earnings announcements.  Google surged with a 26% increase in earnings, to $1.35 billion.  IBM was suffering a tad in Europe, but still posted gains from a year ago, earning $2.8 billion.

It took a minute for that to sink in.

Google, founded in 1998, now earns 48% as much as IBM.  A large portion of that money comes from a new market, tapping a need for people to reach each other at time of transaction, and sell their wares.  Most of this money is automated, moving through online systems, running in a cloud.  Contrast this with the labor, hardware-driven revenue of IBM and companies like them.

Amazing.

I don’t want a generic application in the cloud.  That’s probably what Richard Stallman is all in a tizzy about.  Just handing over all my private data to some application, that I have no idea where its running, scares the bejeezus out of me, too.  Maybe others are comfortable uploading keys to their life savings to Web 2.0 companies like mint.com.  Personally, it freaks me out.

Yet I’m still a huge proponent of cloud computing.  Why?  Simple.  Its the computing part I want.  I want raw storage, raw computing, raw bandwidth.  Let me encrypt my data and build my systems by hand, the open source way, so I know they’re safe.

Give me 60Hz 120V power and a three-pronged plug.  IBM, give us Blue Plumbing. 

Allow us to buy a raw Intel machine or Z-series capable of running any OS.  Sell us a wavelength on fiber for a truly private network.  Open up those gorgeous SAN machines and Netapp Filer wanna-be’s and let us buy reliable storage by the bit. Create vast arrays of cheap SATA drives like MySpace and Photobucket, then sell it to us as an S3 competitor. Sell us a chunk of your AT&T contract as cheap CDN bandwidth for a nickel a Gig. Turn MQ Series into an uber-reliable messaging service.

IBM engineers build incredible infrastructure; they often invent it. Its in their Poughkeepsie and Watson Blue blood. Don’t sell ‘em short by packing the Web equivalent of crapware into the cloud.

Amazon Web Services is a true cloud provider, at least in my book.  In 15 minutes I was able to Google for new information on their Elastic Block Storage, download the appropriate 128-bit key encrypted tools, create a 100Gig slice of a Netapp filer, and allocate it to my virtual machine that runs this blog.  That blew me away.  For kicks I created another computing instance in EC2, dropped my first instance, and re-attached the drive.  That was another 5 minutes. The cycle time is incredible.

Google gives us Python, forcing a language abstraction.  IBM forces us to eat a version of social networking.  Salesforce hawks leftover CRM systems as an app platform.  Those guys are living in the (recent) past, where applications running on a server somewhere have been relabeled as “cloud computing.”   If that’s all Larry Ellison is reading and hearing about from his marketers, no wonder he calls cloud computing “complete gibberish.” That’s not the cloud I want, either.