Friday, September 11, 2009

Grand Central Dispatch now Open For Business

I recently gushed about Apple's new Grand Central Dispatch code at the heart of the recent Snow Leopard Mac OS X release. Good news. Apple has decided to Open Source the code under an Apache license.

Drew McCormack over at MacResearch speculates as to why in Grand Central Now Open to All | MacResearch:
There could be one last reason why Apple has taken this step: they want to use Grand Central to push the adoption of other technologies, in particular, blocks. Blocks are an extension to C which form the basis of Grand Central Dispatch. Having your operating system based on a non-standard language is not a good position to be in, and Apple would surely like to see blocks incorporated into the C language. By offering Grand Central to the broader programming community, they may be hoping it will catch on, and make the argument for incorporating blocks in the C standard that much stronger.

I agree.

The next step is for this code to be picked up in a Linux Distribution. Having simple concurrent programming, integrated tightly with C code (and C code derivatives) would be a boon to many industries looking to exploit multicore.

Thursday, September 3, 2009

Petabytes on a budget: Cheap Storage

The folks at Backblaze provided a very nice, and detailed look into how they create cheap, reliable storage to host their customer's data. Via Petabytes on a budget: How to build cheap cloud storage | Backblaze Blog:
At Backblaze, we provide unlimited storage to our customers for only $5 per month, so we had to figure out how to store hundreds of petabytes of customer data in a reliable, scalable way—and keep our costs low. After looking at several overpriced commercial solutions, we decided to build our own custom Backblaze Storage Pods: 67 terabyte 4U servers for $7,867.

Nicely architected 4U storage pods. It's very unusual for a company to explain, in such detail, how they built a core layer of their system. I love that they included the 3D design of the enclosure, as well as a detailed Bill Of Materials list. All of which was really only possible because of the tremendous amount of work tha went into Linux and the Linux storage subsystems.

Let's hear it for Open Source software (and Open Source design)!

2009 OpenSourceWorld: Presentation Picks

Didn't make it to the 2009 OpenSource World, Next Generation Data Center and CloudWorld conference? No worries. Enjoy the presentations right here, from your recliner: OpenSourceWorld: Presentations .

Be sure to check out the following (in no particular order):


Wednesday, September 2, 2009

Islands of serialization in a sea of concurrency

John Siracusa does an awesome job writing his review of mac OS X Snow Leopard for As Technica: Mac OS X 10.6 Snow Leopard: the Ars Technica review - Ars Technica.

Grand Central Dispatch is far and away the coolest technology that I've seen created in a long time for super-simple concurrent programming that fits your existing programming style/model.

If they were handing out Oscars for programming, the developers of GCD and the "blocks" extension to the language would get one. Heck, if they were handing out Nobel Prizes for programming, this would get one!



Makes me happy I'm an AAPL shareholder.

Tuesday, August 25, 2009

Bill Coleman Joins 3tera Advisory Board

I think this move surprised a number of people, since Bill recently wrapped up Cassatt Corproation, getting the technology and people acquired by Computer Associates.

However, I was not surprised at all. The announcement, via 3tera Welcomes Bill Coleman:
You may or may not have seen the recent press realease.  Bill Coleman, IT/Silicon Valley luminary, Founder and CEO of BEA Systems, has joined 3Tera’s Advisory Board.

Yes, this alone is a great testimonial to what we have accomplished in our field.  Getting dignitaries such as Bill does not come easy.  But here’s the best part - this has a lot more than just marquee value and I doubt that Bill would have joined us if that was the case.  Bill, especially since his most recent stint as Founder and CEO of Cassatt Systems, is an extremely knowledgeable visionary in the area of utility and Cloud Computing; and, data center automation.

So, Bill will be extremely valuable, reviewing and tweaking both our business plans and technology as we forge ahead to maintain our lead at enabling Cloud Computing in enterprises and service providers.

Bill and I, both Founders of Cassatt, clearly share the same good taste in great technology. I've been following 3tera since 2005 and continue to be impressed with their accomplishments each year. I initially wrote about them on this blog back in 2006 in the article Amazon EC2 Killer Apps – Meet 3Tera.

I wish Bill and 3tera the best of luck as they move forward chipping away at the cloud. Bill will be an excellent addition to the 3tera Advisory Board (really, he should be on their Board of Directors, but nobody asked me...).

Friday, August 21, 2009

YACCI - Yet Another Cloud Computing Intro

From Liming Liu of, check out: Cloud Computing Introduction. Nicely done, and is an indication of how fast this market is changing. Finally.

Friday, August 14, 2009

Databases in the Cloud

Ken North writes a great article in Dr. Dobb's Journal on Databases in the Cloud. Check out Dr. Dobb's | Databases in the Cloud: Elysian Fields or Briar Patch? | August 3, 2009:
The cloud is an elastic computing and data storage engine, a virtual network of servers, storage devices, and other computing resources. It's a major milestone in on-demand or utility computing, the evolutionary progeny of computer timesharing, high-performance networks and grid computing. The computer timesharing industry that emerged four decades ago pioneered the model for on-demand computing and pay-per-use resource sharing of storage and applications. More recently Ian Foster and Carl Kesselman advanced the concept of the grid to make large-scale computing networks accessible via a service model. Like computer timesharing and the grid, cloud computing often requires persistent storage so open source projects and commercial companies have responded with data store and database solutions.

The more things change, the more they stay the same...

Sunday, August 9, 2009

Doubling Down - Good Money After Bad

This post by Fred Wilson certainly resonated with me, and caused me to reflect about my time with Sevin Rosen Funds. Via Doubling Down:
Like most VCs, I am guilty of sticking with our investments too long and putting too much money into the ones that are not working. It's an occupational hazard. As I've gotten more experience in the venture business, I've gotten better at this part of the business, but it is still a challenge for me and most VCs I know.

Bliss McCrum, one of the two VCs who taught me the venture business early in my career always said, "if you are going to put more money into a company that is not working, make sure to change the strategy, team, or cost structure, or all three." It's good advice. You will not get a different result doing the same thing.

During my time at Sevin Rosen Funds, I saw quite a few presentations made by existing portfolio companies looking to raise a Series C or Series D round of investment. As the "new guy", I didn't have the history of what transpired to get the company to this point -- I was just evaluating the business as it stood at that moment.

More often than not, my advice was that the company was not a good candidate for follow-on investment. Invariably, the company did receive their follow-on investment.

It wasn't until I fully understood a number of things about VC that this all started to make sense to me: The way that VC funds are managed, how VCs are compensated, how money is allocated for follow-on rounds in the fund, the timing intricacies of when a new fund can be raised, the details hidden within LP agreements on how investments can be made when multiple funds are being managed by the VC entity, and the psychology that comes into play once your investment in a company is a public one (on the Portfolio page of your firm's web site) and you have become emotionally attached after a series of dozens or hundreds of board and strategy sessions.

Needless to say, my comments to SRF must have been heard as "your baby is ugly". And, let's face it, nobody wants to hear that their investment did not turn out to be a good one. With money allocated in a previous fund, and the LP agreement of the current fund prohibiting new investments in prior funds, say, some VCs may in fact be conditioned to double down more often than you would expect.

I completely agree with Fred Wilson and Bliss McCrum. Doubling down only makes sense when you change something core to the business. All too often, the VCs will do too little, or nothing at all, when doubling down. Such a strategy amounts to nothing more than hoping for some external change ("the market will finally start to take off; the hockey stick is just about to shoot up!") to make your investment pay off. That's bad business. That's throwing good money after bad. And the LPs will not like that.

Monday, July 27, 2009

Zappos Founders Do Well In Sale To Amazon

Last week, Amazon acquired Zappos for roughly $847M. The founders did very well. Via What Everyone Made from the Zappos Sale.
It’s much harder to see how much Hsieh and Lin are making, because no one knows how much of the proceeds of the Venture Frogs shares goes to their pockets or to their LPs. But they clearly did well. Hsieh made at least $214 million; Lin made at least $18 million, with the Venture Frogs shares netting an additional $163 million. If that’s a forced sale, the two are crying all the way to the bank.

Zappos was all about execution. When a founding team executes this well, they certainly deserve the spoils. Many congratulations to them and all the Zappos employees for building a really great company.

Investor Sequoia returned about $160M on their $35M investment. That's a nice cash-on-cash return for them. Congratulations to them for supporting their investment through multiple rounds and helping to build a really great company.

Amazon gets a great asset.

Customers win in the end.

I love seeing these kinds of results for entrepreneurs, investors, employees, and smart acquirers. Congrats to all!

Google Voice iPhone App Denied?

Looks like the Google Voice iPhone application has been denied from the Apple iPhone App Store. Via Apple Is Growing Rotten To The Core: Official Google Voice App Blocked From App Store:
Earlier today we learned that Apple had begun to pull all Google Voice-enabled applications from the App Store, citing the fact that they “duplicate features that come with the iPhone”. Now comes even worse news: we’ve learned that Apple has blocked Google’s official Google Voice application itself from the App Store. In other words, Google Voice — one of the best things to happen to telephony services in a very long time — will have no presence at all on the App Store. If there’s ever been a time to be furious with Apple, now is it.

I'm completely bummed, and disappointed in Apple. And, I'm a huge fan-boy, so this is hard for me to admit.

Looks like I should have purchased the GV Mobile appication while it was still available, instead of waiting for the announced upcoming release of the official Google Voice application.

If AT&T doesn't want free SMS, then they should disable that within the applications, not completely remove them. Using Google Voice doesn't reduce the number of minutes that I burn with AT&T, since my Google Voice account is linked directly to my AT&T iPhone. Everybody should be able to win here.



Saturday, July 25, 2009

RackSpace Opens The Cloud, But Needs To Do More

Rackspace is wise to go open with their API and work to build an open community around cloud services (via RackSpace Opens The Cloud):
Rackspace is open-sourcing the specs for its Cloud Servers and Cloud Files APIs under the Creative Commons 3.0 Attribution license, enabling third-party developers to copy, implement and rehash them as they see fit.

In addition, The Rackspace Cloud (formerly known as Mosso) has made available Cloud Files language bindings along with technical guidelines for Java, PHP, Python, C# and Ruby under the MIT license through GitHub. Rackspace aims to offer a reference implementation in Python soon and in a press release casually mentions it “is aware of Ruby, Perl, Java, and Twisted Python Cloud Servers bindings”, which are all in the process of being developed.

However, I don't think it's enough to offer up an API. When I think about how successful NFS became, it's clear to me that it was the combination of offering up the API as well as a reference implementation and test suite. The NFS team also sponsored "Connectathon" events where everyone could check the interoperability of their ports.

I know that the two domains are not the same. I know that the NFS implementation was not "open source", but rather commercially licensed (there really was no equivalent open source concept like the Creative Commons in the 1980's).

The point I'm making is that if Rackspace wants this to truly become the industry standard, they need to actively foster the community. Even to the point of enabling their direct competitors with a reference implementation of their server side abilities. Rackspace should be confident enough in their lead and implementation to do so.

Wednesday, July 22, 2009

How Much Information? 2009 Update

I recently saw this:
Together, the cloud and the next generation of intelligent infrastructure will also help us navigate a world where information is growing exponentially. With less than 20 per cent of the world’s population currently online, the amount of digital information doubles every 18 months. According to industry analyst firm IDC, 988 exabytes of data are added annually to the digital universe, about 18m times the information in all the books ever written.

via / Technology / Digital Business - Technology critical to turning mass of data into useful insight.

And it reminded me of the seminal Berkeley paper, How Much Information, most recently released in 2003, I believe.

Yeah, there might be a market for archive and storage...

Doubles every 18 months, eh? Sounds a bit like Moore's law, but applied to bits. "The number of bits of information in the world doubles approximately every 18 months". I like that thought.

Monday, July 20, 2009

In VC deals, It's Pre Money + Option Pool Size That Matters

Jeff Bussgang of Flybridge Capital has a great post up about the nuances of early-stage VC financings. For first-time entrepreneurs, there is a lot of confusion about what financing terms really matter in the term sheet.

via Seeing Both Sides: In VC deals, Price Doesn't Matter - But The "Promote" Does

Jeff talks about the "Promote", which most entrepreneurs really don't consider. He says, "The “promote”, as we have called it, is the founding team’s ownership percentage multiplied by the post-money valuation."

An example:
This relationship between option pool size and price isn’t always understood by entrepreneurs, but is well-understood by VCs.  I learned it the hard way in the first term sheet that I put forward to an entrepreneur.  I was competing with another firm.  We put forward a “6 on 7” deal with a 20% option pool.  In other words, we would invest (alongside another VC) $6 million at a $7 million pre-money valuation to own 46% of the company.  The founders would own 34% and we would set aside a stock option pool of 20% for future hires.  One of my competitors put forward a “6 on 9” deal, in other words $6 million invested at a $9 million pre-money valuation to own 40% of the company.  But my competitor inserted a larger option pool than I did – 30% – so the founders would only receive 30% of the company as compared to my deal that gave them 34%.  The entrepreneur chose the competing deal.  When I asked why he looked me in the eye and said, “Jeff – their price was better.  My company is worth more than $7 million”.

In this example, the "promote" works out to be about $4.5M in either case, so the offers are exactly the same from the "promote" perspective.

However, the better response for this entrepreneur would have been to go back to the VC with the $9M pre-money valuation and argue that the option pool of 30% was way too big (and, it really is -- what the hell are you going to do with a 30% option pool in your Series A timeframe? Really?). Reducing the option pool to 15% would have made the "numbers" side of these deals a lot more palatable, and resulted in a "promote" of $6.75M. Clearly the better deal.

Many thanks to Jeff Bussgang for this excellent article.

Friday, July 17, 2009

iPhone App Store + Ad Revenue == Developer Gold Rush

What I find most interesting about this story is that it is possible to make $2,000 a day with the most simple of simplest iPhone applications -- based ONLY on Ad revenue and a FREE application. The app shows elegant frames around a black screen that you can use as a pseudo-mirror. That's it. $2K a day! Check out the graphs on ad hits here:

Top iPhone App Developer Was Losing Out On $2000 A Day Because Of Sloppy Coding .

The iPhone App Store is an amazing gold rush for application developers. I said it before. Apple has nailed it with their App Store, and the recent additions in iPhone OS 3.0 (subscriptions and in-game purchases) confirm my belief.

Apple is taking the friction out of developing and releasing mobile applications. That's why their mobile platform is succeeding.

Can others take as much friction out?


Thursday, July 16, 2009

Sand Hill Slave Returns to Blogging

I completely missed that Sand Hill Slave has returned to blogging last month! Entertaining reading for anyone in and around the VC industry.
No apologies, no excuses. I'm similar to the man who dates a woman forever, and then goes Casper on her. If it's meant to be, he comes back with that one grand gesture- out from a robin's egg blue box...a glorious stone to slip on her finger! Like that guy, I've also returned with a simple gesture no less; now it may not involve a ring, but it definitely involves a certain finger….

via Sand Hill Slave: Even Though I Walk Through the Valley in the Shadow of Tech, I will fear no VC... - Journal - The Office Space Scheherazade Returns- .

First Look: The Palm Pre Mojo SDK and Palm Acquisition Thoughts

First glance at the Palm Pre Mojo SDK is that it's a pretty nice piece of work. Developers doing web-based programming today will have no problem making "chicklet" applications by the thousands.

Thinking out-of-the-box for a moment, Palm would be a good acquisition target for Adobe. The Mojo SDK could certainly use the Adobe expertise in web-based development tools. Alone, Palm really doesn't have a chance of catching up with Apple.

Unfortunately, the Palm "App Store" will not be available until fall:
Thousands of developers have participated in the Mojo SDK early access program since it began in early April. New applications are in the pipeline for the Palm App Catalog, and the App Catalog submission process will be opened to all developers beginning this fall.

via The Official Palm Blog: Mojo SDK available to all .

Wednesday, July 15, 2009

Twitter + Wordpress Ultimate Integration

Here's how I integrated my Twitter status updates into my Wordpress blog. I wanted to load all of my old, archived tweets as "Weekly Digest" combined posts, but I didn't want those digests to show up on my home-page, and I didn't want those digests to show up in my RSS feed. I wanted all of the digest tweets to be accessible via the new Tweets category that I just set up.

Here's what I did:

  1. Create a new category to hold your new tweets. I created the Tweets category.

  2. Download, install and activate the Lifestream Wordpress Plugin. I pulled straight from the Subversion repository with the command (from the wp-content/plugins directory):
    svn co lifestream

  3. Add a new "Twitter" feed to the Lifestream plugin. This will get you the most recent 20 tweets, which was not what I wanted. An earlier version of the plugin could download all archived tweets, so I downgraded to version 0.99.4 with the command (from the lifestream directory):
    svn sw

  4. Re-add the Twitter feed to Lifestream, which should give you all (or, at least most) of your archived tweets.

  5. Unfortunately, Lifestream does not set up Weekly digest posts for all the existing events/tweets that it loaded. So, I had to hack on lifestream.php to do that. If there's interest in the comments, I will post my diffs here. Let me know!

  6. I ended up with 42 weekly digest posts being created. Unfortunately, they all showed up on my blog home page, as well as in the RSS feed. Not what I wanted.

  7. Download, install and activate the Ultimate Category Excluder Wordpress plugin.

  8. Configure the Category Exclusion plugin to exclude the Tweets category from the Main Page and from the Feeds.

And, that's all there is to it.

Your Main Page and Feeds stay clean, but your Tweets (and any other feed/stream that you want to include) will get archived with your blog.