Zack Rosen's blog

Our long-time colleagues at Advomatic have just rolled out a service that Pantheon users have asked for since we first announced the project. They are the first hosting provider to offer direct end-user support for the high-performance "Mercury" stack, which is now available on all of their fully-managed hosting plans.

We're thrilled to be working with them on this and it's a big step forward for Pantheon. You can read the press release here, and you can learn about Advomatic's hosting service here:

http://www.advomatic.com/hosting/mercury

This means that you can now get all of the performance benefit of Pantheon without having to do any of the work to get your server up and running and keep the system up to date. Advomatic manages everything for you. They also offer 24X7 support, a 100% network uptime SLA, and very flexible hosting plans. They even offer an extended support for Pressflow/Drupal maintenance: no more worrying about security updates!

Have any questions? Send Advomatic an email.

There is something just awesome about using a Drupal site hosted with Pantheon. Snappy page loads will make you happy, and when your pages generate up to ten times faster you will really feel the difference. Even though it makes Josh nervous (since we're still in beta testing) it has been really fun for us to play with the first Pantheon powered sites out in the wild. In all we have tracked over 1,000 Pantheon servers launch so far during the beta test phase of the Mercury stack.

We're getting very close to a stable 1.0 release for the Mercury Stack. So far, the biggest feedback we have heard from testers was that A) The stack needed to be able to run everywhere, not just on Amazon EC2, and B) Once a server was launched, improvements and changes to Mercury needed to be portable to live environments.

To address B, we recently added support to Mercury for BCFG2, a server configuration management system. Thanks to the magic of BCFG2, all changes to the Mercury stack (managed in a BZR source code repository) can be easily pushed to live Mercury servers. This provides the best of both worlds, all the power of the Mercury stack available to those that need the flexibility of managing their own server. Those wanting to roll their own can branch a config-set from our public launchpad repository right now.

To address A, we have a lot of exciting announcements in store soon, but first..

New Pantheon Website

Say hello to the new Pantheon website designed by our own Nica Lorber. Click around, and you'll soon discover our big announcement...

Announcing the Mercury on Demand Service

Soon, we will be offering a Mercury on Demand hosting service built on top of the Rackspace Cloud! We will begin private beta testing with our first customers later this month.

Our first hosting package will be comparable to Slicehost, Linode, and other VPS providers in terms of service and cost, but with one big difference -- our server environments, available at the click of a button, will come set up with the full Mercury stack capable of running Drupal 200+ times faster.

Also, we're excited to leverage the wonderful libcloud project in offering our on-demand service. This powerful open-source python library allows us to use a single interface to talk to a growing list of cloud hosts, meaning we'll be able to utilize other clouds in subsequent service offerings going forward.

If you are interested in being part of our private beta program please sign up here. Otherwise, stay tuned, there will be more announcements shortly.

We just got out of Dries's 'State of Drupal' presentation here at Drupalcon Paris. With excitement for the Drupal 7 release mounting it was an opportune time to reflect on not just how quickly Drupal is progressing, but where it is exactly this project and all of us are going. At Chapter Three we are becoming increasingly convinced that Drupal has a fast approaching destiny to run - as a developer environment or as a hosted product - on cloud infrastructure.

To that end, we have been doing a lot of work researching and hosting Drupal on Amazon's EC2 service. EC2 is ready to use, it is relatively cheap and instantly available. And, with its ability to run pre-packaged server set ups (Amazon Machine Instances, or AMI's), it allows developers and users to get a lot more power from their server architecture "out of the box" and spend their time doing more Drupal and less system administration.

With the strong interest Josh has been getting on the Project Mercury AMI, we're doubling down on the effort to create open source, fast, and dependable cloud based development and deployment environments for Drupal. The goal is to make this effort a more community focused effort and work together to create divine server packages that run in the cloud.

We're calling this effort Pantheon:

We have three Pantheon Packages available for alpha testing, they run but they are all a work in progress:

• Mercury: Liquid metal fast site hosting environment for Drupal 6 that includes Varnish and Pressflow Drupal running on a highly tuned LAMP stack. While running on a single Amazon instance, Mercury can handle over 2,000 requests per second for cached pages while mantaining a server load of 0.02.
• Vulcan: Continuous integration/testing environment for Drupal based on Hudson. When you can run 2,345 Drupal unit tests every time you commit a line of code — along with Selenium browser tests and automated coder.module compliance checks — you sleep better at night.
• Aegir: Use Aegir to deploy Drupal sites literally at the push of a button. This initial release allows users to take Aegir for a test drive without having to devote/configure a whole server for that purpose.

Get started

If you are familiar with administrating LAMP, deploying a server from a Pantheon AMI package for the first time is relatively quick and painless even if you aren't familiar with Amazon's EC2 services. From start to finish it should take approximately 15 minutes. Read our set up instructions to get started.

Get involved

All of the development work for this project will be done out in the open on the Drupal EC2 group on groups.drupal.org. Say hello on the group if you want to get involved.

Meet up in Paris at Drupalcon

We are organizing a Amazon EC2 / Project Mercury / Pantheon Birds of a Feather session at Drupalcon Paris. Come find us at 2:50PM on Friday.

You should also feel free to find Josh, Matt, or myself here at the conference if you want to learn more about Pantheon, to get in touch just shoot us an email: zack [at] chapterthree.com, josh [at] chapterthree.com, matt [at] chapterthree.com.

Follow the project

We will be posting regular updates on Project Pantheon to our blog. We are also on twitter: @pantheon_drupal.

Forty years ago software ran on mainframes. In the eighties it moved to the personal computer. Right now, before our eyes, it is moving to the internet (or 'the cloud' if you will). Of course, this will create a spectacle. Over the next decade we will witness Microsoft grapple with fate — licensing software in boxes is a dying business — and Google's presumed ascendancy when much of the world's information moves online. But this isn't simply the titanic clash of industry giants battling it out for market share. The technology industry is not continuously overturned only by faster processing and the ongoing invention of new tools. Rather, the real uprooting is accomplished by disruptive new models for doing business with technology.

New models always seem outlandish at first. When IBM allowed Microsoft the licensing rights to resell DOS on commodity PC hardware, it had no idea it was giving away the keys to a kingdom which at the time existed only in Bill Gates' head. One billion PCs later, it is Microsoft stumbling into the new computing model's market with a product they pay more to create feature by feature than their start up competition does.

Folks out here call these cycles of technology "waves". For Drupal, I would call it a massive opportunity. Like any bell curve of adoption, what happened last year is approximately 1/4th as interesting as what happens next year. And what is happening in the coming years with commercially available "cloud hosting" services such as Amazon EC2 will be integral to the future model of delivering software and services over the internet.

These new services aren't just cheaper and easier to use than commodity hosting servers, most importantly they are imminently hackable. Suddenly it doesn't seem so outlandish that our Drupal, the same one that can get the job done on GoDaddy if needed, could soon run and scale on the same infrastructure powering the top websites in the world.

And if Drupal could do that... what else could it do?

I saw the utility of Drupal distributions first hand at CivicSpace. Beyond technology, the ability to discuss a web technology solution in the context of a specific market (in this case non-profits) is very powerful. Suddenly, you can translate abstract technology concepts ('tags', 'wikis', 'blogs') to specific pain points and use cases ('event organizing', 'taking donations', 'creating a dialogue with supporters'). This is a subtle difference of syntax to developers, but for a non-profit looking to build a website it's the difference between seeing Drupal as a potentially interesting technology rather than the tool best suited to solve their problem.

Unique markets require unique solutions. In the Drupal world this means different modules. Outside of the Drupal world this generally means different companies with unique web software. In the non-profit web technology market there are more than a dozen companies, each with different technology platforms. Convio, Kintera, Democracy In Action, and Blue State Digtal are the big ones. Look into any other web-site software market vertical and you will see the same pattern.

How many intranet products are there? Too many. But how many of them are as straight forward to develop for and extend as Drupal? This 'feature' instantly makes Open Atrium a compelling option. While its initial feature set is not extensive, it inherits a powerful API and a huge developer community. Most importantly, it does this without losing the specificity and nuance required of software competing head-to-head in the market place for intranets.

I think Open Atrium could be the start of something big. At the very least it is a beautiful product.

First things first, Drupal just isn't a very good product. I've spent the better part of the last five years working on Drupal. Not on the core patches (I'm no good at those) but on the other side of the technology, the part the people use. And I'll tell you for certain, while people very much want websites, they don't want just Drupal. Even if they thought they did, they won't soon after they log in for the first time with UID 1.

Wordpress and Joomla solve real-world problems, they make blogging and simple content management relatively easy. The code may be relatively brittle, the API unworkable, but when you turn them on they do something obviously useful. Can we say that about Drupal?

Of course, this failing hasn't mattered much to Drupal's growth so far. The fundamental metrics of Drupal's health as an open source project -- core patches, community members, deployed sites -- are still on an unbelievable exponential growth curve. For Drupal firms, it is still sellers market, there are more demand for Drupal services than qualified supply. Can Wordpress and Joomla say that? Not exactly.

The internet is comprised of more than more than two hundred million websites with the top 50 sites claiming 40% of the total internet traffic and the long tail of sites are scattered across thousands of commodity hosting service providers. Most of these websites are rather complex, all of them are unique.

I don't know if a Joomla, or a Wordpress, or Drupal will ever run a significant percentage of the internet's websites, but I do know that most websites require functionality beyond blogging and simple content management. A website can serve any digital purpose, that's the beauty of the web's organic and distributed invention. But this freedom brings inherit complexity for technology projects endeavoring to power a great many websites. This is a hard problem. It's the kind of problem that requires thousands of engineers working collaboratively for a number of years to solve.

While Drupal's success as a product so far is limited, it's success as a tool web developers use to build websites is remarkable.

I returned last week from South by South West. It was my third showing and by far my best experience. SXSW can be overwhelming, it draws thousands of messianic and work obsessed technologists plucked from the front lines of the internet and buzzing with the fervor. Yes, the over-hype is distracting, but the few days of conversation with the innovators at the edges of the internet over world class BBQ makes it well worth it.

I first went in 2003, that was six years ago. The internet as a medium has changed dramatically since then, it's become much more useful and compelling. More exciting though are the underlying advancements under formulation for years that are just beginning to bear fruit. My trip to Texas got me as excited as ever about what the future will bring given what is now possible.

Compounding advancements in computing technology and the increasingly streamlined distribution channel available on the web for new businesses will yield plenty of surprises over the next five years. In the past five years there has been a flurry of innovation; blogging, Facebook, Youtube, Twitter, Flickr, virtually every staple of 'Web 2.0' was created or popularized in the past five years. Discerning which new applications will take hold over the next five is a $100BN question, but recent advances in platform technology available on the web give a glimpse of what will soon be possible:

1. Cloud computing services (Amazon EC2, Google App Engine etc.) are driving the cost of web hosting and storage down dramatically for innovators, but more importantly, are providing a clear path for applications to reach internet scale. In just a few years access to these services have gone from theoretically possible, to technically possible, to now totally possible and even advisable.
2. New application distribution channels (Facebook, iPhone) are making it vastly easier and cheaper to create and market innovative applications and bring them to scale.
3. Data interchange standards (Oauth, OpenSocial, etc.) are finally starting to make possible the dream of seamless and costless data interchange between web applications. This will enable a host of new applications that can re-use existing data sets and content published on the web.
4. Web frameworks (Drupal, Django, Ruby on Rails) are making it vastly cheaper and easier to develop custom applications. For application developers, building on a web app frameworks is quickly becoming the norm.

Back in 2003 these tools were either in their infancy or still untested ideas, this year at SXSW they were all common knowledge.

Kieran and I from Acquia are organizing a brief, lively, and tasty introduction to Drupal for geared for professionals in Washington DC at DrupalCon. It will take place in Thursday, March 5 from 12:30 to 1:30 pm during the DrupalCon conference – an event that’s bringing 1,300+ Drupal developers to learn, share ideas, and advance Drupal. For those not familiar with Drupal and its community, being immersed in DrupalCon DC – with some instruction – is the perfect way to find out more.

Our discussion will be geared towards professionals reviewed Drupal. We’ll demonstrate the software’s capabilities, discuss the support services available, and give an orientation of the open source community behind it. We’ll also dive into case studies of Drupal websites and will explain the production process behind them from start to finish. After the presentations, we will have plenty of time for questions and answers with Drupal experts.

The event costs $20 and includes lunch.

Details:
Thursday, March 5
12:30 – 1:30 pm
Washington Convention Center
Enter in the L street entrance, by 9th Street
Mount Vernon Square/Convention Center metro stop on the yellow and green metro lines

Register:
http://dc2009.drupalcon.org/drupal-lunch

If you have any questions about the event, please contact our event coordinator Rachael Boggan: rachael [at] chapterthree [dot] com.

We gave our first Drupal Lunch presentation with Kieran Lal at Acquia last Thursday. Thirsty Bear served a tasty buffet and we lucked out with attendees. From e-commerce companies investigating a Drupal re-launch to communication consultants researching the latest web technology, we got a chance to meet with the earliest of Drupal adopters climbing the Drupal project's learning curve. John Faber graciously recorded the audio of the presentation here if you are curious. It was a real pleasure for Kieran and myself and we are looking forward to producing the next events in the near future.

We are working with Development Seed to hold a Drupal Lunch at the DC Drupal Con in a few weeks. Registration will be open shortly and we will follow up with more details. We are also planning the next San Francisco Drupal Lunch, likely to focus on designing sites with Drupal. We'll be posting more on both of these events in the near future.

Last week my friend Mark Dwight got in touch with me and asked if I could fill in for him presenting for the Teens in Tech conference. Mark had a very successful run heading up Timbuk2, taking the company from $4M in sales to $20M before selling to private investors and famously netting the sewing staff at the factory more than $1.2M in bonuses. Mark has a new start up going a few blocks from the Chapter Three office called Rickshaw Bagworks. They make some truly beautiful bags and are very serious about minimizing their environmental impact, focusing on recycled and domestically sourced materials.

The Teens in Tech conference assembled a scarily young group of technology entrepreneurs to discuss new trends in technology adoption and to compare notes on entrepreneurship at a young age. Mark was scheduled to speak on "Going Green" as an entrepreneur which he can speak to with a lot of experience. I on the other hand, lacking Green cred, talked about my experience in entrepreneurial pursuits in politics during the Dean Campaign, with non profits at CivicSpace, the world of Drupal at Chapter Three and our increasingly more serious bicycle business.

It's been true for some time now that computing technology wields such powerfully disinter-mediating social and market force as to afford even young neophytes the ability to create or up-end entire industries. In some monied pockets around here in Silicon Valley inexperience and youth can even be considered a virtue in running a technology start up. While I wouldn't go that far myself, it is clear that advances in computing technology can progress so ridiculously quick that creating real innovation is possible and relatively accessible to anyone with the passion and a computer. The Drupal development community exemplifies this. If you can install a web server, download Drupal, and can collaborate with others, you have everything you need to impact the world.