A little after I wrote up the news yesterday that salesforce.com was acquiring Heroku I came across a quote from Marc Benioff, tweeted by @timanderson
“Ruby is the language of the cloud 2”
Computerweekly has a fuller version:
“Ruby is the language of Cloud 2 [applications for real-time mobile and social platforms]. Developers love Ruby. it’s a huge advancement,” said Benioff. “It offers rapid development, productive programming, mobile and social apps and massive scale. We could move the whole industry to Ruby on Rails.”
Apart from conflating Ruby on Rails the development framework with Ruby the programming language, something which many Ruby developers hate, but which its easy to slip into (so easy i slipped into it yesterday) Benioff was clearly not talking the language of the line of business, but rather of technology.
It seems to me we’re witnessing a sea-change at the moment. Many of the commentariat, industry analysts and so on, still seem to think the purchaser is king – that IT will simply be bypassed by savvy business users. But what about developers? We are currently emerging, blinking, from a dark period when many enterprises, and their advisors, truly believed we could just draw up a business process diagram and ship it offshore for coding by commodity developers. There was a Taylorist view of software development, an artefact of Waterfall development. But the model was broken. Innovation comes from code, and code comes from software developers. Coding is social. The idea developers are a commodity is as broken as the notion that a free market consists of independent actors making independent decisions that always lead to better outcomes. We are people, we are herds, we might as well be wildebeests. Well Benioff wants to be the alpha male, showing the way to the next watering hole. He isn’t alone.
I wrote a couple of months ago about a bravura performance from VMWare CEO Paul Maritz. He said:
“Developers are moving to Django and Rails. Developers like to focus on what’s important to them. Open frameworks are the foundation for new enterprise application development going forward. By and large developers no longer write windows or Linux apps. Rails developers don’t care about the OS – they’re more interested in data models and how to construct the UI.”
And this was an IT Operations audience!
The bottom line is this – if we’re really going to consumerise enterprise IT then that means developerising it (please excuse the hideous neologism). Just try and find me a successful consumer tech company which hasn’t fostered amazing, more often that not local, relationships with software developers. The old model was busted. Great users experiences generally come from developers and designers working closely together. That’s the bottom line. We all owe Apple a debt for helping business owners to understand the value of developers.
It seems to me that Benioff and Maritz are reflecting a powerful change. As I said yesterday:
Salesforce avoids IT to sell to the business, while Heroku avoids IT to sell to developers
Powerful macro forces are at work, driven by the cloud, the appstore, open source, social media and so on. CEOs need to get a lot more developer savvy. Its not enough to pimp your own apps and APIs, you need to embrace the herd. Maritz and Benioff are wise to that. Developerforce. Yup.
disclosure: salesforce and vmware are both clients.
further notes: in RedMonk parlance dorks, geeks and nerds are good things. we are all about the makers.
Rick Bullotta says:
December 9, 2010 at 3:25 pm
Been there, seen that. A lot of new innovation occurs when creative technologists are provided with tools (and time/freedom) that enable them to do things better, faster, cheaper.
i claim that very little competitive advantage was EVER achieved from purchased line-of-business applications. it’s the innovations around the core and in whitespace areas (or microapps) that have a real effect. there are zillions of high value uncaptured opportunties each year that traditional IT approaches leave on the floor.
we need a new class of application platforms, though, for a new class of problems and a new generation of users and developers (we’re working on one of our own at ThingWorx). i don’t think enough attention is being paid to the importance of domain-specific platforms overall, either.
And I know that you can build a healthy business in this space. Did it once. Plan to do it again.
once the developers and business owners get their chance to unleash their creativity and put their ideas into live application in a fraction of the time/cost, then the hard ROI stories start appearing that help build the case for the traditional buyers to support these creative developers/users and take the innovations more broadly across their companies.
i think these new technologies can start a wave of “continuous innovation” (or microinnovation) just like we saw “continuous improvement” as an operational mantra in the 90s and 00’s.
happy to share some specific experiences in this area anytime you want.
Developer-focused Heroku wins big with SalesForce Acquisition « Lew's Blog says:
December 10, 2010 at 4:07 am
[…] and developers took note. New Relic shares this same passion to simplify the job for the newly re-emerging developer who increasingly has responsibility for the app’s success in production. For many Heroku […]
Did Heroku really buy Salesforce.com developers? « rand($thoughts); says:
December 10, 2010 at 9:14 am
[…] would like you to think. Here’s how Salesforce.com’s CEO, Marc Benioff, described the motivation behind the acquisition: Ruby is the language of Cloud 2 [applications for real-time mobile and social platforms]. […]
Pat Patterson says:
December 11, 2010 at 4:54 pm
Don’t let Benioff hear you calling him a ‘software CEO’ 😉
Even the salesforce.com HQ wireless SSID is NO-SOFTWARE 🙂
Pat Patterson says:
December 11, 2010 at 5:34 pm
Seriously, though – we’ve seen exactly this kingmaking happen over the years with Solaris (and the other Unixes) vs Linux. Developers and Sys Admins picked up an operating system that was close to what they were using at work, but could run at home (utilizing Clay Shirky’s ‘cognitive surplus‘).
As the Linux ecosystem grew, dorks slid across so they could work with the same stuff all the time. As this continued and became a virtuous cycle (for Linux), the pool of Solaris/Unix talent dwindled to the extent that enterprises’ decisions were influenced by the scarcity, and therefore, cost, of that talent. I’ve seen this first hand, from both sides of the fence – outside Sun I worked on a project for which ZFS would have been a great fit, but a strategic decision was taken to use Linux and write a lot of the equivalent functionality from scratch, driven in no small part by the difficulty in finding Solaris dorks.
James Governor says:
December 14, 2010 at 5:29 pm
Pat – right on the money. thanks for taking the time.
Stu says:
December 29, 2010 at 4:52 am
This is way, way too simplistic. Sorry if this comes off as a bit of a rant, but, I think you hang out with vendors too much if you really believe that developers are the new kingmakers.
So far, all evidence I’ve seen points less to an awakening from a Taylorist mindset to bifurcation in the world between Enterprise IT and Consumer Web approaches to How Business Gets Done. One is heavily outsourced, optimized, and process focused; the other is more collaborative, agile, and developer-driven. Many, many enterprises still do not do agile software development; or if they do, it’s really more of a label on top of sloppy waterfalls instead of a true discipline. Outsourcing is growing, not shrinking. Offshoring continues unabated. Agile software development is not universal, nor is it even uniformly agreed upon by IT practitioners. Annual capital budgeting continues to rules IT expenditure and projects.
I think the worlds and mindsets of Enterprise IT vs. Consumer Web technology are so fundamentally different in their assumptions and priorities that while there will be bridges and practices that are cross-adopted from consumer to enterprises, they will be selective and hard to predict, and for at least the next decade it’s going to be “business as usual” in many respects. Benioff and Maritz want to be the ones to make those bets early, but it’s far from clear that they’ve called them correctly. Frameworks like Django and Rails do not solve the vast majority of enterprise IT problems — most of which have to do with managing, integrating and incrementally improving a legacy, rather than building on a greenfield. Part of the challenge is to get people who have worked in a consumer web environment to work in an enterprise (or large outsourcer) to foster the transition. Most developers would probably prefer to be part of a greenfield and roll the dice on stock options rather than deal with migrating 10 year old technology to 5 year old technology that needs to integrate with 20 year old technology. Yet no amount of developer ingenuity, marketing, or “cloud” will eliminate the reality of the legacy.
I’ve not even really touched that developers often (I’d say more often than not) get it wrong. For every blazing open source success, there are a thousand corpses floating on Freshmeat.net. For every reasonable attempt at building a RESTful API, there are tens to hundreds of mockeries of the architecture, often just out of “developer convenience” or “not thinking” rather than a weighing of tradeoffs. Then there are debacles like Digg’s move to Cassandra, because the “developers” knew best about how to manage data at scale.
Don’t get me wrong, there is plenty to learn out there, and huge opportunities in the enterprise to find the bridges from consumer web tech. But to call the developers kingmakers is silly. That would be like saying all startups just need are good developers, yet it’s almost a cliche to see good technology fail because of horribly flawed management or marketing strategies. You can’t produce results with just one role – they’re all involved in “making”.
In sum, this “sea change” seems more to be like a five to eight year journey that has barely begun. Patting developers on the back doesn’t really make it move any faster, as they’re not sufficient. It’s a fucking huge hill to climb. But it’s there, so, why not try? Let’s just be a bit more realistic about what it will take. There are lots of talented developers, fewer talented architects, and even fewer insightful businesspeople that can draw a line between a business opportunity and the tech. It’s never just been about the code. It’s also about the architecture, and then, at least in Enterprise IT, it’s about the business application of the technology. It’s rare for a developer to maintain insights in all of these areas.
Ken says:
December 30, 2010 at 11:10 pm
Wow, Stu. Some great points on the divide (Monkchips, it goes without saying, great insight as always). Although I differ on the nature and permanence of the divide.
Having come from a Python on some god-awful framework on our own hardware to Ruby on Rails on AWS, I am amazed at the increase in speed of development and length of runway available to build scalable applications. It’s not even close what dynamic languages and frameworks, massive open code libraries, rich ux/ui sharing and learnings, and the cloud can bring.
Yes, legacy is legacy and existing enterprise applications will be around for a long long time. But for enterprises that want to develop new business lines, that want to reach customers in new ways — whether it’s new frictionless lines of business or collecting data from smart devices (autos, garden sensors, network adapters) or whatever — they’ll do well to do things in a Consumer Web way.
I don’t see it as a them vs thar. I just see the Enterprise as late to the game but waking up. The Consumer Web understands how agility and speed allows you to iterate and beat your competition (Boyd and OODA loops and all that). They also understand minimal viable product and measuring every click and action. On creating loss leaders (via freemium) to create connected relationships and better understand where they provide value and what segments will pay what. On using realtime a/b testing of pricing, headlines, buttons, and just about everything that goes on a page.
I think you take the Monkchips use of the term “developer” too literally. The Consumer Web developers are leading the way not only in many development areas (cloud, NoSQL, realtime a/b testing, ci, etc) but also in the use of the marketing and product development approaches mentioned above. RoR and Python/Django developers and development shops aren’t just developers. They’re business/product marketing/ux-ui/development — sometimes all in one person, other times as part of small agile teams.
Enterprises do need to wake up. They need to understand that the speed of IT and the speed of product development has changed by many x. And that cloud and dynamic languages and frameworks (and consumer web patterns and principles) are no different than what IP, email, http, HTML, CGI, and Perl were to enterprises back when they were still taking MIS and why would you need a website.
James Governor says:
December 31, 2010 at 3:02 pm
Stu – i can’t really argue with a great deal of what you say – at redmonk we take an aggressive advocacy position on the value of good developers and ops people – but we all understand legacy pretty well. Ken’s arguments are also powerful – and i think he states them better than I do. The status quo really is not an option – and Taylorist approaches just don’t work in software development. If enterprises are going to cope with business disruption they’re going to need to wake up fast. “Late to Game and waking up” – great phrase.
Developer-focused Heroku wins big with SalesForce Acquisition | New Relic blog says:
February 1, 2012 at 9:02 pm
[…] and developers took note. New Relic shares this same passion to simplify the job for the newly re-emerging developer who increasingly has responsibility for the app’s success in production. For many Heroku […]