Native Client – a technology that allows developers to run native compiled C and C++ code as part of their web apps – has long been a part of Google Chrome. Even though other browser vendors haven’t adopted it yet, Google is clearly putting quite a few resources behind this technology and at I/O this year, it announced Portable Native Client (or PNaCl, which Google says we should pronounce as “pinnacle”). PNaCl is now available in developer preview in Chrome 29 and will slowly find its way into the stable version over the coming months.
PNaCl, the company says, will allow developers to write web applications “that are truly architecture-independent.” It’s essentially an architecture-independent version Native Client, so unlike now, developers can write their apps and know that they will run on ARM and X86 (both 32-bit and 64-bit). PNaCl, the team says, uses an LLVM compiler infrastructure with a “compile -> link -> translate” workflow that creates an intermediary bitcode, which is then translated locally for the specific infrastructure.
That’s some pretty complicated stuff, but essentially it will allow developers to write high-performance apps that offer near-native speeds for today’s existing platforms and they can be sure that these apps will also run on new architectures as they become available without having to rebuild their apps (assuming, of course, that Google will continue to support this product).
Other browser vendors, of course, are also trying to speed up web apps. Most recently, for example, Mozilla announced the asm.js project, which ports C and C++ code to asm.ja – a subset of JavaScript. This scheme, Mozilla says, allows JavaScript code to run at speeds within 2x of native performance. That’s not quite what Native Client can achieve, but the advantage of Mozilla’s approach is that the JavaScript code that the system generates will run in any browser – just slower than on Firefox.
Suddenly all this backend stuff is hot. Who would have ever thought that data pipes and the spigots would get so much attention? Salesforce is getting into the game and now so is Rackspace with the launch of its mobile push. Rackspace does not call its new offering backend as a service (BaaS). Instead they call it a “mobile-ready” stack that pre-packages the backend for developer so they do not have to reinvent the wheel every time they start a mobile project.
Rackspace CTO John Engates wrote in an email that the company is packaging its expertise and experience to cut deployment time from days to minutes. The goal is to let developers focus on building the frontend of the apps, like user experience, while Rackspace deploys and runs the backend for them. Engates said Rackspace has also built in its own reference architectures that developers can use to optimize the development process. Engates said the first stack is for a LAMP PHP based deployment with MySQL, Varnish acceleration service, memcache and other components that optimize for a mobile backend deployment.
All of this brings me back to a conversation online about the meaning of BaaS and the connections to platform as a service (PaaS). Whatever you want to call it, the whole backend is getting abstracted and the big players want in.
And that’s just fine for the likes of Kinvey, a BaaS provider that has specialized in providing its own service, which serves as a mobile SDK so developers can connect to systems of record and pull that data into their apps.
CEO Sravish Sridar seized the moment this morning and added Salesforce to its “map,” so to speak.
Make no mistake, despite the conspicuous absence of arguably the hottest buzzword in cloud service, BaaS is exactly the category Salesforce just entered – and for that reason, we’ve added them to our category-defining “Subway Map” graphic.
The map has 29 vendors. Sridar writes that one analyst says there are as many as 47 vendors in the space. And here’s why. Sridar accurately and insightfully draws the correlation that the CIO sees mobile as the way to transition workloads to the cloud. That really sums it up. Mobile is the path to the cloud and companies like Rackspace and Salesforce want to automate as much as they can for developers, the ones who are building out the app-centric enterprise.
It’s sure been an interesting week for this of us who cover browsers. Last weekend, we heard that Internet Explorer 11 will probably support WebGL and SPDY. Then, on Tuesday I got an email from Mozilla, asking if I had time to get on the phone with Mozilla’s CTO Brendan Eich to talk about the organization’s next generation browser engine Servo and the Rust language it is written in. Turns out, Mozilla Research wasn’t just going to work on this alone, but managed to get Samsung to help out with bringing this new engine that’s optimized for multicore and heterogeneous computing architectures to Android and ARM. Given that Mozilla had remained relatively quiet about Servo until now, it was a bit of a surprise that it was now ready to put it into the spotlight.
Interestingly, I had heard from Google’s Chrome team earlier in the week and they, too, wanted to talk on Tuesday. Oddly, the PR team was unusually wary about giving me any details before the call (often they’ll give you some background to make sure you can prepare for the call). When Linus Upson, Google’s VP of Engineering, and Alex Komoroske, the product Manager for the Open Web Platform team, told me that Google was going to fork WebKit and launch its own rendering engine Blink based on WebKit, I had to backtrack a few times to make sure I had really heard this right. Outside of a few WebKit insiders, few people expected Google to do this. WebKit is generally seen as a major success and given its dominance on the desktop (thanks to Chrome) and mobile (thanks to Safari), few would call it anything but a massive success.
It’s Okay To Blink
Google’s WebKit is fork is obviously the most controversial of last week’s announcements and leaks. Google says the reason for its fork is essentially technical, but given that WebKit was, at this point, one of the few projects where Apple and Google were working together relatively closely, it’s hard not to think that there were some political motivations behind this, too, especially given that Google was quite a bit more active in sharing its code back to the project than anybody else.
We’ll still have to see what the ramifications of this move are, but I’m pretty optimistic. Sure, it means web developers will have to test their code against yet another rendering engine, but I do believe the folks on the Chrome team when they say that this was “not an easy decision” for them. Google is obsessed with speed and with WebKit being part of so many other browsers, the development just didn’t move fast enough for Google.
The Chrome team believes that “having multiple rendering engines—similar to having multiple browsers—will spur innovation and over time improve the health of the entire open web ecosystem.” I agree. The sudden arrival of Chrome on the scene in 2008 sure pushed every other browser vendor to innovate faster. At the time, the competition was mostly about who could develop the fastest JavaScript engine (all the WebKit-based browsers use different JavaScript engines). Now, with Blink and Servo out in the open, we may see a similar arms race for rendering engines, too, and at the end of the day, that can only benefit users and developers who can look forward to faster, more stable browsers.
Cisco’s Lew Tucker stood onstage today at Cloud Connect and pitched the networking giant’s “Internet of Everything,” an app-centric world that will be worth $14.5 trillion over the next couple of years. Whereas the Internet of Things is all the objects in our world, Tucker says the IoE is the smart grids and, really, the entire supply chain and its transformation.
Big enterprise companies are good at this kind of thing. They talk about huge market opportunities and great futures with tremendous upside, but it’s a question of how nimble they can be with startups innovating so fast. Tucker, however, gets credit for explaining how an app-centric world ties in with software-defined networking (SDN) and the switch from traditional, heavyweight systems of records (ERP, CRM) to systems of engagement (apps, lightweight services that provide feedback loops).
Tucker, citing Cisco’s own study, says there is $4.9 trillion in immediate opportunity through the development of such things as smart grids, smart factories, smart buildings and smart cities.
Here is the key and why Tucker is correct in what he has to say. We live in a futuristic, app-centric world that has to operate in an old universe filled with increasingly outdated “physical” routers and switches. That’s why SDN is so important and why Cisco faces challenges of its own with new players such as Big Switch and VMware, which acquired Nicira last year. With SDN, the network goes to the application instead of the other way around. The compute, the storage and the network are continually optimized based upon the demands of the customer.
His statements mirror what you hear from someone like Amazon Web Services CTO Werner Vogels who talks about cost-aware architectures that enable companies to manage IT and app workloads according to its own feedback loops. Data goes in to the data center and is read by a variety of apps and services. Analysis comes back and the customer adapts accordingly using standard, reserved or spot instances that it trades in an exchange.
It also further validates the rise of any number of IoT startups. For example, SmartThings has raised $3 million to develop a platform for connecting everyday objects to the Internet. According to our Natasha Lomas, London and Cambridge, U.K.-based accelerator Springboard is launching a dedicated program for hardware startups, focusing on the IoT.
Tucker describes a virtuous circle for the IoE that is all about this app-centric world. The Internet of Things means more data and that requires a cloud infrastructure. Apps are multiplying and will continue to increase as we plant more sensors with APIs into our personal items and everything else: roads, bridges, etc.
The IoE also provides a context for the ways we interact with this deep fabric of connected things. An ERP system will become less relevant for companies. Instead, systems of engagement will put us right in the center of a feedback loop that allows us to measure our own selves and in the process connect to all the other smart aspects of our life. That might be in the city of San Francisco when trying to find a parking spot or the smart factory where we order our data-generated personal things.
“What really matters in the end are the apps,” Tucker said. “They drive a change in what we need in our infrastructure.”
I could not agree more.
Tealium has raised $15.6M in a series C financing round led by Tenaya Capital. Battery Ventures and Presidio Ventures also participated in the round for the for its tag management company that captures clean data streams for enterprise marketers doing analytics on web sites, mobile sites and mobile apps. Tealium has raised a total of $27.2 million.
Tealium’s single universal tag lets companies manage all their vendor tags, including analytics, advertising, affiliate, PPC search and other online marketing. The system lets marketers manage their vendor tag deployments using a drag-and-drop interface. Customers can add new tags, remove or modify existing tags without requiring additional IT resources.
The company will use the funds to accelerate sales, expand its mobile platform and extend its clean data cloud service. In mobile, Tealium will provide support for different mobile architectures and operating systems. Tealium DataCloud will extend to capture more data from email marketers and other vendors. The cloud service collects the data, segments it and makes it available for export that marketers use to do analysis. Lunsford said Tealium has gone from 5 to 250 customers in the past two years, showing how the universal tag is becoming the most valuable piece of data on the web site.
Tealium’s team have a history in data tracking that dates back to the late 1990s when they worked with Webside Story, developing the first JavaScript tags to make it easier to look at website data as opposed to the standard log files.
The practice became popular and soon companies had an assortment of tags that became a giant hairball to untangle. Tag management has emerged as a way to codify the JavaScript tags, making it easier for companies to manage the data.
Tealium CEO Jeff Lunsford said the whole industry has reached a pain point that tag management helps solve, providing a clean source of data from a company’s various vendors.
The tag management market is still relatively young. Lunsford compared it to the web analytics market in 2002 which now is a $1.3 billion market. The vendors in the tag management market have a combined revenue of $70 to $80 million and the market is growing faster than the analytics market did in its early days.
