For some, the definition of software freedom begins and ends with source code. Such people have apparently never heard of market competition.
It's arguably even more important, and doesn't necessarily derive from a software license (though it's no doubt better when protected by an open-source license).
Over the last few weeks, we've seen signs that key open-source vendors are waking up to this fact, with both Canonical (Ubuntu) and Mozilla (Firefox) sniffing around Yahoo/Microsoft to replace Google with Yahoo/Bing as their default search engine.
Choice, you see, is good, even when it's not created by an open-source license.
In the case of Google, its search technology is 100 percent proprietary. Yes, the open-source world prefers Google, but letting that preference turn into a fixation is both bad business and bad policy.
It creates too much reliance on one vendor. That's not good for software freedom, and certainly not for market freedom, no matter how "not evil" we might presume Google to be.
Apple, the uber-proprietary technology giant, gets this. That's why it, too, is looking to lock Google out of its iPhone, as BusinessWeek has reported. The more Apple competes with Google, the better Microsoft looks as Apple's "pawn" to keep Google's search dominance at bay.
It's not about loving Microsoft. It's about preserving options...and competition.
With ZDNet, I applaud Google's efforts to take on Apple in key markets like smartphones where Apple increasingly dominates. I don't want to serve an Apple monopoly any more than I liked living under a Microsoft monopoly.
Apple, even more than Microsoft before it, is committed to a proprietary path, one that threatens to stifle outside innovation and becomes more proprietary with every product release.
But I also don't want a Google monopoly. While I deeply respect Google's open-source credentials, it is not necessarily always the model open-source citizen, as its experience with Android suggests.
In a perfect world, open source would create the "magic" that wins over consumers and fuels new markets.
Until we arrive at that perfect world, however, there's good old-fashioned competition, which sometimes, indeed often, will have nothing to do with source-code licensing and everything to do with market competition.
As arguably the world's largest open-source company, Google has a big stake in maintaining its place at the heart of the open-source ecosystem. Recent events, however, suggest that Google can't rest on its laurels if it wants to secure the hearts and minds of open-source developers.
Make no mistake: Google needs those developers. Android, Chrome (and Chrome OS), and other Google initiatives depend upon fostering vibrant open-source communities that can help it to surpass Microsoft and Apple.
Does Google need to search for new friends?
There have been rumblings that Mozilla would look to Google alternatives for the default search engine within Firefox, despite Mozilla pulling in 91 percent of its revenues from its Google partnership. Mozilla employee Asa Dotzler, though not speaking for the foundation, says that he'd welcome a switch from Google given its rising dominance over the Web.
Mozilla executive Mitchell Baker, for her part, has noted that alternatives (Yahoo, Microsoft) would likely pay Mozilla more money and give Firefox shelter from Google, which has been building a rival browser, Chrome.
If Mozilla needs a nudge, Canonical just gave it one, defaulting to Yahoo search for future versions of its popular Ubuntu Linux operating system.
Mozilla and Canonical represent the heart of the open-source community. If they move from Google, it paves the way for other communities to do so, too.
Mozilla and Canonical arguably have their financial self-interest at heart in such deals, but Gartner Research Vice President Brian Prentice points out an even bigger issue that could drive a wedge between Google and the wider open-source community:
Patents.
Google, after all, was recently granted a patent on its MapReduce parallel programming model. This puts the Apache Software Foundation Hadoop project firmly in its sights, even if Google likely has no intention to sue. It's the uncertainty that may end up hurting the open-source community.
Prentice queries whether
So, does that mean that it's only a matter of time before Google's legal team starts sending out letters seeking license fees? I don't know. And that's the point. No one else does either.
I would suggest to you all that the greatest threat to any open system is, in fact, uncertainty. And what Google has done here is to dump a whole lot of uncertainty onto the market.
To those asking Google to submit such patents to a commons to ensure they are only used for defensive purposes, Google's response, as Prentice goes on, effectively amounts to "trust us - after all we're not evil."
Don't get me wrong: Google is and will likely remain a very strong proponent of open-source software, with projects like Chrome that impress and push the boundaries of innovation.
But in its rush to serve a wide variety of customer needs, it may end up overlooking or stepping on its erstwhile partners (like Mozilla), and could pursue policies (like its MapReduce patent or even its H.264 video codec stance) that threaten the open-source community just as much as Microsoft has.
Google is powerful, but it still needs friends. Open source has been a chief ally of Google to date. Will it remain such? That's an open question.
Never have developers mattered more. As Apple readies its tablet announcement party for later Wednesday morning, the big question remaining is whether developers will join, or whether they'll join Google's Android and Chrome initiatives.
It's Apple vs. Google. And it's all about developers.
Microsoft has ruled the software market for decades because it won the hearts and minds of developers. But Microsoft has been slipping lately, and Apple has eagerly taken its place. Despite some early fits and starts, Apple has consistently given developers ways to make piles of cash, even as a global marketplace for mobile applications has expanded.
Developers have returned the favor by making mountains of applications. Consumers, in turn, buy into this rich application ecosystem. And so the virtuous circle feeds itself.
Or will, unless Google can intervene. Google isn't reported to be building a tablet. At least, not yet.
But that's largely because Google's strategy is to allow its hardware partners to build Android-based devices like tablets, Netbooks, smartphones, etc. CES was filled with them.
The strategy seems to be working, with IDC projecting Android to grow by a 150.4-percent compound annual growth rate through 2013, reaching 68 million units shipped that year.
Importantly, Google's strategy is very different from Apple's. Apple has set up an excellent, but very closed, developer sandbox where Apple retains complete control. Google's strategy depends heavily on open source, giving developers freedom, flexibility, and a paycheck at the end of the process.
Apple, however, could spoil this strategy. As Gartner's Brian Prentice reasons, Apple may be changing the rules of developer engagement just as Google starts to close the gap:
Rather than focus strictly on the addressable market for developers what about the addressable market of developers....
[H]earkening back to the 80s again, what if Apple, either today or some point in the future, is able to resurface HyperCard. Not 1980's HyperCard. But a new millennium iteration of it.
Doing so would massively increase the number of people that can create mobile applications and, by extension, create a new market for the more traditional developers to create widgets and components for this new class of developer - all available at the iTunes Store.
If Apple can pull something like this off [then] there is the potential for them to bind an enormous new army of next-generation consumer-developers to their proprietary platform. And that would be an outcome advocates of open source should ponder over long and hard.
In other words, what if Apple moves the goalposts? Apple could pull a Microsoft, making development much easier, while keeping the resulting applications safely entrenched in iTunes/the App Store.
Apple's advantage with the tablet is that developers already know how to write iPhone applications. It could be that writing specialized applications for the tablet won't be worth the bother, but it's likely that Apple learned enough with its iPhone experience to ensure developers will get on board the tablet train, too.
All of which could spoil Google's developer party, and end up crimping Android's growth just as it was set to explode. How will Google respond? How would you?
Despite booming enterprise server sales, some in the industry continue to grumble that Linux is too hard. Designed by geeks for geeks, the theory goes, Linux will never be mainstream.
Reality hasn't been kind to such arguments.
Consider the fact that Linux-based Google Android saw 350 percent growth in 2009, according to Myxer data. I've yet to hear anyone talking about Android being hard to use. My teenage neighbors bought their Android phones and have had little trouble texting, browsing the Web, and installing applications.
It's Linux. It's not hard.
Where Linux does sometimes fall down has little to do with Linux, and everything to do with the focus application developers place on Linux.
Some report having trouble installing applications on Linux, like the cutting-edge Firefox release. While some Linux defenders suggest that an easy way around this is simply to wait until official support/packaging arrives within Ubuntu/SUSE/Fedora/etc., that won't do for those of us who sometimes want to veer off-piste and use beta software.
Application developers make sure their beta software installs easily on Windows and Mac OS X. They don't pay as much attention to Linux.
The same is true for hardware vendors. Linux works very well when pre-installed on Dell hardware: just as well as Windows does. The difficulty is in getting the hardware vendors to pre-install Linux with all necessary drivers.
This, however, is an opportunity for the Linux distributions. The distribution with the best application and hardware support will win.
Red Hat demonstrated this in the enterprise server market, and is approaching $1 billion in revenues for its trouble. Canonical, the company behind Ubuntu, is taking the initiative to determine for which third-party applications the Linux community wants easy installation.
Google, of course, has already done this with its Android platform. No Android user is ever going to have to learn about "sudo" or "apt-get install" or any other elements of the Linux command line.
That's how it should be. At least, for the mainstream.
Google also promises to be good for "the Linux desktop" by removing the need to worry about the operating system at all, instead focusing on Web applications. I've already had a taste of that this past week as I've been installing Ubuntu Linux on different hardware, running a range of Web applications as easily as I do on Mac OS X.
We're rapidly approaching this future when the OS takes a backseat to the applications that run on it. It's going to be an awesome time for consumers, as Microsoft, Apple, Google, and others duke it out for supremacy on the proliferating range of hardware devices that consumers and enterprises use: desktops, laptops, smartbooks, Netbooks, phones, etc.
Linux has a starring role to play in such a future. The company that makes Linux easiest will win.
Few will have noticed that Amazon.com is getting into the developer game for its Kindle e-reader.
It doesn't help that Amazon is launching its third-party developer program in the midst of Apple's tablet hoopla, but that's not the core problem.
The big problem is that Amazon's program offers developers less than Apple's equivalent. A lot less. And it's way too late, as ZDNet's Jason Perlow opines.
We are presently inundated with application stores. Every device seems to come with its own app store these days--each an island within the larger ocean of mobile computing.
Amazon might have been able to stand out when it first introduced the Kindle, but today there are application stores for everything from iPhones to Nokia phones to BlackBerrys to Intel-based Netbooks to...you get the picture.
To stand out, the Kindle developer program needs to offer developers more, much more, than they'll get with competing application development platforms and associated application stores. In this Amazon strikes out. Big time.
As reported by The New York Times, the Kindle program requires developers to pay 15 cents per megabyte for wireless delivery of their applications to end-users.
This requirement doesn't apply to free applications, as noted on Amazon's Kindle developer page, but it could be a healthy chunk of a 99-cent application, when Amazon is then going to take another 30 percent of the net on the deal.
And will Amazon stick to this arrangement? The vast majority of applications (80 percent) distributed for the iPhone cost users $0, which means that Amazon is going to be paying for a lot of wireless distribution.
It can afford to do so now, but what about when developers get paid for in-application advertising, rather than the application itself?
Gartner predicts that by 2013, 25 percent of app store sales will come from advertising on free applications. Will Amazon tweak its program to take a fee on the advertisements? Or will it push the 15-cent delivery charge onto these free applications, anticipating that the developers will get paid?
In this scenario, it might be that developers will underwrite delivery charges to get to downstream advertising revenue, but that's a risk many developers likely can't afford to take.
Amazon might be able to overcome this hurdle it offered a significantly more open development platform (and approval process) than Apple, but it doesn't appear to be doing so. As Fast Company notes:
The [Apple] App Store's true rival isn't a competing app marketplace. Rather, it's the open, developer-friendly Web.
To the extent that Amazon can open the Kindle to that developer ecosystem and shoulder the costs of application delivery, it could have a winner on its hands. But so long as Amazon simply apes Apple's somewhat closed approach, while potentially burdening developers with the cost of reaching their audience, it will fail.
Mail that only a mother could love
(Credit: Apple)Most Mac users default to Apple's good-but-not-great Mail program to manage their e-mail. But given Mail's serious deficiencies (e.g., weak to nonexistent integration with calendaring and contacts), a variety of open-source initiatives are underway to bring a full-featured e-mail client to the Mac.
That's the good news.
The bad news is that there appears to be little to no coordination between the different groups, leading to duplication and fragmentation of efforts.
Zimbra, Mozilla Thunderbird, Mozilla Raindrop, and a new project initiated by NetNewsWire founder Brent Simmons, Letters.app, all seem to be heading in the right direction. But they're taking different roads and seem intent on reinventing the e-mail wheel.
Why not band together?
There seem to be a number of good reasons. Zimbra, for its part, is geared toward the enterprise messaging market. It's also licensed in such a way that it's unlikely to be palatable to competitive efforts, given that it requires developers to include Zimbra logos and other trademarks in the UI of the derivative work.
As for Mozilla's Thunderbird, while its browser cousin, Firefox, continues to expand its market share due, in part, to its active and enthusiastic developer community, Thunderbird still has a nascent community. So does Raindrop, Thunderbird's younger and nimbler sibling.
With a still-small community, there's less reason to accept the other trade-offs that come with building on the Thunderbird or Raindrop source code.
Those trade-offs may well be significant.
For example, Mark Mason, a Thunderbird user and Letters.app developer, argues that "it would be much more work to fix up Thunderbird than start from scratch."
There's also the potential problem, pointed out to me by Dj Walker-Morgan, that Thunderbird is not a native Cocoa (Mac) application and doesn't integrate seamlessly with the Mac OS X interface," an issue acknowledged by Mozilla's messaging head, David Ascher ("[Letters.app's] primary focus on OS X is a real difference").
I say "potential problem" because Thunderbird's cross-platform approach to e-mail (i.e., it works equally well on Mac OS X, Windows, or Linux) is also a great strength.
After all, most of us don't have the luxury of spending 100-percent of our time on just one platform, be it Mac, Windows, or Linux (or any of the myriad of mobile devices out there.)
Such thinking doesn't sway the Letters.app crowd, which has the stated goal of creating "a lean and programmable IMAP e-mail client, with plug-in and automation APIs, designed for developers and power users." Given this goal, building from cross-platform Zimbra or Thunderbird simply may not work.
Even so, Ascher remains "hopeful that there'll be room for collaboration in places, and cross-pollination." I do, too. The Letters.app community is energetic and has attracted some significant names from the Mac community, including John Gruber and Gus Mueller. These would be great people to help build a next-generation e-mail client.
The Mac definitely needs one. And the Letters.app community could well be the ones to build it. But it would be ideal to combine the talents of it with other open-source e-mail communities working toward that goal, rather than fragmenting their efforts.
This might not mean a sharing of source code, but perhaps it could mean a sharing of ideas, techniques, and even a common standard for add-on extensions?
Martin Luther King Jr. dreamed of a day when we would "transform the jangling discords of our nation into a beautiful symphony of brotherhood." In the technology world, perhaps that notion might be applicable to the dream of a day when we'd get beyond incompatible operating systems to have a truly interoperable software application industry.
Slowly, but surely, we're getting there...on both counts.
Sorry, IE: Windows just isn't enough anymore
The need for heterogeneity has always existed, but it's becoming more pronounced as devices proliferate, devices that run diverse operating systems, especially Linux.
Last night I fired up my Ubuntu-based Netbook. It would have been an uncomfortable experience, given that I'm used to working on my Mac, except that several of my favorite applications were already there: Firefox, Zimbra, and Thunderbird. Between those three, I had access to all the other applications (Facebook, Twitter, RSS feeds, etc.) that I use on my Mac.
You may not have a device running Ubuntu/Linux today, but tomorrow it's very likely that you're going to have to get out of your Windows/Mac comfort zone. Perhaps it's that new Android-based phone you've been eying. Or maybe that Netbook AT&T and Verizon want to give you.
Somewhere, at some point, you're going to have to confront a new operating system. When you do, you're going to appreciate the comfort of cross-platform applications (a fact not lost on the Ubuntu folks, who are trying to ensure popular but proprietary applications run on the open-source operating system).
Firefox is likely to be the Big Kahuna among them, as it's the application that launches you into the Web, that wonderful world that (mostly) doesn't care whether you're a Mac/Windows/Linux person, and just delivers great applications.
But despite Google's HTML5 protestations to the contrary, we don't live in a Web-only world just yet, and probably won't for a long time. We're going to need client-side applications for the foreseeable future (remember Apple's ill-fated attempt to relegate us to browser-only iPhone applications?), and the winning applications are going to be those grown-ups that span multiple operating systems.
Sure, it's more work for Mozilla to ensure Firefox works on Mac OS X, Windows, and Linux, in addition to a growing array of mobile operating systems, but it's that work that keeps many of us devoted to Firefox.
It's no longer sufficient to fixate on one platform, no matter how popular. Hardware devices are quickly expanding, running less and less Windows and Mac, and running a wider variety of Linux distributions.
In addition, the Web is changing expectations. Just as we expect CNN.com to work, no matter our underlying OS, so, too, will we come to expect our applications to follow us, no matter our choice of platform.
Zimbra gets this. Mozilla and Google do, too. Do you?
Facebook adds Apache to its gold friends list.
Web properties used to treat open source as a resource to be strip-mined. Increasingly, however, successful Web companies like Google and Facebook are giving back, helping to replenish the open-source ecosystem from which they derive so much value.
Facebook and Apacheannounced on Tuesday that it is becoming a "gold" sponsor of the Apache Software Foundation, expanding its open-source investments from code to cash.
Facebook arguably couldn't exist without open source. As David Recordon, Facebook's senior open programs manager, told me by phone on Monday and writes on Tuesday, Facebook owes a lot (and gives a lot) to open source:
From the day Mark Zuckerberg started building Facebook in his Harvard dorm room in 2004, the [Facebook] site has been built on common open source software such as Linux, Apache, memcached, MySQL, and PHP. In that time, we've open sourced more than 20 different technologies, and scaled Facebook to reach over 350 million people around the world....
[I]t's not possible to scale a site like Facebook simply by sharding your databases, but rather takes a combination of specialized technologies. Open source allows us to not just to make technologies like memcached scale beyond its original intent, but to release technologies like Thrift for others to build upon as well.
Take Hadoop, an Apache Software Foundation technology used and initially developed by Yahoo, Facebook, and others to scale their operations. Facebook uses and contributes to Hadoop, but also initiated an Apache Software Foundation-hosted subproject called Hive, a "petabyte scale data warehouse" technology.
Facebook joins Hewlett-Packard in sponsoring the foundation at the gold level. (Google, Microsoft, and Yahoo are the "platinum" sponsors.) The cost of gold sponsorship is $40,000 per year and provides little tangible benefit to the sponsoring organization beyond a logo on the foundation's Web site and a quotation for a press release.
If that seems skimpy, sponsors like Facebook aren't generally looking to advertise or get other quid pro quo benefits through sponsorship.
Among other reasons cited by Recordon, Facebook is contributing both code and cash to Apache because the foundation provides a forum in which Facebook is able to engage different ideas on code direction and development. It's also a way of feeding an ecosystem that has brought so much value to Facebook.
Indeed, Recordon noted that it's sometimes hard to know in advance how Facebook or the wider open-source community will benefit from its contributions. He cited the Cassandra project, similar to Google's Bigtable technology, which Facebook seeded but which Rackspace and Twitter have advanced.
Intriguingly, Facebook, like Google, increasingly defaults to the Apache Software License.
Given that Facebook and other Web companies aren't subject to the requirements of traditional open-source licensing (i.e., since they don't actually distribute software they are not required to contribute back), one would assume they would either a) not contribute back at all or b) contribute under a variety of licenses, given that any of the traditional open-source licenses is effectively a blanket license to use the software in a hosted environment however one wishes.
Facebook prefers Apache, according to Recordon, because it encourages frictionless adoption, a theme I've covered before. According to Recordon:
Apache is our preferred license. We really like Apache because it's easy to understand (non-lawyers can understand it) and includes clear patent rights language. Plus, Apache's contributor license agreement works nicely with it. We'd like the contributions we make to be used as widely as possible, which is more likely to happen with a license that is easy to use and understand.
I've suggested that theGPL is the new BSD on the Web, but apparently Facebook and its peers think that Apache/BSD is the new Apache/BSD.
Regardless, Facebook's gold sponsorship of the Apache Software Foundation is a coup for open source, and a signal to its peers: it's not enough to simply build on open-source software. You also have to give back.
Microsoft is far from dead, but it's hemorrhaging on all sides, and particularly in markets closest to consumers like mobile where it is steadily losing market share.
As one example, though a potent one for me, a longtime friend and Microsoft employee wrote on Facebook that he had finally capitulated and bought an iPhone.
This is a man who dutifully stuck to Windows Mobile while the rest of the world fled. He's a man who resolutely continues to promote Microsoft for many good reasons..
He's now gone to the "dark side." Or the cool side, as the case may be.
Microsoft can hardly afford to lose its marketing mojo with mainstream enterprises and consumers (and boy, has it lost some with these ads), but when its own employees turn on it...? It's not looking good.
To stanch the bleeding and rediscover its glory years, Microsoft should take a page from Apple, a company that nearly imploded years ago only to rediscover its design and financial prowess.
Yes, from Apple.
The reason that the world gave Apple another lease on life is that it created lust-worthy products. The desire has become so intense that the markets starts to salivate, Pavlovian style, at the merest hint of Apple creating a tablet...or trash can. We will apparently buy whatever Apple sells, sight unseen.
Sure, Apple still has anemic market share compared with Microsoft. And, yes, Apple serves different markets--content with consumers while Microsoft owns the enterprise.
But this is cause for concern, not complacency.
The enterprise is increasingly becoming personal. Employees are being given more leeway about what technology they buy, and Macs and iPhones are increasingly within the general guidelines as to what enterprises will support.
They have to be. Employees demand these Apple products.
Microsoft generally makes technologies that consumers have to use. Apple makes technology that consumers want to use. Over the long haul, which position do you think will win? If you're recruiting the best and brightest, for which company do you think they'd want to work?
This isn't a suggestion that Microsoft should start shipping purple PCs. But it is a call for Microsoft to make more software that consumers want.
Yes, even within its core market of enterprise software. Enterprise software can be sexy. Well, sort of.
Look at Atlassian. It does phenomenally well by creating software like Jira that developers want to use. For that matter, Microsoft technology like SharePoint inspires the same reaction: it's not sexy in the Apple sense, but dramatically more usable than its curmudgeonly competitors like Documentum.
SharePoint, in a way, is Microsoft emulating Apple, albeit in a stodgy, enterprise sort of way.
Microsoft needs a lot more of it, and it needs to step up its game in mobile and other markets that directly touch consumers. Windows 7 is a good first step in reinventing the Windows brand. The company should go further.
Whether you're an open-source advocate or not, you likely run open-source applications on your laptop or desktop. From Firefox to VLC to Handbrake to Adium, some of the best applications for Mac OS X, Windows, and Linux are open source.
The iPhone, however, is a relative wasteland for open source. Should Apple care?
There's no open-source app for that.
OStatic points to four good open-source applications currently available for the iPhone, including Funambol for sync functionality and WordPress for blogging, but these are the exception to the rule. A Google search turns up others, but few that match the quality of open-source applications available on the "desktop."
Why?
It can't be a question of open-source developers shrinking from the proprietary iPhone platform. Much of the best open-source application development already happens on proprietary platforms like Mac OS X and Windows.
It's also not a question of understanding: O'Reilly and others walk open-source developers through iPhone application development.
Even so, it could well be that open-source developers have resisted the closed nature of the iPhone development and distribution process. In this, they wouldn't be alone.
First, you need a Mac to write iPhone applications. While the open-source crowd has embraced the Mac, many developers won't want to leave their Windows or Linux machines. (XMLVM could offer an alternative, as well as enabling developers to write iPhone applications in languages more familiar to them, like Java, instead of Objective C.)
Beyond hardware and programming languages, there's still the burden of running the gauntlet of Apple's application approval process. Open-source development has tended to provide unfettered playgrounds for development like SourceForge and Google Code. iPhone development is anything but unfettered.
Yes, Apple opened the door a crack to open source when it stopped requiring an NDA for development, but the door must be opened wider.
Otherwise it stands to miss out on applications like VLC, the open-source media player (which actually wouldn't likely be approved by Apple due to its competitive nature with Apple's native iPhone media player).
You can get a fork of VLC (called VLC4iPhone), but not through the official Apple App Store, and only if you jailbreak your iPhone. Few mainstream users will choose this route to install applications.
Even fewer open-source developers will go this circuitous route to lead them there, as hinted by noted open-source luminary Doc Searls, which leaves the official App Store largely bereft of open-source applications. This is Apple's loss, and it is a loss to the general public.
Apple's iPhone browser is great, but we'd be better off if mobile Firefox gave it some competition, just as it has on the "desktop." Fring is nice, but I'd much rather have Adium running on my iPhone. And so on.
Open-source development could provide needed competition to the mostly proprietary applications that dominate the App Store. It also could supply fresh innovation.
CNET has reported that iPhone users and Google Android users have much in common in terms of their usage patterns and demographics. Their developer audiences, however, are increasingly different, and that's to Apple's hurt, especially as Android grows in market share.
Android, after all, stands to scoop up a significant swath of mainstream users by attracting both proprietary and open-source application development, while Apple's iPhone predominately serves the proprietary software set. That's the bulk of the market, to be sure, but it stifles the experimentation and innovation that open-source developers could be bringing to Apple's iconic iPhone.
Does Apple care?
Follow me on Twitter @mjasay.
