A month ago at Mobile World Congress, Microsoft announced that the Windows Phone 7 copy and paste update, known as “NoDo“, would ship some time in the first half of March. That’s not going to happen now, as the company has formally announced in a blog post that the update is delayed until the second half of the month.
The reason given is so that Microsoft can take a little extra time to ensure that the update “meets [Microsoft’s] standards, your standards, and the standards of [Microsoft’s] partners.” The statement assured users that the updates scheduled for later this year won’t be impacted by this delay, and that multitasking, Twitter integration, and an Internet Explorer 9-powered browser are all going to ship as expected.
The announcement also outlined the way that updates are signed off, first by the hardware companies, then by the carriers. This is a complex process: Microsoft’s update delivery channel is used not only for software updates for Windows Phone 7 itself, but also for hardware-specific updates such as new drivers or, I believe, the low-level firmware used to control the cellular radio. This is a complexity that a company like Apple doesn’t have—for Apple, all the hardware is its own. On top of that, Microsoft has, regrettably, given carriers the ability to block updates.
It’s good to talk
It’s good that Microsoft has finally spoken a little about what’s going on. An official confirmation that there will be delays is better than rumor and innuendo, and the claim that the delays won’t have any impact on future updates is also a little reassuring. But the response to the blog post in both the comments and around the Web has been largely hostile, and it’s not hard to see why. What we wanted to see was some straight-talking. Something taking ownership of the problems, something to provide confidence that the problems had been solved, something to assure us that these issues were isolated—ultimately, something that would show us that Microsoft would deliver on the promise of its platform. What we got instead was PR-laden equivocation.
A big problem is that the announcement wasn’t really telling us anything we didn’t already know. A statement made on an official French blog said that NoDo would not roll out until the second half of March. The new blog post just makes clear that this is the case across the board.
On top of this, it offers no real explanation either for the past problems or for the new delays. This makes it hard to believe that the future is going to be any better. Though Microsoft has fixed the problems experienced by many owners of Samsung handsets for most users, there are still some who are having trouble installing the initial February update unless they completely wipe their phones, and since that update is a prerequisite for NoDo, those troubles are unlikely to disappear.
The partners aren’t the important ones
The major issue, however, is not so much these details as it is a systemic lack of effective communication. Microsoft was slow to act in response to the initial Samsung problems, and the news about the delay was similarly circulating for several days before this official confirmation. Microsoft is still treating the OEMs and carriers as if they were its customers. While they’re the ones Microsoft is dealing with directly, treating them as the customer will kill the platform. The early-adopting end users—the people who actually bought the platform, are buying the applications, and are encouraging their friends and families to follow suit—have to be treated as king. But they’re not, and they’re not happy.
Microsoft, instead, is covering for its partners. It covered for Yahoo when a Yahoo bug caused sky-high data usage. And it’s still covering. Instead of useful information such as “which carriers have blocked the February update,” or “which carriers will block NoDo when it eventually arrives,” we get vague comments about “working closely” with the carriers. Even though we already know that some carriers can actually treat their customers with utter contempt and block updates, Microsoft still skirts around actually admitting it—a fine example of PR standing in the way of actually relating to the public. Joe Belfiore already acknowledged that they could, so there’s really no need to dance around the issue.
Obviously this is information that might be considered “sensitive,” but putting the carriers’ needs over the users’ needs serves only to alienate the users, and it’s Microsoft who looks bad as a result.
Nor does the post do anything to actually explain why the updates are taking so long to deliver. It talks about the OEM involvement and the carrier involvement. But none of it really makes sense.
The OEM issue is certainly the more sticky one. The fact that OEM-specific updates can be pushed out means that there is necessarily communication and testing between the manufacturers and Microsoft. This is a complexity no other phone platform has, and while it will be good in the long run, it’s somewhat understandable that there are issues in the short term. However, the blog post doesn’t paint a very pretty picture of the process. It may just be an oversimplification, but it sounds like Microsoft basically has to take it on trust that the OEMs have done their job properly. As is obvious from the Samsung issue, that trust isn’t earned.
In its previous post about the February update, the company proudly claimed that the update worked for 90 percent of users, with a further 5 percent failing due to trivially diagnosable conditions (lack of free space) that the software didn’t actually bother to warn about, preferring instead to try to install and then fail. If a Windows Service Pack only worked for 90 percent of users, leaving 5 percent perplexed with useless errors and another 5 percent with some mystery malady that risked leaving their machines useless, it would be a dismal failure.
The PC space is far more diverse than any cellphone ecosystem, and yet Microsoft makes patching that a routine and highly successful experience. Not flawless, but not far off. That it can’t do so for nine different phone models, each using damn near the same hardware, is astonishing.
What would be good to see is an acknowledgement that this process broke down. Instead, we we’re told that laboratory testing “can simulate—but never quite equal—the experience of delivering software to thousands of real phones. So basically, we’re getting a shrug of the shoulders, and it’s probably going to happen again.
The carriers can only ever make things worse
As for carriers, the blog post also made clear that carrier involvement was a part of any phone’s testing. This is probably true, but it’s completely missing the point in a couple of ways.
First and foremost, carriers really don’t need to get in the way here. The testing they do just can’t be all that important or significant. How do we know this? Because they’ll let you stick any phone on their network (at least in the world of GSM networks). Even in the US, with its frequency-based barriers to entry, T-Mobile and AT&T will sell you a SIM card with no phone, for you to slap into any old device and use as you see fit, and this is repeated around the world. Some operators (mostly MVNOs who depend on another network’s hardware) don’t offer any handsets at all; only SIM cards and the expectation that you’ll provide your own handset.
If the carriers can do this safely—and overwhelmingly, they acknowledge that they can—then they can also allow firmware updates through without interfering. If it’s safe to use a firmware update on an unbranded handset on a network, it’s safe to use a firmware update on a carrier-branded but otherwise identical handset. Plainly an untested firmware can’t screw things up too much, or else the only phones they’d allow on their networks would be carefully vetted.
The carrier’s involvement with the firmware should end at branding and perhaps a little configuration: the custom boot screen, the custom highlight color, a handful of other settings that carriers are allowed to touch, and the preinstalled applications. And that’s it. There doesn’t need to be extensive testing (they’re not testing every phone on their networks anyway), and there certainly shouldn’t ever be reason to allow carriers to block or delay updates. It should take mere minutes for a carrier to produce an appropriately branded firmware and give it the thumbs up.
And if they really can’t do that? Microsoft (technically, if not contractually) can put out minimally branded firmwares of its own. That’ll lose the preinstalled applications, but so what? They’re all downloadable from Marketplace anyway! If the carriers continue to ruin the experience, that’s got to be an option.
And that gives way to the second point: we’ve already seen updating done better, and that’s the standard Microsoft needs to live up to. We’ve already seen the iPhone. Apple already works with dozens of carriers (to whatever extent that it’s necessary to do so—Apple prohibits carrier branding), and yet in the same timeframe since Windows Phone 7’s launch, Apple has managed to push out four releases containing fixes and/or new features (plus a fifth release for CDMA support). And while the number of different Windows Phone 7 models might create more work for Microsoft, they shouldn’t make much difference to carriers, since most carriers only offer one or two models anyway, just as they do with the iPhone.
Plainly, regular updates without carrier blocks are feasible. Not just feasible: they are the standard that smartphone users should be demanding. If Microsoft can offer an iPhone-like update experience with an Android-like selection of handsets, that is undoubtedly a good thing. It is one of the promises of the entire platform. A high-quality update experience is one of the major things that can set this platform apart from Android, just as the delightful user interface already does.
So instead of straight talking or useful facts, we’re left with a delay to NoDo so that the company can “learn all [it] can from the February update,” and, uh, basically nothing to reassure us that it won’t happen again.
Microsoft knows that regular updates are valuable. As the blog post says in its unique PR-laden way, “Delivering regular updates to your phone is a key part of our innovation plans.” And we also have quite a bit of evidence that copy and paste has been done for months. A build with copy and paste was shipped to developer handsets on December 5th last year. That build was number 7338. New and presumed-to-be-final builds that have leaked for certain HTC handsets have build numbers of 7339. The emulator that developers use to test their software on NoDo also has build 7339.
That strongly suggests that not too much has happened since the 7338 test build; that even then, the feature was all but done. The company also described it as the “January update” in a webpage (though the original page has long since disappeared into the memory hole, and now no longer mentions January), and even before Windows Phone 7 launched described it as coming in “early 2011”; the tail-end of March hardly qualifies as such. So we’re looking at a delay of more than three months.
All this trouble, and for something that at the end of the day isn’t even a major update—just copy and paste, some performance tweaks, and some small improvements to Marketplace. And yet Microsoft and its partners are struggling to deliver that much.
One has to wonder what’s going to happen if an actual important update needs to ship. A gaping security hole in the browser, for example, that allows rooting the phone and malicious access to personal data. Something where a three-month turnaround is simply not acceptable. Are carriers and OEMs still going to be able to drag their heels and screw up the updating process? Or will this finally give them the spur they so badly need to swing into action?
Not fatal… yet
It’s still early days in the life of Windows Phone 7, and none of these update troubles are in themselves fatal. But the communication has been abysmal. It has been consistently reactive, patronizing, and so laden with PR wording that it tells you next to nothing anyway. It’s better than silence, but only just, and it’s no wonder that the community feels short-changed by it.
Ideally Microsoft would be releasing updates in a timely manner, every month or two, adding new features incrementally and justifying the commitment made by the early adopters. Let the updates flow as fast as they can be developed. We know it’s possible.
But if that doesn’t happen, at least have the courtesy to tell people why. Tell people that the update is going to be delayed beforehand. Don’t just keep silent and claim, “Well we never actually meant it to go out in January, that was all a misunderstanding.” The userbase isn’t that stupid. Tell people which carriers are doing what. We know carriers can block things, so don’t fob us off with vague claims that we’ll get the updates eventually—make sure that if someone is denied an update then they know who to blame.
The current strategy looks bad. It is causing a loss of confidence among early adopters, and that hurts. These are the people who will fall in love with the platform and who will do the real job of selling it to their peers. They have a reach and influence that no marketing budget can buy, and if they feel that the platform is lost, and that Microsoft can’t deliver on its promises, and if they no longer have the faith that it will develop into the platform that it deserves to be, they will defect to other operating systems, taking their influence with them. Both the customers and Windows Phone deserve better.