I wrote the underlying article - thanks for catching that I wasn't using 2007/iPhone as the timeline base.
The 2002 date was when I was saying smart phones reached general consumer availability, not when the very first one was invented. As I mention in the article, it was 2002 when Blackberry phones and Microsoft Pocket PC phones came out, as well as the aforementioned Palm-based Treos.
I wrote the original blog post about Teslas bricking. This IDC/IDG post is embarrassingly wrong about basically everything. The cars are most definitely not "just in need of servicing". Tesla's not even arguing that - I even have a written statement from them that it's $40k to fix.
This alleged "expert" can even get the number of cells right (6,831 by the way).
My takeaway is that the Tesla Roadster is a very expensive toy at this point. There are some electric charging stations now at my work (I can only assume they would work with the Roadster, haven't look into it), but it really sounds to me like the early adopters are going to be paying out the nose for electric for a few more years.
I have a Civic Hyprid, which is pretty underwhelming, but functions pretty much like a regular car. If I don't end up giving it to my step-daughter, I almost assuredly will keep it until the more efficient electric vehicles are as reliable.
You should to post that statement regarding the $40k (with any sensitive info blacked out, of course). Seems like a lot of the back and forth on this issue hinges on whether they are really charging that much to correct this problem.
Motorola Milestone XT720 released June 2011 (USA: August)
Runs 2.1
No bugfixes
No updates
No upgrades
Buggy as hell. 2.2 upgrades and flash support pledged by Motorola sales/support at the time of release in the US -- later clarified that those employees had been "confused" when the mothballing was made official in late November.
Those Apple green bars look quite nice, but I'm not a fan. Maybe add some Windows phone.
The problem as I see it is that since Android is Open Source, the manufacturers don't abstract their innovations into a HAL. For example, Motorola's FM radio and HDMI support is peppered all over inside the core eclair framework. So, someone has to mix it all together again with each new release. Usually in the FOSS space, we like to believe that the effort of maintaining private forks encourages companies to contribute their efforts back to the open source project. That doesn't work here, because manufacturers like Motorola also have the viable option to simply abandon their forks. Which is even less effort.
Apple has a huge economy of scale--only four or five devices (most very similar) with relatively humongous market share per device. The personnel/device ratio obviously supports a much better customer experience than any Android device can offer (with the exception of perhaps the Android developer phones).
I can only assume Microsoft knows what they are doing and enforces abstractions and barriers that limit the scope of manufacture monkeying.
I agree they're maybe not essential to a lot of consumers. Lord knows most people never thought about whether their dumbphone's OS was outdated or not, but I do think it's less than ideal for the ecosystem in the long term.
Yes, it's 7 that were "never" and 5 more that were "weeks", so 12 total that were "weeks or less". I struggled with that phrasing a bit - either way seemed wrong.
As an owner of a iPhone 3G I don't agree with the solid green color that it gets. It was rendered unusable by iOS4.0, and although it was on "the current version" up until the release of iOS5 I don't recall it getting any actual feature from iOS4 other than faster javascript in the browser. So things are not as binary as depicted there.
It also seems like you're looking at Android through an Apple lens where the only phone updates come via the OS and having the latest version is the key metric. This means that e.g. any delay in releasing an iOS update has no impact on your graph, but if the Nexus team bust their chops to get an update out early, it makes every other phone look bad in relation.
What would it look like if you charted when the phones last received an update to their Maps, Youtube, or Gmail client apps from Google?. For Maps they would all be green I believe, less so for Youtube which is currently 2.2+ only so 15% of current devices wouldn't be able to get the latest version.
Glancing through the list of new features in iOS5 it seems to be equally split between a) things Android already had (twitter, notifications, no pc required updates), and b) things that are considered Applications, not part of the OS in Android (e.g. To-do apps, email/calendar improvements, e-magazine reader app) so there doesn't seem to be as much reason for Android users to have to upgrade.
So generally, from this point of view, you seem to be a bit loose with equating "support" from Google to mean new OS version e.g. your phrase "tracked down every update that was released for each device" isn't strictly true if you consider a newer version of Maps or the Market an update. And I would imagine the average person in the street would.
You saved me writing a post. Totally agreed on the features part which matters most for the updates second to security.
The only open issue with Android is security updates. Manufacturers often release minor patch level updates (HTC did recently for their logger fiasco for e.g., My Atrix went 2 minor updates before getting Gingerbread etc.) but that still leaves the phones vulnerable to browser based attacks at least.
With Google making Browser and many other system apps uninstallable/replaceable in ICS may be they can update it like every other app - through market.
With Android 5.0 Google should focus on near complete modularization - all user space apps / libs can be updated by Google independently of the phone just like GMail. This may not be 100% possible but they should do as much as they can practically.
I got my Nexus one around 14 months ago. I choose it because its was the official Google developer's phone and so mostly likely to get updated! Except of course we now know that the Android benchmark device that's less than 2 years old will not be getting any more updates.
BTW - I am willing to bet that when ICS source is released someone works out how to get ICS on it. I.e - I doubt that there's any fundamental reason it can't be done?
I think I bought it about 6 months before the 3GS came out. I no longer use it as a phone, but still use it for other stuff.
I also have just got a Nexus One (and have another Android phone). I'm not particularly disappointed that it's not getting ICS. I've not figured out one actual feature I care about that this announcement will deprive me of yet[1] (and as you say, I'll probably be able to put it on anyway. The key sticking point seems to be limited storage space, which custom ROMs have worked around before). Perhaps it'll mean I get better scores on javascript benchmarks with the default browser, but I'm keen to try Firefox anyway now that I have an ARMv7 device. But if the only thing I get is version number bragging rights, I'm not sure I'll even bother with ICS.
[1] Face unlock? No front facing camera on the Nexus One, Android Beam? No NFC chip. Better browser? I'm not sure how much the benchmarks translate into reality. Built in VPN capatablities? actually that's interesting. Folders? We're scraping the barrel now. Different multitasking UI?
Good points. Also since there will be loads of users stuck on Gingerbread / Froyo the N1 will probably be useful as a development test phone in my case.
I actually originally started doing it that way, but thought it was more relevant vis a vis release date than calendar date. I do intend to include that when I finish doing all of the phones. (Yes, I'm a bit OCD :) ).
Google apparently says less than 50% are on Gingerbread, despite way more than 50% of Android phones having been sold since 2.3 shipped.
I do think manufacturers might have gotten a bit better, but I guess we'll see. That's one of the reasons I want to do the next 6 months or so of phones.
That's every Samsung phone released in the US through mid-2010. I do intend to keep at it and add more. (& yes, Motorola does appear to be particularly bad)
Yeah, the European marketshare is a little different: Motorola releases few of their devices in The Netherlands (more in Germany & UK), but Samsung has a very high market share here.
If you're coming to the chart to see how well a manufacturer supports their products, the current layout is best. If you're trying to understand the history of phone releases and their current state, your suggestion might be best. I prefer the current layout.
It might be interesting to include whether a given phone is running stock Android or OEM skins like Sense. General consensus is that stock phones tend to be upgraded more often and faster, but it'd be nice to have some data to back that up.
Concur. Similarly, general consensus seems to be that the skins actively hold things up because manufacturers have to port them to each new OS on each device.
It'll be interesting to see how much life the skins have left in them in the post-ICS era.
I think you have proven something that many of us have experienced - Android may have a number of technical advantages but its rollout management is unreliable and chaotic and so many non technical users may actually be better off with IOS devices! Technical users may be able to update themselves with some effort (E.g. via XDAdevelopers) if they choose devices carefully and have a bit of luck !
Over the last few months I have been advising my non technically minded friends / family members to "just get an IOS device if you can afford it. Android is great and more flexible but App quality is generally lower and you can't be sure you can EASILY get updates even 6 months after you buy. With IOS you know the device will be supported for around 2 years and maybe even longer. With Android there is no guarantee!"
For me this is sad. I've got both an Android phone (Nexus one) and a tablet (HTC flyer). I am seriously thinking of getting an Iphone for my next personal phone and just accept the few restrictions - BAH !
At the same time, the US is one of the biggest markets around, so it provides a reasonable case study and it tends to be the first to get any upgrade. It would be a nightmare to try to do the same for Europe for example because of the fragmentation.
is the US really the first to get an upgrade? Most of Samsungs phones are launched first elsewhere, and US versions later come with their own carrier specific versions. The plain Galaxy S is not on that chart, and it would have received the upgrade before all the US variants.
It's an apple fanboy piece, presenting heavily skewed data, be it from either straight-out falsity or improper definitions:
1) Original iphone can't run iOS 4, but it gets green all the way through. Similar issues with 3G and iOS5. EDIT: I see now that the way you've done the timelines makes it confusing, and seems to be done intentionally to make your point rather than clarify what's going on.
2) 2.1 to 2.2 is a minor release. 2.1 to 3.0 is a major release. And Android 3.0 isn't a phone release at all - specifically stated as such. It's impossible for a phone to be "3 major releases behind" when there are only two major releases available.
Go back and remake the graph fairly. It will still show the story you want to tell, but you will be able to claim impartiality - and with it, better authority.
I was surprised at the iOS 4 omission, but for the love of god, please don't argue based on the version numbers alone. They never tell the whole story, and it depends on both software maturity and industry whether an upgrade is major or minor. There's also the fact that enough minor upgrades with a feature here and there over time can definitely accumulate into what would otherwise have been a major release.
For example, Android 2.2 added huge performance gains, which definitely makes a difference for what app developers can do and what you can run. It also added very useful new features like wifi hotspot that make the device much more useful.
Meanwhile 2.3 had a streamlined UI theme and added a new keyboard, NFC, gyros, WebM/AAC, multiple cameras, etc. Is that just minor?
As an Android user, I feel the graph is pretty representative of the ecosystem and how my experience compares to iPhone users around me.
For example, Android 2.2 added huge performance gains ... It also added very useful new features like wifi hotspot that make the device much more useful. ... Is that just minor?
Yes, I think so.
Compare to this snippet:
March 9, 2011 – iOS 4.3 arrived, delivering hotspot functionality, faster browsing, lock switch settings and improvements to AirPlay and Home Sharing. [1]
iOS gets to call hotspot and faster browsing a minor version, why not the same for Android? It's comparing apples and oranges. Similarly, the cycle for Android 2.0 to 2.3 was around a year (well... 14 months) which is equivalent to the cycles for iOS major revs, which are about a year each [2]
I'm currently trying to upgrade a bit of software at the moment at work that's a few minor points out... but those minor points contain some delicious features. Doesn't mean it's not a minor revision upgrade.
I think you should touch on frequency of updates: how often does Apple release, how often does Google?
Also it's not really fair to expect that every phone is going to run the very latest Android code, especially ice cream sandwich, which expects a forward facing camera, doesn't use the navigation buttons and so forth.
I do agree with the general premise that Android carriers/manufactures tend to release and abandon. They have no interest in updating the OS after the sale, which explains the popularity of Cyanogen.
As JoachimSchipper said, provide a timeline and mark release dates of Os versions on it as well. This way it would be much cleaner and gives more context.
I actually had letters on there ("E", "F") etc for releases but on top of the sales windows & the support windows it was just too much on one chart.
I do plan to provide all that information as a comprehensive list of phones/updates, ideally along with the ability to sort by carrier, manufacturer, etc.
Could you also make a chart showing, how long the phone is supported after it stops selling? For example, 3G was being sold up until iPhone 4 introduction, but the iOS4 was the last update it has seen. Similarly, 3GS is still being sold, but that does not mean that it will be supported for next three years.
This is exactly the kind of in depth nerdery I love. Thanks for putting it together. It seems like it would be valuable as a standalone resource. You might even be able to get some help from the community in a wiki format for keeping it up to date.
Excellent, to the point visualization of an insight. I love it when journalists can use the table-chart format to create a data-deep yet simple graphic that tells a better story than any plain narrative can.
You should include iPod Touch and how apple forces a charge for updates on that device.
This chart is also skewed a bit and should be based on intervals of releases rather than "years since". I don't recall seeing any iOS5 backports to iPhone 2G or 3G.
No, the original iPhone & 3G don't support iOS 5. But they did support the current version of iOS for 3 years after they were released - far more than any Android phones of the period.
(And, no, they don't charge for iPod Touch updates any more - that was technically an accounting regulation issue that the government changed.)
Though to be honest, for the iPhone 3G, the longer green bar is easily misinterpreted as a positive thing, when for the most part, iOS4 rendered the phone very painful to use. 15-30s wait times for camera shutter, frequent hangs, constant crashes in Maps, being the main pain points with the two that my wife and I owned. If they'd have continued security and minor app updates for iOS3 on the 3G, I'd have been much happier.
The first releases of iOS4 for the 3G were pretty bad but the subsequent releases greatly improved things. I recently sold my old 3GS so I updated it to the latest 4.x version and it was totally usable. Slower than iOS2 but for a 3.5 year old device I thought it was completely acceptable performance. In the rush to release iOS4 I think Apple didn't have time to optimize the 3G release. They seem to have learned their lesson with iOS5 for the 3GS.
There is an enormous difference between a 3G and a 3GS. My wife and I have one of each, both running 4.2.1. Her 3GS is usable, my 3G really isn't. The 3H suffers from lots of OOM-killed applications, and a huge number of 30-40 second pauses that freeze the entire UI. (Although AT&T seems to be equally good at dropping calls placed from either.)
I hear you on the 'Apple didn't have time to get the 3G release right'. That's fine. I have no problem with being stuck on iOS 3.x. Good iOS 4.x would be best... Good iOS 3.x would be almost as good... but bad iOS 4.x is awful. They either should have invested the time to do the release correctly, or just not ship the update to the 3G.
The chart could definitely be made easier to read. I also thought it claimed that the original iPhone got iOS 5 since current = 5 right now. Instead, the chart says that after three years, the original iPhone was still running the latest version at that time (iOS 3)
"Years since" is the right metric, because mobile phone contract lengths are measured in years, not in some ad-hoc multiples of the times between releases of the OS.
I own a one year old Galaxy S, and lately I have been holding my breath hoping Ice Cream Sandwich will be released for my terminal, whether by Samsung or by Cyanogen.
I agree with many of the points people are making (there are some differences, there are many Helvetica/etc based fonts, etc). Really the most amusing thing to me is that Google made a big deal out of the type change - it was the first point in their Ice Cream Sandwich launch. Can you imagine if at Apple's iPhone 4/iOS 4 launch they said "and our first big feature is: we're dropping Helvetica! For Helvetica Neue!"
Here's an animation/toggle I did comparing Roboto & Helvetica: http://und.st/pqCj4I.
In the lowercase, there are only 5 letters (a, e, f, g, k) that a normal person would have much chance of distinguishing, even some of those are subtle. The other 21 letters appear virtually identical.
The 2002 date was when I was saying smart phones reached general consumer availability, not when the very first one was invented. As I mention in the article, it was 2002 when Blackberry phones and Microsoft Pocket PC phones came out, as well as the aforementioned Palm-based Treos.
FWIW, here's the source article: http://www.technologyreview.com/business/40321/ And the HN discussion thread (empty, just submitted it) for it: http://news.ycombinator.com/item?id=3951352