Other than the low price (only $199?!) and the fact that Google is getting absolutely nothing out of Amazon’s use of Android, I couldn’t really bring myself to caring too much about the Kindle Fire (Apple and/or Microsoft patent lawsuit in 3… 2… 1), but there is one aspect that intrigued me – Amazon’s beefing up of what at its core is Opera Mini.
The browser on Amazon’s Kindle Fire is interesting because it basically does what Opera has been doing for a long time (and BlackBerry too, if I’m not mistaken) – render websites on the internet in the cloud (sorry, this was getting confusing – I have to use the buzzword here) so the local rendering engine doesn’t have to do as much.
“Instead of a device-siloed software application, Amazon Silk deploys a split-architecture. All of the browser subsystems are present on your Kindle Fire as well as on the AWS cloud computing platform. Each time you load a web page, Silk makes a dynamic decision about which of these subsystems will run locally and which will execute remotely,” the Amazon Silk team explains, “In short, Amazon Silk extends the boundaries of the browser, coupling the capabilities and interactivity of your local device with the massive computing power, memory, and network connectivity of our cloud.”
Of course, this raises a whole bunch of red flags when it comes to privacy. Basically, Amazon will be storing all the sites you visit in its cloud, effectively maintaining a lengthy log of all your online activities. This is the holy grail of advertising, and you can bet your sweet bum that Amazon will be exploiting this opportunity. Not only for itself (i.e., suggesting items from Amazon’s own store), but of course also selling it to third parties.
Jon Jenkins, director of platform analysis at Amazon, told TechWorld that Amazon will not store personal information – but Silk’s terms and conditions tell a different story.
Luckily, the browser does have a no-cloud mode (probably built-in in case the Amazon servers go down), and you can turn it on to have all rendering done locally. Rendering, by the way, which is done using the WebKit engine, through SPDY (HTTP is still used between the website and Amazon’s cloud). There’s also hints of being able to use Amazon Silk on non-Kindle devices too, suggesting Amazon may be bringing this technology to other devices – Android phones maybe, or maybe even desktops and laptops.
This might just be me, but wouldn’t it be possible to do something like this locally? I mean, wouldn’t it be possible to have the portion currently running on Amazon’s cloud running at home, on your powerful desktop? My quad-core 8GB workstation would be perfect for the job – have the workstation render the complex bits only to send it on to my iPad 2 and Windows Phone. This would sidestep any privacy issues, and while it wouldn’t be very useful out and about, the fact of the matter is that the Kindle Fire doesn’t have 3G anyway so that point is a bit moot.
Seems like a nice enough idea to warrant some discussion, I suppose?
To me, it seems like this might be a proper start of “solid inexpensive Android tablet” (or ~Android, whatever); “shit just got real” saying nicely sums up the Amazon tablet entry in that regard, I think. Only too bad the sub-100 USD e-paper Kindle is so relatively late.
Though OTOH – a black rectangle with rounded corners?! How dare they?! That might put Amazon into BIG trouble… Also, for their sake and mine (cheap headline shots would quickly get tiresome), I hope Amazon pushed for far-beyond-standard levels of making sure that the batteries are unlikely to catch fire
Too bad this version doesn’t seem to have Bluetooth – straightforward BT data connection, provided by inexpensive (so called) feature phone, would make the thing an even nicer deal.
Also, I would like to propose a new buzzword of choice for what was once the internet, or now cloud: fog.
And remote desktop connection via fast Wifi when at home, PC browser displayed on a tablet, would probably provide some sort of basic makeshift local equivalent to Silk.
PS. Before the inevitable… yes, it is an iPad competitor, people will do on Kindles virtually the same type of stuff as on iPads (with the exception of camera-related functions of course; large part of which are a quickly forgotten toy with most people); that puts them in the same product category. And as for the Nook Color… its manufacturer doesn’t want to compete in the first place, they won’t even sell it to me (wrong country, continent, or hemisphere, it seems)
Edited 2011-09-29 20:47 UTC
When refering to hosted services, I like to use the more accurate term of “hosted service”. Buzwords are only of use to marketing folk trying to re-sell the exact same hosted service under a new hip add comapain.
it seems like this might be a proper start of “solid inexpensive Android tablet”
The analists say that Amazon is losing $50 for every tablet sold, so its cheapness is finanzed by Amazon.
Yes, but that is what it will likely take… Hardware profits are not the point, it is all about owning the distribution points.
Amazon is at it’s heart a digital commerce company – tablets are nothing more than razors to them – they want to sell the blades…
If this works for them and uptake on Kindle Fires picks up dramatically, I fully expect them to push the pricing down even further. Put another way, a really good way to compete with Apple is shrink their market with price pressure. Apple has historically never dealt with this well – they don’t want or like to make “bargain” technology.
It sounds good, but the real losers are the ones who don’t have the services Amazon has to offert and just make money from the hardware, so It could do more damage than benefit.
I agree. I’m not saying it is good, but it is what it is. I would argue that the reality now is that if you are not Apple, Amazon, or another media powerhouse you have already lost…
I left out Google on purpose. They are essentially farming out the hardware so they can play middle man for all the commerce on their devices. They are going to suffer the most by a price war, because Android tablet makers who got into the game for hardware profits will simply have no reason to stay around… Hence why Amazon essentially took Android and annexed it for their own use. If tablet makers use Android the way Amazon has, Google cannot monetize it…
This could actually be a legal problem for Amazon down the road. If Amazon is in fact selling these things for below cost, I could see lawsuits for anti-competitive behavior coming from some other tablet vendor who doesn’t have comparable ways to recoup losses. Amazon’s massive bundled service infrastructure definitely has the potential to push fringe hardware competitors out of the market if they can leverage those services to sell their tablets below cost.
I have no idea whether or not any such suit would be successful as proving predatory pricing is very difficult, but I imagine there has to be some limit to how far you can take this “razor blade” model as being legitimate competition.
I doubt it.
Back in the grand old days, with the NES and later SNES Nintendo introduced the “loss-leader” concept to the video game market, where the game system was relatively inexpensive and the games were the real profit bringers. This practice carried over to the other console manufacturers like Sega and later Sony. In fact, from what I remember only the Neo Geo was priced for a profit on the console itself, and it was damn near unaffordable at ~$600. They didn’t sue Nintendo, Sega or Sony for selling loss-leader hardware from what I recall.
That’s not to say it won’t happen here, but again I doubt it. It’s been in practice for nearly 30 years now in the tech field (not to mention the razor blade manufacturers as you mentioned) and I don’t see it being challenged by other companies in court.
That’s not to say it won’t be challenged in the field though. Brother laser printers, for example, use chipless “dumb” cartridges that seem to be designed to be refilled by the consumer; all you have to do is pull an easily accessible plug, fill with toner and replace the plug, and reset a gear on the side. A Philips screwdriver and five minutes of your time save a lot of money. That combined with the fact that their cartridges are also very inexpensive to begin with, and the printers themselves are great quality, and you have honest pro-consumer competition against the gouging tactics of HP, Lexmark and Samsung.
Edited 2011-10-04 03:39 UTC
So? That doesnt make it not cheap and it doesnt make it not solid.
Apparently these estimates came from appleinsider which is a site know for idolizing apple technology (not so bad) but also for trying to damage every other bit of competing technology (the much worse part). As such these estimates MUST be treated as suspicious.
It’s a loss leader, and Amazon will undoubtedly make it up when people buy e-books and e-magazines.
Ahh .. yeah … Man in The Middle attack ftw …
“This is the holy grail of advertising”
Just like some ISPs, for example in the UK, have been caught replacing ads on webpages. Amazon is in the perfect position to do the same.
They say do don’t store anything.
The most interesting aspect of the Fire to me is the fact that, at the ‘default’ price, you agree to receive ‘special offer’ (aka advertising) essentially constantly on your device (it’s displayed on the home screen).
The version without advertising will cost you and extra $40.
It’s the same for all the new kindles.
Amazon is gunning for Google in a big way with this. I wonder if there will be some kind of Amazon powered search in this or will Google be the default.
]{
where do you see that? I cannot find the ads spam info on the kindle fire website ….
… Opera comes up with good-to-great ideas, but coupled with poor execution
Ah so that’s why Opera Mobile is the nr1 mobile brower if you look at the international picture?
I thought the market share was falling? Newer androids don’t come with opera bundled either. Don’t get me wrong — I love opera. I use it on desktop, my iPod Touch, my dumb-phone and android. But I don’t think the company as a whole is doing a great job in expanding their marketshare
I love Opera mini on my iPhone 3G. It’s the only fast application I have on this dinosaur
Opera Mini is the one with significant userbase – and that is because it is a not-completely-useless browser that runs on cheap feature phones that come with completely-useless browsers (or in many cases come with Opera Mini itself)…
Mini is for feature phones, Mobile is for smartphones, and it has a dramatically smaller userbase because most smartphones come with decent browsers.
That is not a knock against Opera or anything, but you are twisting the truth a bit. You make it sound like people are tripping over each other to run Opera because of its worldwide popularity…
Mini is widely used for no other reason than for most phones it runs on it is the only decent browser available. Virtually no one runs Opera on Android or iOS (although it does have a small but significant presence on Windows Mobile 6.1)…
No, that’s why its not the number one mobile browser, if you don’t look at the “international picture”. The execution of the idea sucks on devices used in top tier mobile devices.
How is Opera Mini a poor execution of its (merely similar!) idea? It even targets completely different kinds of devices – look at the top handsets in country snapshots of their stats ( http://www.opera.com/smw/2011/08/ ); some of those mobiles (Nokia 2330 certainly, C1-02 most likely, not the only ones) can be had for less than 50 bucks, new, without contract (majority of the 5+ billion mobile subscriptions)
Even such phones work well with Opera Mini, giving quite full access to the web; show me a browser which works better on them.
Oh, while saving huge amounts of data costs (particularly important for the millions who want to be online but can afford only inexpensive phones)
Well, then I must say I chose the wrong word there.
I use opera mini on my ipod touch at home even when I have it connected to my WiFi; so I know how useful it gets. Even on snappy networks it can get snappier — the execution I’m talking about is the polish and marketting, rather. Take speed dial of opera vs chrome; or browser sync, or pretty much any awesome concept that has me stuck using it. It works, but it lacks polish until a competitor picks it up.
As for marketshare — the “completely different devices” — that’s my primary issue. It feels like people are content with it being the niche browser. All markets will slowly migrate to smartphones (I live in india, so I can fairly say I have some idea) and I rarely see people use opera mini on their ‘droids.
Edited 2011-09-30 00:52 UTC
Oh well, that’s also sometimes the result of being the pioneer, I guess – when the idea is not yet clarified fully, not yet refined…
And there are still billions of subscribers with phones which can’t run ever Opera Mini (S30 class devices), there should be place for such browser for quite some time. Either way, Opera so far is profitable and on a healthy, sustained rise (also during recent economic turmoil; they are a publicly traded company, you can check their financials)
Actually, opera mini isn’t the only java browser using cloud rendering technology.
There are also Bitstream Bolt, ucbrowser and one specific for Nokia S40 devices.
As for sites compatibilty Bolt runs circles around Opera Mini supporting even ajax heavy websites and offering usefull full screen rendering on screens as low res as 200×400.
Nowhere did I claim it is the only one; but show me a better one on such inexpensive sub-50 devices
They aren’t the target of non-basic Bolt, resolution you cite is enormous / reflow is basically required on ~50. And either way Bolt seems to be a no-go for most of the world alphabets, it doesn’t work at all for many people (with those using “niche” alphabets probably disproportionally represented among people who have access mainly just via an inexpensive mobile)
Also, looking at the Opera Mini top visited sites ( http://www.opera.com/smw/2011/08/ ), “even ajax heavy websites” don’t seem to be a problem.
In addition to “poor execution”, with the revised meaning of “lack of polish and marketing” you made clearer, I’d add that there is also some stillness i Opera when it comes to features and bugs. Issues that have been reported remain open for months if not years.
For instance, the default download folder is still not available as an option in the file download dialog, there is still no ‘delete’ button for items in the opera:cache, the tabs extension API has never been completed and its last update was in January, I’m still waiting for the option to have locked tabs behave like they did until 11.00 mimicked the Chrome “shrink and move to left side of tab bar” behavior, the support for non-ASCII characters in keyboard shortcuts is still buggy, custom form fields are nowere to be found… I could go on for pages.
On the side of never-to-be-seen features, how long have we been asking for PGP or S/MIME integration in M2? How about restoring the lightspeed-fast back and forward navigation we had in Opera 8 in lieu of the joke that the current “fast history navigation mode” is? I guess some things will never be implemented. But who knows? We’ve waited equally as long for HTML emails and now we have them.
Oh how I miss this feature
Lots of ISPs (well my home connection) run caching proxies…lots of businesses do the same – saves bandwidth and generally gives better performance, at least for the more commonly visited sites. If Amazon are smart they’ll add on some of the other advantages along with SPDY, like zero day virus and malware blocking, ad blocking and anything else they can think of…
P.S. Funnily enough, my ISP even has a bittorrent “proxy”, i.e. a local peer – gives great speeds on popular torrents, not so sure about the legality though…
Caching proxies à la Squid have seen their usefulness degraded by the overuse of: no-cache, inter website linking, “over-dynamicizing” (note the double “over”) of static-compatible content (of which the culprit is advertising), and the likes of Google Analytics. And Javascript libraries didn’t make the whole situation better. So many bad coders and clueless webmasters.
The great thing about is that it is just 200 USD and runs Android and it has a good CPU OMAP 4330. The bad thing is that it isn’t available in Europa for preorder.
I don’t think Amazon ships to the outer planets yet.
Shitsville seems to be fine though.
At $80 for a simple e-reader and $200 for a tablet, Amazon is doing some nice price-cutting there. I remember when, a few years ago, they told that the thing couldn’t truly take off until it starts costing less than $100, time to check this declaration I guess.
Now, I’d blindly hand my credit card if only they could do an A4 version of their e-ink product, something like the former Kindle DX. But I guess demand is not sufficiently high. Only scientists read documents printed on A4 paper after all
Edited 2011-09-30 05:41 UTC
“The thing” being e-readers, of course… I’m so tired…
Check out iriver Story HD. Not exactly A4, but apparently has a much better PDF reader that can reflow text. They also slash the price from 139 to 99 for the period until Oct 15th.
Edited 2011-09-30 23:28 UTC
Thanks for the reference, didn’t know about this device ! However, after looking around a bit and trying to print A4 documents in A5 format to see how it would look like (I don’t trust automated reflow), I think I’ll still wait for cheaper A4 readers, should they come one day…
(I’ve found some from other manufacturers, but they are also priced in the 250€-300€ range)
Edited 2011-10-01 08:20 UTC
Cyanogen Mod!!!
I’m sure Amazon will try to lock this down just as B&N did with the Nook. After a few months I took the plunge and rooted my Nook and installed Cyanogen and have never looked back. Sounds like a great alternative to the Nook for people who have a few skills and don’t mind applying new firmware.
But good Linux apps pretty much all tend to be cross platform….. that’s part of the point of Linux, you aren’t locked in. Good don’t stay Linux specific for long.
How you move people to Linux is you move the apps they use to free ones. Then just change the OS underneath for a faster more secure free one. 😉
Sorry, wrong article. I’ll show myself out… 😉
The comparison being made to Opera Mini is drawn by people who seem to have very little actual knowledge of mobile tech.
Opera Mini converts at server side requests into a binary markup format (OBML). The browser in “turbo” mode becomes a glorified picture viewer with some limited scripting ability (ie xmlhttp requests are supported but only in a very strict sense) and is analogical to SILK only in the sense that they are both proxies.
If anything given the information presently at hand, the Amazon system is closer to the Blackberry browser (which was decried as a horrible browsing experience by many of the same idiot pundits hailing SILK simply because it is attached to Android). Interestingly, people seemed to trust Blackberry with respect to MITM and now distrust Amazon. I guess because RIM has never been caught blatantly taking advantage of its users by selling and/or making use of the private information of its users for profit ala Amazon and Google.
There is one feature of the system that really strikes me as troublesome. Amazon is using “machine learning” to prefetch pages/resources before a user actually requests them. Have to imagine if this algorithm is not very good this would lead to requests being made and resources drained from web servers having nothing to do with Amazon (which kinda violates the spirit of the web) and increased data plan usage by the consumer.
SILK purports to offer an interesting innovation that I haven’t seen before in either Blackberry or Opera (but which may exist given that both are very tightly held proprietary systems) that Amazon engineers claim that they will compile javascript server side into ARM binary for consumption client side which combined with the massive caching resources available to Amazon will effectively be creating an ARM-only compilation farm/application server for the web bypassing the javascript engine in the embedded browser. Adds a new twist to the native v html mobile app debate.
Could be significant speedups depending on their compilation toolchain but I have to imagine you will quickly run into serious hurdles wrt to expanding to a very fragmented Android userbase if they are compiling beyond Thumb (such as the use of softfp v hardfp, NEON, ThumbEE which may or may not be present, etc). So keeping SILK exclusive to the Kindle lineup makes some technological as well as business sense.
Will be interesting to see it in action and test it out on complex web apps. But unlike some pundits who jump on bandwagons for pageviews without ever seeing, evaluating, or understanding the system they are writing about (such is the state of “tech journalism” nowadays) I’ll reserve judgement until it can be put to some realworld tests.
Like most software, it will probably suck until release 2.0.
I kinda started down this road a couple of years ago, creating a proxy “compiler” that translated javascript not into ARM binary but java bytecode for consumption by an app leveraging the Jazelle RCT hardware acceleration present on pre-ARM7. Quickly ran into obstacles getting any real documentation and put the project on the shelf (ARM for whatever reason decided to license the SDK and so its adoption for this and other reasons was limited primarily to major JVM implementors and I think even dropped entirely from Cortex A8).