Chrome Netbooks a Year Off

Google No Responses »
Nov 222009

Google says it’s working with a bunch of unidentified “top OEMs” to create a Chrome notebook based on a Google hardware reference architecture that will hit market around this time next year in time for the holidays.

It expects the widgets to change the basic computing model by transferring everything to the cloud.

Google accepts that such a device won’t suit all use-cases but it should suit those, particularly youngsters, who spend most of their time on the web, to a “T.”

Google will position the thing as a “companion device” to one’s more robust and conventional laptop or desktop.

It means to tackle laptops and desktops somewhere down the road.

Google said its reference architecture, whose components are still apparently being spec’d, won’t support hard drives, only solid-state memory, and will demand a somewhat larger form factor than today’s notebooks to provide for a full-sized keyboard and touchpad. Its connectivity will be Wi-Fi.

Google disclosed its plans Thursday in a Chrome OS webcast, where it demoed the thing and open sourced the code so developers can play with it and contribute to the project.

Remember, Google’s never developed an operating system before. It’s big and it’s hard. It’s also a matter of winning mind share and an ecosystem.

The open source contingent will reportedly work with the same source tree as Google developers. What with may be problematical since Chrome OS doesn’t work on conventional machines; virtualized is a possibility. Google will have a list of validated hardware.

Chrome OS, whose watchwords are “speed, simplicity and security,” is supposed to make the Internet instantly accessible like a TV – Google is currently trying to get its boot time to under seven seconds – and – in a challenge to Microsoft – every application it runs will be a web app.

It promises no software install, maintenance or management since the software will be in the cloud along with all the data, which is synced to the cloud, making offline use possible via cached data.

All data, Google said, including local data, is encrypted by a random number algorithm. And, unlike your conventional operating system the root file systems is read-only.

Provided the cloud is functioning, it’s supposed to solve the security problems of the traditional PC. All applications are treated as a hostile, Google said, and locked down in a “security sandbox, separated from both each other and the operating system.

Chrome OS looks like the Chrome browser with application tabs at the top of the screen. And Google talks about Chrome OS as a browser.

Google declined to discuss the cost of the reworked x86 netbooks, begging off that it was up to the hardware makers and component costs can change in a year. Apparently it’s shooting for current price levels. It suggested that other browser makers could develop Chrome-style packages of their own.

Google also begged off discussions of an app store.

The widgetry supports Flash and where possible open source drivers and the usual slew of storage devices and mice. Android apps, not being web-based, won’t work on Chrome OS. Chrome OS apparently doesn’t print yet but it’s supposed to before it comes out.

Chrome the browser, meanwhile, now in its twentieth update is close to launching on the Mac and Linux. Google clocked the Windows version as 39x times on JavaScript than IE. It claims 40 million users.

Posted by rhall2091 at 6:15 am

Unix Co-Creator Writes New Open Source Programming Language for Google

Google No Responses »
Nov 122009

Ken Thompson and Rob Pike, a couple of the revered old soldiers who wrote Unix, Plan 9 and Inferno at Bell Labs – and programming languages like Limbo and B, without which C probably wouldn’t exist – have come up with an experimental new programming language called Go that Google, their current refuge, open sourced Tuesday under a BSD-style license.

If it catches on, it would be the first major systems language to emerge in a decade.

One of the small Go team compares it to Java, calling it “performant but garbage collected” and “vastly more enjoyable to code in.”

Go’s primary attribute appears to be that it’s fast. It reportedly compiles to machine code very quickly thanks to a couple of Plan 9-beholden compilers.

Google says “Go combines the development speed of working in a dynamic language like Python” – Google does dote on Python – “with the performance and safety of a compiled language like C or C++. Typical builds feel instantaneous; even large binaries compile in just a few seconds. And the compiled code runs close to the speed of C.”

Go has been in the works for the last two years, but has only been a full-time project at Google since the middle of last year.

It was born, the Go team says, “out of frustration with existing languages and environments for systems programming. Programming had become too difficult and the choice of languages was partly to blame,” a statement that sounds very much like Ken Thompson.

“One had to choose either efficient compilation, efficient execution, or ease of programming; all three were not available in the same mainstream language. Programmers who could were choosing ease over safety and efficiency by moving to dynamically typed languages such as Python and JavaScript rather than C++ or, to a lesser extent, Java.”

“Go,” says one of the team’s FAQs, “is an attempt to combine the ease of programming of an interpreted, dynamically typed language with the efficiency and safety of a statically typed, compiled language. It also aims to be modern, with support for networked and multi-core computing. Finally, it is intended to be fast: it should take at most a few seconds to build a large executable on a single computer. To meet these goals required addressing a number of linguistic issues: an expressive but lightweight type system; concurrency and garbage collection; rigid dependency specification; and so on. These cannot be addressed well by libraries or tools; a new language was called for.”

Go descends from C in syntax and borrows declarations and packages from the Pascal/Modula/Oberon family and concurrency from Limbo, but it’s still a completely new language, its creators say, meant to escape the “bookkeeping, repetition and clerical work” of current languages as well as their “clutter and complexity” while still being “sophisticated.”

It’s object-oriented but without the type hierarchies.

Although it’s meant to make writing the servers and other software Google uses internally a lot easier, Go isn’t mature enough for large-scale production use yet. Its development team is experimenting with it as a candidate server environment and the server behind http://golang.org is a Go program.

Since Thompson’s involved, one would assume that the project’s called Go after the oriental board game that’s sort of like chess. He worked on Belle, the first chess program to get a master’s rating.

See http://golang.org/doc/go_faq.html and http://golang.org/doc/go_lang_faq.html#top.

Posted by rhall2091 at 7:40 pm

Google Apps Lands its Biggest Fish Yet

Google No Responses »
Oct 172009

Rentokil Initial, an odd combination of a modern-day charwoman who in her off-hours is a courier service, exterminator, insurance adjuster, heating and air conditioning installer and water and electrical tester, has gone with Google Apps, expecting to be the largest deployment of its paid Premier Edition yet.

The cloud widgetry will replace Rentokil’s 180 e-mail domains and 40 mail systems – everything from open source to Microsoft – as it rolls out to 35,000 employees in 50 countries and six operating divisions by the end of next year.

The UK-based conglomerate called it part of a “five year turnaround plan to enhance capability and introduce operational excellence” ands said the decision was less about saving money than “creating capability.”

The widgetry will go to 20,000 desk-bound PC users and 15,000 mobile folk who currently don’t have a company e-mail address. They are expected to use PDAs, home computers or shared PCs at their branch office.

The company also expects to use Google’s integrated chat and video features to support training and improve productivity and collaboration within and between divisions and functions. And it’s got a real use for Google’s automatic e-mail translation and the real-time translation in Google Talk.

Rentokil ran a 100-day pilot at its Ambius office plants division that involved 800 users in global locations.

Google Apps includes Gmail, Google Calendar, Talk, Docs and Sites. The Premier Edition, which includes 25GB of e-mail storage, usually runs $50 (40 euros or 33 pounds) a year per head. Terms were not disclosed.

Iron Mountain: A Rock Slide on Rival Cloud Storage?

Iron Mountain – where companies have been storing their IP and data for a couple of generations now – has opened its Archive Services Platform so developers and ISVs can offer its cloud storage with their applications.

The great $3 billion-a-year pack rat has never extended its platform to outside application developers before.

Besides long-term data storage for data-intensive apps, Iron Mountain is also offering use of its own advanced indexing, full-text search, retrieval, metadata support, retention and destruction.

The company claims to be more practiced at this cloud stuff than other people. It’s been providing cloud storage solutions since 1995 when it brought out Connected Backup for PC and started using the Internet to provide secure backup for companies’ and individuals’ PC data.

It subsequently developed a Storage-as-a-Service suite of solutions for everything from secure e-mail to file and image archiving, online backup and recovery and software escrow.

Founded in 1951, Iron Mountain has 140,000 corporate clients worldwide and can argue security and reliability. On the face of its reputation it should pose a very serious competitor to the other cloud storage wannabes.

Taneja Group director of validation services Jeff Boles said, “The cloud is bursting with companies eager to deliver storage services to business customers. Success will require a lot more than a simple API and a bucket of storage on the remote end of a wire.”

Iron Mountain has set up a developer program to provide third parties with the tools to integrate with its enterprise-class cloud archiving services either by coding to a Web Services API or through a set of command-line tools.

Besides the SOAP and REST Web Services interfaces, the platform uses standard NFS and CIFS interfaces and offers petabyte scalability.

Data is stored using 256 AES encryption technology and the platform is WORM-enabled when retention is applied.

Data is stored redundantly within and across geographically split data centers, in other words, four times across two data centers.

Stored data is full-text indexed when it comes in and can be searched by ID, content and metadata. The platform computes SHA-256 hashes of the data and check hashes for discrepancies on each and every step of storage and retrieval to guarantee data integrity.

Iron Mountain employs both Secure Socket Layer (SSL) and Transport Socket Layer (TSL) security and certificates, good for a year, instead of just user IDs and passwords in transmitting digital assets to and from its data centers. Each transaction includes bilateral authentication.

The company’s also got a lot of physical security: underground sites, real-time closed circuit TV monitoring, a commercial-level power grid with generator backup, on-site firefighting apparatus and personnel, and multi-tiered access controls. Only authorized personnel with biometric passes are reportedly allowed in specific areas.

The company’s vague on the actual numbers but says that users are charged a monthly per/GB storage fee and “there are no hidden charges for transmitting or retrieving data.” Developers will be sent a bill itemizing each customer’s storage use every month.

They can also use a co-branding “Protected by Iron Mountain” logo.

Posted by rhall2091 at 5:28 pm

Google Wave Hits Wider Beta

Google No Responses »
Oct 012009

Google Wave, the amorphous open source widgetry that Google has trouble explaining but contends – silly Google – will replace e-mail, the most viral application ever, started moving into a wider test group of some 100,000 users Wednesday ahead of still wider release in December.

It’s akin to a limited launch. Wave reportedly got a million requests for early access.

Wave’s Australian-based developers describe e-mail as the Internet version of snail mail; e-mail isn’t fast enough or multitasking enough.

The latest cut of Wave includes people who signed up early for the preview, developers active with the thing and some Google Apps users.

The communications widgetry – from the same people that created Google Maps – is suspected of being either Google’s answer to unified communications or to social networking.

Since Wave was unveiled in May, the company has reportedly been fixated on scaling it, stabilizing it, making it faster and more useable. It admits it’s still bug-infested, crash-prone, slow and “quirky.”

Wave is supposed to combine highly shareable elements of e-mail, chat, wiki documents, blogs and photo sharing on a single platform, creating a so-called “hosted conversation” or “wave.”

A user can, say, chat about a document or do document edits in real-time; e-mail can be edited by several people simultaneously; a replay any conversation.

It’s estimated Wave is still a year away.

Posted by rhall2091 at 5:56 pm

Google, the Verb, Versus the Bing Ding

Google, Microsoft No Responses »
Jul 312009

Microsoft’s long, tedious, temper-flaring pursuit of Yahoo has finally culminated in an exclusive 10-year deal.

It’s not the deal people expected which is why Yahoo stock took a 12% fall in the hours after the companies laid it out Wednesday.

Yahoo is going to sell search ads for the both of them using Microsoft’s AdCenter platform. It’s going to use Microsoft’s newfangled Bing search engine on its web sites. And it’s licensed its search technology to Microsoft to integrate.

The companies expect to get regulatory approval early next year, which Google will attempt to block, and it will take about two-and-half years before they’re operating at the top of their combined game.

It’ll take them that long to unwind what they’ve currently got geography by geography starting with America and to integrate whatever they’re going to integrate.

Microsoft will pay Yahoo 88% of the search revenues generated by its sites for the first five years of the agreement.

There’s no “boatload of money” upfront for Yahoo like CEO Carol Bartz suggested there would be a few months ago. She says she traded it for a sustainable TAC. “Having an upfront payment,” she said, “didn’t really help us from an operating standpoint.”

“Boatloads of money” has now been translated into “boatloads of value for Yahoo,” which gets to focus on media, marketing and sales.

The contract doesn’t cover display ads, a province the companies will continue to compete in.

But Bing will get the traffic Microsoft wanted.

Microsoft is widely seen as having gotten the best of the bargain.

Advertising Age, which scooped the story Monday, has observed that the arrangement eliminates Yahoo as a Microsoft rival and consolidates ~30% of the US search market on Microsoft’s widgetry. Bing’s early results suggest it’s eating into Yahoo’s market share.

Google’s still got 65% of US traffic, 92% in Western Europe according to Microsoft CEO Steve Ballmer. The great question is whether Bing and friends can ding that commanding lead any.

Dubbed a “decision engine,” Bing is supposed to provide better results – or will with more traffic to analyze. As part of the deal its name will appear on the Yahoo results page where it will say “powered by Bing.”

It’s supposed to tickle Yahoo’s share of search advertising. Advertisers are supposed to be attracted by the improved scale of the combined companies and the simplicity of working with a single platform and sales operation.

Since Bing was introduced in June backed by a reported $100 million ad campaign Microsoft still only has a thin 8.4% of the US market.

Yahoo will also save the money it would have cost to continue to support its own search engine, money it can put into sales staff. Yahoo anticipates saving $200 million in CAPEX and $500 million in OPEX a year within two years of regulatory appoval. It also thinks it’ll see an operating cash flow of $275 million.

Certain Yahoo personnel will shift to Microsoft and some will simply lose their jobs, Bartz said.

The arrangement is a far cry from Microsoft’s original proposal that it take over Yahoo’s search and search advertising – a year ago Microsoft said it would pay a billion dollars for Yahoo’s search business – and may be the result of antitrust concerns. It’s further still from the $47.5 billion acquisition bid Microsoft put on the table last year.

The complex talks reportedly broke down briefly last week because Yahoo wanted hundreds of millions of dollars upfront plus revenue guarantees potentially worth billions over the life of the deal.

It also wanted more revenue for search that originates on its site and leads to a purchase and for clicks on ads. And there have been issues over Bing branding and the amount of search data Yahoo would get.

Yahoo emerged from the negotiations with limited access to the search data.

The pair expects to push paperwork to the regulators next week. It will argue that their alliance heightens competition.

The deal is not expected to set off the same kind of fireworks that the government-squelched “we’ll sue you if you do” Google-Yahoo deal did but it won’t go through without a fight. Microsoft raised regulatory antennae to the Google-Yahoo deal and Google will return the ball. An Obama Justice Department is also likely to paw over the thing. Reuters got to Senate Judiciary Committee chairman Herb Kohl, a Democrat, who also apparently wants his turn at bat.

The pair has set up a site at www.choicevalueinnovation.com.

Posted by rhall2091 at 6:07 pm

Google To Develop Operating System To Counter Microsoft

Google, Microsoft No Responses »
Jul 092009

The shoe that Google’s been itching to drop – the one everybody knew was dangling – has finally dropped.

With Pearl Harbor-like timing of a declaration of war, Google VP of product management Sundar Pichai and engineering director Linus Upson said in a midnight blog post Wednesday morning that Google’s going to develop a fast, lightweight, open source operating system based on its nine-month-old Chrome browser to compete against Microsoft.

The so-called Chrome OS, described as Chrome running in a new windowing system on top of a Linux kernel, won’t be ready for prime time until the second half of 2010, a lifetime in Internet minutes.

The schedule gives Microsoft, which is busily detaching its browser from its operating system in Europe, plenty of time to figure out how to retort if it doesn’t already have a plan other than Windows 7.

Google’s aiming to put the thing on ARM- and x86-based netbooks first and sometime after that on full-sized desktops, both places where Microsoft overwhelmingly dominates. Google says it’s working with “multiple OEMs to bring a number of netbooks to market next year.”

It later identified them as Acer, Adobe, ASUS, Freescale, Hewlett-Packard, Lenovo, Qualcomm, Texas Instruments and Toshiba.

As open source Google says Chrome OS will be free, a clear attempt to drain Microsoft’s treasury.

Chrome OS is not Android, the Linux-based operating system Google is pushing onto phones, although the two admittedly overlap since Android is also supposed to fit into netbooks.

Instead Chrome OS is described as a new Linux project that will be open sourced later this year so the community can contribute.

Google says existing operating systems, meaning of course Windows without saying so, were designed before the web was invented. The web and applications running in the browser – not the operating system – are central in Google’s thinking. Chrome OS is supposed to get the user onto the web in seconds via a minimal interface designed to stay out of the way.

The company says it will “completely” redesign the underlying security architecture to avoid viruses, malware and security updates. (Cookie-free might be more useful at this point but unlikely coming from Google.)

Google promises that all web-based applications will automatically work and any new applications written for the thing will also run on any standards-based browser on Windows, Mac and other Linux operating systems “to give developers the largest user base of any platform.”

Google uses Linux internally so it can presumably make an operating system robust enough to hold up. But netbooks are not the stuff of mission-critical work, Linux has so far failed to loosen Microsoft’s grip on PCs (not even on netbooks), and the Chrome browser has yet to make much headway against Microsoft or even the Google-subsidized Firefox.

By Google’s count it’s only used regularly by 30 million people, a relative rounding error.

Despite its talk of community development, which is probably more viral marketing than anything else, an operating system is also going to cost Google a pretty penny or two to extend, support and maintain. (Ask IBM. Its futile OS2 efforts cost it close to a billion dollars a year back in the day.) And Google might actually have to talk to people to support the thing, something it has no cultural skill at.

Heck, it might even have to learn how to write a press release – or talk to the press. A blog posting is not marketing.

The object of Chrome OS, aside from harrying Microsoft, which is trying to harry Google back with its new Bing search engine, is of course to sell more ads by driving traffic to Google Search and other Google “services.” It is not about getting work done.

The Guardian reckons Chrome OS will “strip whatever hardware it runs on of most of its usefulness, without actually reducing the price by very much.”

Tim Negris, the former IBM and Oracle VP who invented the expression “thin client” for Larry Ellison when Oracle and its pal Sun Microsystems – Google CEO Eric Schmidt’s alma mater – were pushing the initial concept of a cheap network computer, takes the Chrome OS for a poor play on that last century idea.

He gives Google “all props” for its success so far but calls the Chrome browser “a fat client in a girdle, sneakily sucking cycles off the plates of other (non-Google) apps” and says that “even if that weren’t the case, it will still probably make a crappy app platform. A largely stateless, page-oriented interface is not the best way to interact with most computer applications, other than stateless page-oriented ones like web sites. Windows, Linux and Mac OS (Mach Unix) are far from perfect and I am a big believer in the concept and manifest destiny of the thin client, but, for Google’s particular concept to work, the Internet itself must become an operating system, and not simply be a means of linking one operating system instance to another. Google is taking the web as we know it, including 2.0, as far as it can, but it’s not in a position to drive the changes needed to make the true always-on thin client a reality. That task is in the hands of Cisco, Akami and others who can make the connective infrastructure much smarter than it is today, and that is what it will take to make the endpoints as dumb and simple as Google would like them to be.

“The Plan 9 and Inferno distributed operating systems work that was done at Bell Labs some years back was a lot closer to the right approach than the dog’s breakfast of network computing being served up nowadays by Google, Microsoft, Apple, the Linuxes, Oracle and other ‘end point’ vendors. The Bell stuff was not perfect or complete by any means, and probably flawed in places, but that’s not why it didn’t go anywhere. Plan 9 was smothered in its crib and Inferno sent to a foster home by their impatient, unloving parents on the business side of AT&T. Google would do well to pocket its pride and reach out to people like Rob Pike, Ken Thompson, Dennis Ritchie, Brian Kernighan, Bjarne Stroustrup and others who know history well enough to avoid repeating it.”

Google is ratcheting up its challenge to Microsoft elsewhere too. Hours before its Chrome OS disclosure, it finally ripped those perpetual beta labels off Gmail, Google Docs, Google Calendar and Google Talk to appeal to a wider swath of the business set.

Posted by rhall2091 at 6:45 pm

Google App Engine Learns To Speak Java

Google No Responses »
Apr 092009

Google’s now year-old App Engine infrastructure, previously limited to running only programs written in a particular species of Python, a less-than-mainstream tongue but an internal Google favorite, is learning to accept programs written in Java.

With the move, Google is reaching out to a broader base and interfacing with what it acknowledges are “businesses’ existing technologies.”

Apparently Google is using JVM 1.6 which means it should be able to support Ruby on Rails too.

App Engine’s masters say Java was the first and most popular feature requested and that those requests extended to the other programming languages that have been implemented on top of the Java virtual machine along with the web frameworks and libraries.

It seems that marrying Java to Google’s infrastructure is no mean feat. Google is currently testing the outcome. It can’t guarantee that all the tons of Java code out there, particularly the stuff that ignores sandboxing, will work. It says, “We know that there will be some rough edges when it comes to compatibility.”

To smooth those rough edges out, it’s giving the first 10,000 interested developers a Java preview, looking for feedback.

The widgetry apparently wraps current App Engine APIs with Java standards like the Java Servlet API, JDO and JPA, javax.cache, and javax.mail. It also provides a secure sandbox that’s flexible enough for developers to break abstractions at will.

The widgetry is supposed ensure easy deployment of Java code to all standard J2EE servlet containers including WebSphere and Tomcat.

Google says instead of using the underlying App Engine datastore APIs, developers can program against Java Data Objects or the Java Persistence API as a way to deal with App Engine’s proprietary BigTable database.

Early support includes a Java runtime and integration with a new App Engine-friendly version of Google Web Toolkit 1.6 for turning Java into JavaScript and Google’s Plugin for Eclipse.

Google says these tools provide a unified development experience for writing AJAX apps in a single language from client to server.

Meanwhile, as Google pours Java all over Apps Engine, it’s also more quietly revving Python to make it faster under a project called Unladen Sparrow, a Monty Python reference.

According to a company blog – and having nothing to do with Java – App Engine has also added centrally managed access to on-premise corporate data and applications locked behind the company firewall complements of what’s called Secure Data Connector (SDC), another serious play to put the enterprise in the clouds. It will work with Google Docs and Google Gadgets. The data is reportedly encrypted.

Oracle said its Siebel CRM apps will support SDC, opening the door to next-generation SaaS applications, multi-tenancy support and an “in-the-cloud Internet-based environment.” It’s also got an Oracle Gadget Wizard for Google Apps for its CRM.

There is also so-called Cron support for App Engine to automatically schedule tasks like report generation or DB cleanup at user-specified times and the ability to batch-import gobs of data from conventional databases into BigTable. Google hopes to provide matching export capabilities in the next month.

Google says in the year since App Eng was launched 150,000 developers have built 50,000 applications that now generate upwards of 100 million pageviews a day.

Posted by rhall2091 at 6:20 pm

Google’s Americas Sales Chief To Run AOL

Google No Responses »
Mar 122009

Time Warner dumped the misbegotten team of AOL CEO and chairman Randy Falco and COO Ron Grant late Thursday and wheeled in Tim Armstrong, until that minute Google’s Americas sales chief, as CEO and chairman.

Google, remember, owns 5% of AOL bought at the cost of a billion dollars to keep Microsoft out. The companies also have a search ad deal.

According to the formal announcement, Armstrong’s supposed to help Time Warner “determine the optimal structure for AOL.” Given his heritage it probably won’t have anything to do with Microsoft.

One wonders if Google is keeping Armstrong’s chair warm since Time Warner CEO Jeff Bewkes is thought to favor a spin-off.

Posted by rhall2091 at 6:44 pm

Google’s App Engine Goes Commercial

Google No Responses »
Mar 022009

Google is ready to start charging for its App Engine cloud platform.

For the 10 months it’s been in preview it’s been free to use but limited to 500MB of persistent storage and enough CPU, bandwidth and whatnot to support about five million page views a month.

On Tuesday Google said it was ready to follow through on its intention to offer additional computing resources for a price and allow apps to scale beyond its free quotas. It said it’s been its most requested feature.

However, it’s going to lower its free thresholds in 90 days, claiming it overestimated the resources developers needed to get started. It thinks the free resources will still support five million page views a month.

Under the new regime, it says developers can set a daily budget for their apps representing the maximum amount they’re willing to pay for computing resources each day. They allocate this budget across CPU, bandwidth, storage and e-mail, and they pay only for what their app consumes beyond the free thresholds – prorated “to the nearest penny,” it says.

Mind you there are still no service level agreements to reimburse users if App Engine goes down.

Google figures the resources paid for will scale to around 500 requests per second (qps) or more than 40 million queries a day, which is enough to handle traffic from being Slashdotted or Dugg. “In extreme cases,” it says, “(e.g. your application has been featured on Yahoo’s homepage), you can request additional CPU.”

Undercutting Amazon a trace, it’s proposing to charge 10 cents per CPU core hour and says the price covers the actual CPU time an application uses to process a given request, as well as the CPU used for any Datastore usage.

It’ll cost 10 cents per GB bandwidth incoming, 12 cents per GB bandwidth outgoing, which is supposed to cover traffic directly to/from users, traffic between the app and any external servers accessed using the URLFetch API, and data sent via the Email API.

It’ll also cost 15 cents per GB of data stored by the application a month and a thousandth of a cent per e-mail recipient for e-mails sent by the application.

Google warns users that they may notice an increase in the amount of data stored by their applications and listed in the Admin Console.

Seems data stored in the datastore incurs additional overhead, depending on the number of indexes, as well as the number (and size) of associated properties. It says this overhead can sometimes be significant and it’s been underreporting it.

So it’s doubling the free storage quota to 1GB.

Google wants the bills paid through its PayPal-like Checkout system (add VAT in the EU). And it says you can’t use multiple applications to avoid incurring fees.

There are reportedly 45,000 apps currently running on App Engine.

See http://googleappengine.blogspot.com/2009/02/new-grow-your-app-beyond-free-quotas.html. There’s a FAQ at http://code.google.com/appengine/kb/billing.html.

Posted by rhall2091 at 4:50 am

Gmail Inches Toward Offline Support

Cloud Computing, Google No Responses »
Jan 292009

Eighteen months or so after Google Gears debuted, a period of time scarred by several major Gmail outages, Google’s cloud-accessible-only e-mail is finally getting invested with offline support.

English-speaking US and UK Gmail users who want to test the new skill can now catch up with the rest of the world.

Google calls the feature “early experimental,” but then Gmail is still in beta two years after its general release.

Although Google has been using the widgetry internally “for quite a while,” it warns that there still may be “some kinks that haven’t been completely ironed out yet.” It’s looking for feedback.

Once the feature is turned on, Gmail uses Gears to download a local cache of the user’s mail. As long as a connection to the network is maintained, that cache is synchronized with Gmail’s servers. When Internet connection is lost, Gmail automatically switches to offline mode and uses the data stored on the user’s hard drive instead of sending the messages across the network. Any messages a user sends while offline go to his outbox and get sent when Gmail detects a connection. There’s a “flaky connection mode” for when the user’s on an unreliable or slow connection. Google says it uses the local cache as if you were disconnected, but still synchronizes your mail with the server in the background. It’s striving for the same user experience on- and offline.

Posted by rhall2091 at 8:27 pm
Older Entries Newer Entries
© 2012 LinuxGram Suffusion theme by Sayontan Sinha