Talk:Cloud computing

Cloud Storage

Added some information on it, with some comparison between the traditional way and how it is now. Bruce404 (talk) 19:36, 1 July 2008 (UTC)

• Good idea... need some good refs for this. samj (talk) 08:10, 17 September 2008 (UTC)

Adobe AIR? Silverlight?

Any comments on the suitability of Adobe AIR and Silverlight? I've thus far left them out because they aren't true standards (although some components of AIR have been released?)? samj (talk) 08:34, 1 August 2008 (UTC)

Anyone? samj (talk) 06:37, 2 August 2008 (UTC)

Ok I've done some more research myself:

• Microsoft Silverlight sounds very proprietary and this doesn't seem likely to change any time soon.
• Adobe AIR is a similar story, even if the Adobe Flex component is partly open - the SDK's MPL'd but the tools to view (Flash) and create the content will remain proprietary.

Accordingly I don't see any point in listing either amongst a swarm of other open standards but am still interested to hear about how others feel on this point. samj (talk) 07:19, 2 August 2008 (UTC)

--Potentially incorrect thoughts: Wasn't flash finally released as open source (if not completely free software)? That's why there are third-party flash video players now. Right?

Anyways, it's moot. AIR and Silverlight are just new/different forms of the Java Runtime Environment. They are basically VM programs that allow you to install the same program under different architectures. Think of Azureus, you need JavaRE installed first. Azureus speaks to Java, Java speaks to your computer. This is not cloud computing, it is a program running on top of a program environment running on your computer. Think of it like this: Flash is a web only, web based runtime environment. AIR/Silverlight are full functioning, computer based runtime environments.--deewhite

Not sure about that... Silverlight seems to be more a presentation layer (WPF)... —Preceding unsigned comment added by 194.205.195.34 (talk) 10:29, 23 October 2008 (UTC)

That could definitely be right. Silverlight doesn't seem to be as full functioning as AIR. However, they do have Silverlight for Linux (and Mac?). WPF is a subsystem of .NET & Silverlight is a subsystem of WPF. "Microsoft Silverlight is a web-based subset of WPF that enables Flash-like web and mobile applications with the same programming model as .NET applications. 3D features are not supported, but XPS and vector-based drawing are included." (From http://en.wikipedia.org/wiki/Windows_Presentation_Foundation) On that note, it still allows for the same application to be installed under different OS's without rebuilding the app for the OS specifically. Conversely, AIR is not a subset of anything, it is a standalone compatibility layer.--deewhite

Ok, thanks for this interesting analysis. Do we think it's on-topic for this article or not? I'm still unconvinced, though a generic comment about technologies which enable client-independent delivery (eg Java, AIR, Silverlight) and rich applications could be kosher. 86.68.200.71 (talk) 08:13, 19 November 2008 (UTC)

Is there a Rich Internet Applications article this one can point to? No point duplicating stuff. SteveLoughran (talk) 16:47, 20 November 2008 (UTC)

Yes, and I've recently given it an overhaul. -- samj _in^out 01:07, 22 January 2009 (UTC)

Cloud computing stubs

In the absence of a cloud computing stub and for want of somewhere better, please list cloud computing stubs/requests here for now (in alphabetical order):

• GoGrid
• Heroku
• Mosso (web host)

Security

Are there any legal issues related to the security of cloud computing? For example, what are the legal implications of the use of cloud computing to process sensitive information? 69.140.152.55 (talk) 19:10, 10 September 2008 (UTC)

• Yes, of course there are - that discussion is just starting to heat up now and it would probably be worth documenting here as consensus is reached. Thanks for the suggestion. samj (talk) 07:13, 12 September 2008 (UTC)

This was my first thought when I heard about this sort of system. It's not addressed here at all. Suggest considering a section, even if brief, regarding the privacy or lack thereof of one's information in a cloud computing system. Fraoch Dubh (talk) 00:53, 6 November 2008 (UTC)

Maybe a risks section? The issues are security -can you trust the host not to lose your data. Data Protection: EU rules etc, privacy -does your hosting service look in the data itself. Also, are the legal requirements for the police to access hosted data less than for data on private systems? SteveLoughran (talk) 16:46, 20 November 2008 (UTC)

Agreed. This should definitely be covered in the risks section. If consensus hasn't been reached, then the state of the art should be documented and then updated as it progresses. The legal jurisdiction of the vendor holding your data is also important because it limits recourse when security is breached as well. Here are two incidents relating to violation of medical privacy laws by someone in the outsourcing chain. While these were not a result of cloud computing, similar risks are entailed by cloud computing customers if their vendor allows data/instances to migrate across legal boundaries:

• Virginia VHA records at risk from subcontractor
• Pakistani transcriber threatens to release UCSF medical records

This second case out of San Francisco describes the risk that a single administrator might be able to block access to the cloud computing resources.

• Lawyer says client was protecting city's code

Ellison rains on the cloud

from the Register:

Ellison rains on the cloud

Bizarrely, our tip of the hat goes to America's second-best paid corporate executive - Larry Ellison who punctured the hot air on this year's biggest buzz-phrase: cloud computing. As with web services, service oriented architectures (SOAs), and Web 2.0 before it, "cloud" became abused by marketing drones, start ups, and middle-aged vendors desperate to sound relevant. "Cloud is complete gibberish," Ellison told his OpenWorld 2008 conference, noting - correctly - the phrase is being used to describe everything from Gmail to Salesforce.com's customer relationship management and platform. "What the hell is cloud computing? I don't understand what we'd do differently in light of cloud computing other than change the wording on some of our ads," Ellison said. "When is this idiocy going to stop?" Amen.--Espoo (talk) 01:22, 30 December 2008 (UTC)

Definition of Cloud Computing

The very first sentence of this article is "Cloud computing is Internet-based ("cloud") development and use of computer technology ("computing").", which is not correct - "cloud" in no way implies "internet". Go read a whole bunch of blogs on the subject and you'll realize that no-one can agree on a definition of Cloud Computing right now - whether its internet only, whether its end-user services eg. gmail or an app runtime environment eg. Amazon. My own personal interpretation of the term is more fuzzy - just being that the computation/storage is done remotely from the client by a large pool of computers, and that the load can move around between all of those computers transparently - this could include enterprize applications run in a private datacenter and accessed on a LAN. I also think that the source of the first few sentances of this article is rediculous - its an old article from 1999 (before anyone had through of Cloud Computing), and is about the reasons that people use a cloud-shape to represent the internet on diagrams (btw - people use cloud-shapes to represent LAN's, WAN's, SAN's, cellular-networks, etc. on diagrams) - it has nothing at all to do with cloud computing, and referencing it as a source for an article about cloud computing is just plain dumb. -- 192.197.128.19 (talk) 20:14, 7 January 2009 (UTC)

Well, this article gives this definition of Cloud Computing: Some analysts and vendors define cloud computing narrowly as an updated version of utility computing: basically virtual servers available over the Internet. Others go very broad, arguing anything you consume outside the firewall is "in the cloud," including conventional outsourcing.. That doesn't really sound like it is including intranet-based distributed computing. Also, the start of the article is trying to explain the origins of the term "cloud" in this context, so referring to network diagram conventions seems appropriate. Letdorf (talk) 16:19, 13 January 2009 (UTC).

Thanks, agreed - the origins of the 'cloud' are both important and relevant. It is also well accepted, except perhaps by purveyors of "I can't believe it's not cloud" systems, that cloud computing is intrinsically linked with 'Internet' - anything else is a poor approximation and today virtually all of the stuff being discussed is vaporware or confined to labs anyway while Google Apps (as one example of many) has over 10 million active users. -- samj _in^out 01:03, 22 January 2009 (UTC)

Internet based 'development.... ? This description does not make much sense, for most people "cloud based" means you can run applications in your browser that actually run on servers on the Internet, and can also save your data there. As such the introduction sentence is really crappy to say the least. It obfuscates what Cloud Computing is really about. Mahjongg (talk) 15:22, 12 February 2009 (UTC)

Agreed... see the ongoing discussion below. Letdorf (talk) 15:40, 12 February 2009 (UTC).

The vast majority of topics falling under the 'cloud computing' moniker are developer APIs, components (storage, database, queues, etc.), development platforms (azure, appengine, force.com, etc.), raw compute power (amazon ec2, gogrid, etc.), etc. This is what cloud computing "is really about" - "run[ning] applications in your browser" is called Software as a Service and is just part of the cloud computing landscape.

Furthermore this is derived from the defintion of "computing" which is "development and use of computer technology". -- samj _in^out 17:40, 14 February 2009 (UTC)

For the vast majority of people "cloud computing" has just a single meaning, and it has nothing to do with "development", that is just obfuscating the issue for most people. for the vast majority cloud computing means running your applications "on the cloud", and storing your data "in the cloud", or to quote "Cloud computing is an idea that what makes up your "personal computer" - your desktop, your apps, and your data, - can live and run on the Internet instead of on your hard drive.". That is much more direct than all the techno-babble, here, and much more relevant to the normal users. All the obscure technical distractions are just that distractions. Sooner or later the definition of cloud computing has to follow mainstream usage. Just like the Application software article, there wont be any talk about its development, API's or Libraries, but what a software application does, with "Cloud computing" it will be the same, people are not concerned how applications for cloud are developed, but what "Cloud computing" means for them, and what the immediate consequences are of its existence for them. For a "no nonsense" mainstream description of cloud computing you might want to watch this youTube video [1] Mahjongg (talk) 20:21, 15 February 2009 (UTC)

The term "cloud computing" usually refers to the "virtual hardware" layer where a vendor supplies virtualized hardware and some minimum set of software (for example Xen on ec² or Python on App Engine). Critically, the customer must then install, maintain and deploy their own software on-demand. A "cloud computing" customer still must employ their own software administrator(s) to manage the customer software.

In SaaS computing, on the other hand, the vendor employs the software administrator(s) to install, maintain and deploy all of the software. A SaaS vendor might possibly employ hardware administrators too depending on whether the SaaS vendor maintains their own hardware or uses someone else's "cloud".

Employment of software administrator's seems to be the Occam's razor that differentiates Cloud computing from SaaS. Web 2.0 implementations are deployed using SaaS and SaaS in turn is deployed using Cloud computing.

sn‾uǝɹɹɐʍɯ (talk) 01:48, 27 February 2009 (UTC)

Any references for these definitions? Letdorf (talk) 10:21, 27 February 2009 (UTC).

• Unindented

The vast majority of references to "cloud computing" are either high level passing references in the context of discussions about e.g. Google Apps or Salesforce.com or (more often than not) discussion about cloud computing "under the covers". A good example supporting this is that Salesforce.com has always been the number 1 example of SaaS, while "Force.com is Cloud Computing for the Enterprise" (a developer only platform). Microsoft's horse in the cloud computing race is Azure - another developer only technology, and Google App Engine falles squarely under the "cloud computing" umbrella while Google Apps could be better classed SaaS (even if it sits on top of a cloud computing architecture).

In any case the "developer" component is by no means exclusive to the "user" component - indeed the disputed phrase, "development and use of computer technology", should keep both camps happy so this looks a lot like argument for the sake of it. -- samj _in^out 15:01, 1 March 2009 (UTC)

A couple of references in response to Letdorf's request:

Cloud Computing Information Security and Privacy Advisory Board, Bill Whyman, ISI - International Strategy and Investing

IDC definition of Cloud Computing

Sourced from Tim O'Reilly and used in a presentation about Azure.

sn‾uǝɹɹɐʍɯ (talk) 19:44, 2 March 2009 (UTC)

Yes, the distinction between infrastructure and application is important, but the problem is that the terminology used in your references isn't consistent: you and ISI refer to "cloud computing" and "SaaS", IDC talk about "cloud computing" and "cloud services", O'Reilly talks about "utility computing", "PaaS" and "cloud-based applications". Also I think it's fair to say there's more to cloud infrastructure that just virtualization.

I'm not sure what SamJ means by "developer only technology" - Windows Azure isn't a development environment, it's a cloud services platform! Such platforms may not be directly visible to users (in much the same way as the internals of an OS, or any other "back-end" software, aren't) but it would be misleading to refer to them as "developer" technologies. Letdorf (talk) 11:28, 3 March 2009 (UTC).

Take another look - Azure is of no interest whatsoever to users but the applications which use its components are; if one doesn't look at the components there's very little to say on the subject. It's very much a view of cloud computing as an "operating environment" of sorts, which is an increasingly popular view. -- samj _in^out 19:19, 4 March 2009 (UTC)

Letdorf, yes, cloud computing ontology is inconsistent and an open area of discussion/research. Here are some references (along with some great diagrams) that are sorting out the naming:

* The most well researched appears to be Toward a Unified Ontology of Cloud Computing - Youseff, Butrico, DaSilva

* Cristofer Hoff has produced this very nice and well detailed diagram: Cloud Taxonomy & Ontology - Draft 1.4 - Hoff

* And finally Kent Langley created a very approachable diagram in April 2008, but it does not include IaaS/PaaS layering under SaaS: Cloud Computing: Get Your Head in the Clouds

sn‾uǝɹɹɐʍɯ (talk) 21:54, 5 March 2009 (UTC)

The reason these diagrams haven't (yet) appeared here is, as you say, they're original research. That and they're too technical. Accordingly we've gone as far as we can without crossing over into OR by breaking what we know is cloud computing into half a dozen categories. Until we have something universally accepted (like the OSI Stack) there's not much more we can do - I'm expecting to see something like this emerge from one of the various standards efforts in 6-12 months and be adopted within 12-24. -- samj _in^out 22:13, 5 March 2009 (UTC)

No original research applies to using Wikipedia to publish original research, but the Youseff2008 paper was published as part of Grid Computing Environments (GCE08). I've updated the citation to include the publication info and added a reference for the presentation version. The other two diagrams (Langley, Hoff) might violate WP:SPS (investigation needed), however, other references you've already allowed in this article are of similar quality and so I figured I would mention them anyway. The diagrams in the article might also fall under the WP:SPS policy too. Oh and on the complexity - the Youseff2008 diagram is pretty simple as is Langley's; I agree that the Hoff diagram seems pretty complicated at first glance, but the colors help. sn‾uǝɹɹɐʍɯ (talk) 23:23, 5 March 2009 (UTC)

WP:SPS usually applies to sources written by the article subject - e.g. referencing a company web site, press release or interview in the same company's article to justify a statement like "XYZ is the biggest, baddest company in the industry". It's true that WP:OR applies to images which is exactly why we haven't been too creative as yet. As I say, when there's a consensus ontology we'll accept it but I think it's a while away yet... a lot of even the academic reports have been heavily criticised by the industry. The 3-layer SaaS, PaaS, IaaS option has got some traction but it doesn't go far enough while the others go too far. -- samj _in^out 02:06, 6 March 2009 (UTC)

Oh and by the way, here's the policy. -- samj _in^out 02:08, 6 March 2009 (UTC)

I've been [re]reviewing these references today and it's interesting that IBM should be contributing to the paper given that the resulting diagram is rather incoherent (CaaS? DaaS? !?!?) and that they are actually using a stack almost identical to the one depicted in the article. -- samj _in^out 14:27, 6 March 2009 (UTC)

What's behind "The Cloud"?

Unfortunately, "the cloud" is currently used for everything that's offered in the internet, hiding the real philosophy behind this term. "The Cloud" is about SOA and interoperability. There is a must-read interview with Russ Daniels, CTO and VP of HP's Cloud Services Strategy at arstechnica. He really captures the ideology behind the cloud:

"One of my favorite examples is, I have a car that has an entertainment system that supports Bluetooth so I can use it as a hands-free headset. But to be able to do voice dialing, I have to load the contact list from my phone into the car. [...]

But the whole idea is flawed because my phone's not authoritative for my contact information. The phone has a local cache of information that it gets from Outlook. But Outlook's not authoritative for my contact information, neither is Gmail, neither is my Vonage, neither is Grand Central, neither is the six or eight other things that I have in my life that think they are.

What none of them do is the simple thing of, "tell me the URL for your contact service." Additionally, it has to be a service, not a repository, because in fact the contact information that's relevant for me includes the global address list for HP, and I have to be able to have that invoked... I can't replicate that data and keep it synchronized, so I need to be able to use a federation model behind this single endpoint to answer those kinds of queries."

That meaning of "The Cloud" should be in the article. There's also a lot of other stuff in the interview that should be included in this article. Tdanecker (talk) 23:13, 13 January 2009 (UTC)

While a good example this is too technical for the audience. -- samj _in^out 13:26, 14 January 2009 (UTC)

The example doesn't have to be put 1:1 in the article, but I'm missing the general ideology of the cloud. The article focuses mostly on the "there's a service in the cloud that can be used everywhere" part, and less on the "devices have no local state but only cache things" part. That may be due to the current hype where every service offered in the internet is advertised as "cloud computing". The article should not represent this marketing strategies but the real idea behind the cloud (stateless devices/thin clients and their use of such services). Tdanecker (talk) 15:53, 14 January 2009 (UTC)

We could certainly talk more about "thin" clients like netbooks etc... -- samj _in^out 04:49, 28 February 2009 (UTC)

Intro

So, which is the better first sentence for this article?

"Cloud computing is a term which is used to refer to the use of scalable, real-time, Internet-based information technology services and resources."

or:

"Cloud computing is Internet ("cloud") based development and use of computer technology ("computing")."

Letdorf (talk) 10:31, 16 January 2009 (UTC).

The first one is about SOA+utility computing, and the second one doesn't say anything at all... "The key distinction [between cloud and utility] ends up being around changes in the way you architect software, and changes in the way you architect the work being done." --Russ Daniels Tdanecker (talk) 11:15, 20 January 2009 (UTC)

Okay.... so Mr Daniels says cloud computing is like utility computing, but "architected differently". Should that affect the introductory sentence in this article? Letdorf (talk) 13:27, 20 January 2009 (UTC).

The original (second) definition is at least accurate, if lacking in precision. The proposed replacement is incorrect in that cloud computing does not have to be scalable (but it often is) nor real-time (think batch jobs). A good deal of time and effort went into boiling the many definitions into one compatible consensus view and this was the (generally accepted) result.

Here are a handful of references that "based+development+and+use+of+computer+technology" approve of this definition:

• Cloud Computing Begins to Gain Traction on Wall Street
• The difference between Cloud Computing and SaaS
• Is the Enterprise Ready for Cloud Computing?
• Soonr brings cloud computing to the iPhone
• Cloud Computing Technology Providers and Contributors in 2008-2009
• http://www.cyop.net/cloud_computing.html
• Windows Azure: A clear perspective on cloud computing

Please talk first rather than reverting reverts in future, but feel free to expand on the base definition using the best references you can find (extracting signal from the noise can be difficult). -- samj _in^out 20:13, 20 January 2009 (UTC)

Oops - I just realised the definition wasn't referenced... it is now. -- samj _in^out 20:19, 20 January 2009 (UTC)

I'm not sure what you're saying here - are you claiming that the validity of a definition given in a WP article can be measured by the number of journalists who are lazy enough to cut'n'paste from said article?! I'm sorry, but the phrase Internet ("cloud") based development and use of computer technology ("computing") is far too vague a definition, IMHO. Lots of "computer technology" is developed and/or used via the Internet, but little of it could be considered "cloud computing", even in the most general sense. For instance, is downloading a file by FTP or accessing a remote CVS repository "cloud computing"? Both would fit that definition.

This is a recurrent discussion, but yes, from a user-centric PoV if they can consume raw (FTP, eg CloudNAS) or 'enhanced' (CVS) storage without having to concern themselves with the inner workings then sure it fits; finally we're starting to look at computing from the user's point of view rather than having technology for the sake of technology. -- samj _in^out 12:04, 21 January 2009 (UTC)

When I mentioned FTP or CVS, I meant the kinds of things people were using the Internet for more than 20 years ago; before "Web 1.0", never mind "Web 2.0". That was still "Internet-based development and use of computer technology". Letdorf (talk) 00:28, 24 January 2009 (UTC).

The "scalable" and "real-time" qualifiers came from the InfoWorld article I cited, which seemed like a pretty good introduction to the subject Letdorf (talk) 21:28, 20 January 2009 (UTC).

If it's that good a resource then add it to the external links, and refine the base definition accordingly. I maintain that neither real-time nor scalable are requisites for cloud computing but that's not to stop them being used as examples. -- samj _in^out 12:04, 21 January 2009 (UTC)

Furthermore, of the two references now given for the current first sentence, one explicitly quotes this article(!), and the other is dated January 6 this year, so is highly likely to have been a cut'n'paste from this article. Indirect self-references are obviously not valid references! (see WP:RS). Letdorf (talk) 21:46, 20 January 2009 (UTC).

The articles were selected because they assessed/validated the definition rather than just regurgitating it. In any case I've added a whitepaper on the subject which goes into more detail. Given the InfoWorld article predates most of the discussion around the definition, more recent articles better reflect the consensus of the community. -- samj _in^out 12:04, 21 January 2009 (UTC)

So now there are three refs for the opening statement; they're all dated December 2008 or January 2009, contain phrases that are suspiciously similar to this article around the same dates and hence are almost certainly indirect self-references. These are fundamentally wrong by the rules of WP! Letdorf (talk) 00:28, 24 January 2009 (UTC)

Ok so I've gone over the article and references again and while not an absolute requirement I agree that some reference to "agile, scalable infrastructure", real-time scalability, etc. is useful so I've expanded the intro paragraph accordingly. I hope you are more satisfied with the result now. -- samj _in^out 12:31, 21 January 2009 (UTC)

Okay, to recap, in the absence of any further objection, I'm proposing deleting the first sentence (apart from the obvious "Cloud computing is...") on the grounds that (a) it doesn't really add anything to the current first paragraph and (b) as I say above, the references given appear to be indirect self-references. Letdorf (talk) 17:17, 28 January 2009 (UTC).

The first sentence defines cloud computing; the remainder expands on that definition. You may not like that it's a broad definition, but at least it is accurate; your proposal (real-time, scalable, etc.) is clearly and demonstrably not. As I explained to you above, the three references are not self references but I'll copy the relevant parts here so you can see for yourself:

• "Here's why cloud computing — loosely defined as Internet-based development and use of computer technology — has a future and the likely reason why Ellison doesn't like it: Cloud-based services so far are cheaper than traditional IT products, such as large in-house databases."

• "Answers to the most complicated computing tasks are found in the clouds. Not the clouds in the sky that shower rain but clouds that deliver a super computing power on the internet. Cloud computing is the development and use of computer technology using the internet."

• "Cloud computing refers to the use of Internet-based (i.e. Cloud) computer technology for a variety of services (including storage capacity, processing power, business applications or components)."

The intro has been discussed to death already and we've finally found something that's been stable for months, except that every once in a while someone wanders in and insists on foisting their narrow view on the rest of us. Enough already. -- samj _in^out 00:58, 1 February 2009 (UTC)

FWIW, the intro has been discussed here and here and here and here and here so you can see why I'm bored already... no offense. -- samj _in^out 01:03, 1 February 2009 (UTC)

I'm not really interested in foisting any narrow view on anybody; I'm quite willing to debate whether "real-time" or "scalable" or anything else are defining attributes of cloud computing. What I am interested in is trying to improve an opening sentence that gives a near-meaningless definition which doesn't distinguish cloud computing from pretty much anything you would use the Internet for, and a bunch of supporting references which appear to me to have used this very article as their own source.

Having said that, the somewhat different definition you quote from page 4 of the Micro Focus white paper is rather more specific and does seem at least partly original - IMHO, this would be a better one to use. Letdorf (talk) 12:45, 2 February 2009 (UTC).

Wait, you guys cannot be serious - this is somehow better than what was there before? Software Applications? Mobile Enterprise? WTF?:

Cloud computing is the use of computer data storage, central processing units, business applications, web applications or software applications, via the Internet, as well as software development tools and associated software architecture components like message queues. It is a style of computing in which typically real-time, scalable resources are provided "as a service" over the Internet to users who need not have knowledge of, expertise in, or control over the technology infrastructure ("in the cloud") that supports them. Mobile enterprise is cloud computing for business.

I changed the first sentence to a very light paraphrasing of the third quote User:SamJohnston has given above, which I considered to be an improvement on the previous sentence - see here. User:CloudComputing then attempted to expand this and give another ref shortly afterwards. I assumed SamJohnston would have no objections to me using the quote he himself provided on this talk page, but I agree CloudComputing's additions [2] probably don't clarify matters much. Letdorf (talk) 17:17, 8 February 2009 (UTC).

It seems there were a few different accounts involved, but the problem is that as soon as anyone starts listing anything in this article those of us who maintain it end up spending half our lives reverting additions - within a day or two that list would have grown to 10 or 15 items. You show me a more precise definition than 'internet based development and use of computer technology' and I'll show you an example of something that doesn't fit. -- samj _in^out 00:06, 9 February 2009 (UTC)

Okay, let's try that again without the list:

Cloud computing refers to the use of Internet-based (i.e. Cloud) computer technology for a variety of services.

This is marginally more precise and is copied verbatim from the quote you gave above. By the way, ...spending half our lives reverting additions pretty much sums up my experience of being a Wikipedian too. However, making one editor's life easier isn't a good justification for attempting to stymie other editors' efforts to improve an article. WP:OWN and WP:NVC have some words of wisdom on this. Letdorf (talk) 14:13, 9 February 2009 (UTC).

I strongly agree with Letdorf. "If you do not want your ideas (for article organization, categorization, style, standards, etc.) challenged or developed by others, then do not submit them" --Emre Kenci (talk) 16:05, 12 February 2009 (UTC)

Did you actually read the thread? Anyway I see you've broken the intro again this afternoon - so long as we're talking minutiae about appropriate use of parenthesis and square quotes how about you tell me where it says you can have parentheses without leading spaces? -- samj _in^out 23:04, 16 February 2009 (UTC)

You're right it does not say that anywhere so instead of calling you an abuser like you did, I corrected it. I didnt read the whole threat but I read this part: "WP:OWN and WP:NVC have some words of wisdom on this." by Letdorf. And I still strongly agree with that. --Emre Kenci (talk) 12:14, 17 February 2009 (UTC)

Are you guys having fun yet? The intro's a !#$%@( mess again. -- samj _in^out 03:38, 22 February 2009 (UTC)

The current first sentence of the intro comes from one of the quotes you have cited above, in support of the previous version. Surely you're not objecting to that? I could probably come up with something better, but I decided to compromise, given your intransigence over previous attempts to improve the intro. By the way, have you read WP:CIVIL? Letdorf (talk) 00:20, 23 February 2009 (UTC).

Never mind, I already cleaned it up -- samj _in^out 17:00, 24 February 2009 (UTC)

I've been pondering over the WP:LEAD and have discussed it with a few colleagues... the previous version (which was in place without incident for months) is more informative than the current "for a variety of services" wording and as such has been replaced. -- samj _in^out 04:45, 28 February 2009 (UTC)

The place to discuss making changes to this article is here. Consensus in Wikipedia is built among WP editors, not among one editor and his mates! Letdorf (talk) 18:17, 28 February 2009 (UTC).

These "mates" happen to include some of the most active people in the cloud computing commmunity and until such time as I can convince them to get involved I am their only representation here. In any case, proposing such a controversial edit to a stable article and then implementing it in the absence of consensus is not on. -- samj _in^out 15:08, 1 March 2009 (UTC)

Regardless of the activities of your acquaintances, this is at best original research. Controversial? This time I deliberately used a quote (verbatim) and a source you had given previously in this discussion, in an attempt to be as un-controversial as possible! It's obvious you have no real interest in building consensus here. Letdorf (talk) 15:26, 1 March 2009 (UTC).

It's well sourced so hardly original research and there have been two people outside of us contribute to this discussion over six weeks - one wanted to talk about quotes and parenthesis and the other provided a quote that talked about software architecture (e.g. development). I see that as an indication that it's a non-issue for most people and that we would be better to focus our energies elsewhere. Status quo reflects the views of both users and developers while suggested alternatives ignore the latter. -- samj _in^out 05:24, 2 March 2009 (UTC)

I seem to have inadvertently stepped into a sh*t storm by removing the first sentence. My rationale for removing it was that it's a completely vacuous sentence, and I'm not sure it's even grammatically correct. In any case, it starts out by "Cloud computing is ..." leading me to believe that what follows is a definition of what cloud computing is. Instead I'm told that it is something that is: 1) "Internet based", and 2) a "use of computer technology". That definition matches about a million different terms and tells me nothing of value. It also sets the whole tone of the article as a lackluster soup of marketing terms. I was reverted as "per talk page", but I fail to see any consensus for it here. Instead I see a number of people expressing concerns over the intro in general and the first sentence in particular, and one editor who is disregarding everyone else's arguments.--130.233.154.94 (talk) 12:44, 23 March 2009 (UTC)

Yes, as you can see, I've been trying to persuade User:SamJohnston for some time now with the same argument, but to no avail. A clear case of WP:OWN it seems to me. Letdorf (talk) 13:00, 23 March 2009 (UTC).

Chrome a client?

How is Chrome a client? It's just another webbrowser. If Chrome is listen, then why not IE, Firefox and Opera aswell? It seems to me that people are just trying to draw attention towards chrome. —Preceding unsigned comment added by 212.178.206.30 (talk) 12:13, 23 January 2009 (UTC)

See archives. Chrome is quite a different beast from the other browsers architecturally, but with FF3 here and IE8 around the corner it's less clear that it's alone. -- samj _in^out 00:09, 24 January 2009 (UTC)

The first real cloud

There is some disturbtion about a link I added. Plese take a look at the website I linked and feel free to let me know if the link should be added to the page.

• "Introduction into the first real live "virtual cloud" named virtual storm. Jan, 20 2009"

Virtual Storm is a website that explains what we feel to be the first actual up and running Virtual Cloud. We have separated the client from the hardware using Vmware ESX to virtualise the images. We added a brand new product on the market to seperate the software form the images. In our Virtual storm the image is only 1.2 GB and it connects to a centralised repository where all the applications reside. Every client uses a patented piece of software that redirects the software inside the image when the user has the rights to use the application. Regards Erik Keep Wikipedia clean, but don't leave it emptyErikw11 (talk) 20:07, 26 January 2009 (UTC)

I'd say that if it's notable enough to hold its own article then it may be worth a mention... otherwise there's plenty of similar projects out there. -- samj _in^out 01:04, 1 February 2009 (UTC)

'The first real cloud' is a bit of a stretch BTW. -- samj _in^out 01:05, 1 February 2009 (UTC)

Is Cloud Computing is a type of business computing?

Do people agree that Cloud Computing is a different style of business information management systems. Because looking at the technologies "as a service", they seems like a new model of Internet Hosting Paradigm that derived from distributions.

Paragraph 2 says its utilize Web 2.0. It somewhat support the idea of retail. Examples are Blogs and SMS are the same thing (in a way). They are different "form" of communication communicating through different "mediums."

SMS is through "mobile" Blogs is through "web" Isn't this the concept of retailing?

I am not sure is my idea wrong? Does anyone have any insights for a better guidance? --142.22.48.251 (talk) 04:30, 27 January 2009 (UTC)

If you can find reliable sources to back up your hypothesis, then go ahead and cite them. Otherwise, this counts as original research, which is against the rules of WP. Letdorf (talk) 10:18, 27 January 2009 (UTC).

I tend to agree with Letdorf... sounds like OR, or at least something that belongs more in the Web 2.0 article itself. We could however probably afford to talk more about the effects of migrating capex to opex, etc. but I'm not sure that's really what you had in mind. Anyway web 2.0 is more about usages, cloud about resources. -- samj _in^out 05:04, 29 January 2009 (UTC)

Sun Microsystems?

In March 2006, Sun announced a Grid computing service that was very cloud-like and in many ways a precursor to what's happening in cloud today. Would this be worthy of mention in the History section? Russcastronovo (talk) 19:47, 28 January 2009 (UTC)

Possibly, though if you're talking about network.com, that was recently pulled to undergo some sort of grid->cloud transition. -- samj _in^out 05:00, 29 January 2009 (UTC)

The service sun provided was indeed at network.com. It was a $1 per CPU per hour offering and wasn't particularly successful. It did, however, preceed Amazon's offering by a year or so. Would that make it worthy of a brief mention? 192.18.43.225 (talk) 21:01, 29 January 2009 (UTC)

I don't think so, at least not until it reappears (at which time it may well give Amazon a run for their money). -- samj _in^out 00:44, 1 February 2009 (UTC)

I think Network.com is certainly worth mentioning as one of the first cloud vendors in a historic sense at least as a pioneer. sn‾uǝɹɹɐʍɯ (talk) 03:57, 6 March 2009 (UTC)

It was there when the system was live but not they've pulled it, presumably to move to xVM, the link went too. I don't feel too strongly about that either way. -- samj _in^out 04:37, 6 March 2009 (UTC)

Why do they have to keep coming up with buzzwords that

pretty much mean something that is already there but they want to give it a marketing boost, and an encyclopaedia shouldn't pander to that... —Preceding unsigned comment added by 91.132.224.196 (talk) 05:20, 5 February 2009 (UTC)

I also hate the word "Cloud Storage". It's nothing more than online storage and we also had such stuff previously (without using the word "cloud"). And why has this article has to be based on those marketing terms? IMO, this article should be based on the new ideology of "cloud computing" and those whole hype/marketing-stuff should be mentioned in one small paragraph. Unfortunatly it's difficult to see what "cloud computing" really is because all those marketing/hype-stuff generates so much noise which sadly gets reflected everywhere (including a lot of professional magazine articles). If someone would like to meet my request of overhauling the whole article, I can only recommend reading the arstechnica interview with Russ Daniels (as I've already mentioned twice in this discussing): Part 1 Part 2 Tdanecker (talk) 22:08, 14 February 2009 (UTC)

I also hate the word "Cloud Storage"...my request of overhauling the whole article... hmm that's a bit throwing the baby out with the bathwater don't you think? And based on an opinion piece from a box maker that (unsurprisingly) starts off by equating the cloud to a datacenter and then dives straight into a whole lot of detailed discussion about parallel programming and the like? Interesting, yes, for us programmers, but hardly encylopedic.

Don't stress too much about the "cloud" moniker - it'll fade into the background when everyone's doing it and we'll be back to calling it "computing" again. -- samj _in^out 03:53, 24 February 2009 (UTC)

I hope so. This whole article reads like some sort of marketing piece - 'scalability' is all very good, but what does it mean in practice? 86.135.68.65 (talk) 08:30, 24 February 2009 (UTC)

It means you don't have to worry about the details. -- samj _in^out 04:48, 28 February 2009 (UTC)

Criticism

Ok so it seems we get to have the 'criticism' discussion again. This section is a crap magnet and was previously removed with good reason - Stallman's view is already in the article too: "Richard Stallman, founder of the Free Software Foundation, believes that cloud computing endangers liberties because users sacrifice their privacy and personal data to a third party.".

There has been vocal critisism for cloud computing both for its role as a marketing term with little substance or novelty, and in terms of the privacy issues involved with it. GNU pioneer Richard Stallman had this to say: "It's stupidity. It's worse than stupidity: it's a marketing hype campaign."

Other major figures of the computer industry agreed, with Larry Ellison of Oracle Corporation commenting that: "The interesting thing about cloud computing is that we've redefined cloud computing to include everything that we already do. I can't think of anything that isn't cloud computing with all of these announcements. The computer industry is the only industry that is more fashion-driven than women's fashion. Maybe I'm an idiot, but I have no idea what anyone is talking about. What is it? It's complete gibberish. It's insane. When is this idiocy going to stop?"

It's worth mentioning that the comments from both RMS and Ellison don't necessarily reflect the views of the community or Oracle respectively, and that oracle has since launched the 'Oracle Cloud Computing Center':

Continuing its pioneering role in shaping enterprise computing, Oracle is pleased to introduce new offerings that allow enterprises to benefit from the developments taking place in the area of Cloud Computing. As a part of our initial offering, Oracle has partnered with Amazon Web Services (AWS) environment to offer the following products and services:

Deploy Oracle Software in the Cloud Backup Oracle Database in the Cloud

These offerings may be extended to other Cloud platforms in the future.

Please integrate legitimate criticisms into e.g. the key characteristics section. -- samj _in^out 13:29, 5 February 2009 (UTC)

Removing sourced information this way is rarely a good idea. While the information gives what appears to be undue weight to these opinions, removing them completely is just as problematic. Can the information be incorporated more sparingly into other sections of the article? --Ronz (talk) 15:57, 5 February 2009 (UTC)

Agreed. Tell that to these guys. Anyway the RMS comment was already incorporated into the article before the section was added and the Ellison comment was largely BS when it was said and now not at all reflective of the company's position on cloud computing - people change their minds you know :)

Anyway my main problem is that sections like this instantly attract rubbish in much the same way as the 'new and improved' intro did. I'll integrate it when I have a spare second (unless you care to do it yourself) but life calls right now. -- samj _in^out 13:08, 7 February 2009 (UTC)

The Manual of Style doesn't like them either: {{criticism section}} -- samj _in^out 04:54, 6 March 2009 (UTC)

Open source is required is a bit strong

"Open standards and open source software are also critical to the growth of cloud computing.[13]" is stated in the article at this time. While I do like and uses open source software, and think their are several benefits to that approach. It is by no means a requirement, except in a case where a customers has their own requirement to use open source infrastructure. Also the citation is to a blog where only an opinion that it is required, by no means a strong reference. 209.30.228.224 (talk) 04:14, 22 February 2009 (UTC)

Are you aware of any public cloud infrastructure that does not use open source and/or open standards? -- samj _in^out 18:31, 22 February 2009 (UTC)

Many large, retail clouds run on proprietary/closed source software: Amazon's cloud, IBM's Blue Cloud (IBM) and Microsoft's Azure (MSFT) for example all build their clouds with proprietary code; SaaS vendors Google Apps (GOOG) and SalesForce.com (CRM) similarly power their offerings with proprietary software. Even vendors who started with open software frequently, and sadly, do not release their modifications back to the community and thus are effectively using proprietary software. (GPLv3 section 13 specifically addresses the wide spread exploitation of this "server hole" in GPLv2). Given the widespread use of proprietary software, how about changing the wording to say "Open standards and open source software have been critical to the growth...".

"Open standards" (HTTP, SSL, TCP/IP), however, remain the backbone of cloud computing, public and private. sn‾uǝɹɹɐʍɯ (talk) 23:48, 2 March 2009 (UTC)

It's certainly more reasonable to say that all cloud computing systems are based on Internet/Open standards but you're right, it's less sure for open source software. Amazon is based on Xen (open source), Blue Cloud too (+linux, hadoop, etc.). SalesForce moved to Linux and Google's always run on it. OTOH all of them have proprietary components and it could be argued that not enough is being done in terms of AGPL etc. -- samj _in^out 07:20, 3 March 2009 (UTC)

Architecture

The Architecture section is not very informative, and seems to concentrate on introducing jargon more than on explaining architectural concepts. The reference to "Unix Philosophy" is particularly unhelpful. Rablewis (talk) 00:11, 28 February 2009 (UTC)

Do you have any better suggestions as to how better to explain the component nature of cloud computing architectures? This was the best (well known) analogy that we could come up with and I personally think it does a good job of illustrating the concept. As unix has cat, grep, sed, awk etc. talking over pipes, cloud computing has storage, compute, queues, databases, etc. talking over http. -- samj _in^out 04:47, 28 February 2009 (UTC)

Replace "Internet" with more general word "network" in lede?

Many vendors provide "private cloud" software and other organizations have developed their own private clouds. These resources meet the definition of cloud computing except that they are not deployed on the Internet, but instead use a private network. The lede would be well served by replacing the word "Internet" with the more general word "network". Thoughts? sn‾uǝɹɹɐʍɯ (talk) 23:19, 2 March 2009 (UTC)

The vast majority of references explicitly use the term "Internet" and some of them even explicitly exclude the possibility of internal/private clouds. I for one would strongly oppose such a change and am pretty sure this has been discussed (and rejected) previously. These things are essentially just virtualisation rebadged as "cloud" and they short-sell cloud computing - indeed many of the key features of cloud computing (perimeter free architecture, worldwide access, no peak load engineering, multi-tenant architecture/economies of scale, etc.) are simply impossible to replicate. That's not to stop someone going and writing about "private cloud" or "internal cloud" in another article of course, but previous attempts to do so weren't so successful - it probably doesn't help that virtually everyone pushing this terminology happens to have something to sell and a lot to lose from the success of cloud computing. -- samj _in^out 23:32, 2 March 2009 (UTC)

Both linguistically and based on the key characteristics, "internal clouds" or "private clouds" seem still to be "clouds". Private clouds provide all of the key characteristics listed in the article. "Perimeter-free architecture" and "worldwide access" are characteristics not listed in the article; "worldwide access" is a feature of most private clouds (usually achieved using VPN software) but I'm not familiar with the term "Perimeter-free architecture". A pointer to the previous discussion of this topic would be appreciated. Thanks! sn‾uǝɹɹɐʍɯ (talk) 00:11, 3 March 2009 (UTC)

Here and here some previous discussions on the topic. The point is that users do not connect to a service endpoint as such but "the cloud" (which usually means being routed to the closest next gen data center). There's some talk about security in the cloud here.

Some analysts and vendors define cloud computing narrowly as an updated version of utility computing: basically virtual servers available over the Internet. Others go very broad, arguing anything you consume outside the firewall is "in the cloud", including conventional outsourcing.

Here is a typical quote on the subject and there are many like it, like this:

The desktop is dead. Welcome to the Internet cloud, where massive facilities across the globe will store all the data you'll ever use. George Gilder on the dawning of the petabyte age... Thus, the new computing architecture scales across Earth's surface. Ironically, this emerging architecture is interlinked by the very technology that was supposed to be Big Computing's downfall: the Internet

And the following wording from [this] Pew Internet report:

Some 69% of online Americans use webmail services, store data online, or use software programs such as word processing applications whose functionality is located on the web. Online users who take advantage of cloud applications say they like the convenience of having access to data and applications from any Web-connected device.

These are just a few examples from a quick search - indeed there are ~15,000,000 hits for "cloud computing" vs ~15,000 for for "internal cloud" and 4 of the top 5 hits for "private cloud" refer to this rocking bed. Here is a more specific handling of "private cloud":

That's an oxymoron since cloud computing, by definition, happens outside of the corporate data center, but it's the technology that's important here, not the semantics.

Maybe a footnote or small section on the topic representing this minority view would suffice... -- samj _in^out 07:05, 3 March 2009 (UTC)

The original suggestion was to replace the word "Internet" with "network" in the lede. The other terms (private/internal cloud) seem distracting. An archived disucussion does point out ZDNet's succinct and approachable definition of cloud computing which says in part ``Computing "in the cloud" may refer to a company's own network, but often refers to the Internet...". That seems like a more objective introduction for this article that includes both Internet and private network deployments. While many examples of cloud computing can be found on the Internet, many others are found within private networks.

The suggestion that cloud computing must "happen outside of the corporate data center" seems overly restrictive. For instance, Google's employees use Google Apps to run the business. They are still using cloud computing even though the computers are inside of Google's own corporate data center. The Internet is just one network, albeit a very large one, where cloud computing can be used. Replacing the word "Internet" with "network" in the current lede will improve this article. sn‾uǝɹɹɐʍɯ (talk) 18:48, 3 March 2009 (UTC)

I think Google might be something of a special case here! Cloud computing technologies may be applicable to intranets (in much the same way as TCP/IP is applicable to LANs as well as WANs), but, IMHO, "cloud computing", as a paradigm, currently usually implies services provided via the Internet by external service providers. However, if reliable sources conflict on this point, then there is no right answer. Letdorf (talk) 21:04, 3 March 2009 (UTC).

For all intents and purposes Google corporate is just another Google Enterprise customer - they're remote from the data centers just like everyone else, and may well pay for access just like everyone else too. For now I'd change your statement above as follows: "While virtually all examples of cloud computing can be found on the Internet, some others are found within private labs".

The subject has popped up again a bit with the days-old release of VMware's vCloud... let's give it a bit and see where it goes from here. Oh, and by the way - what did you mean by your latest edits: "The services are accessible anywhere that the required networking infrastructure." -- samj _in^out 00:10, 4 March 2009 (UTC)

The special case seems more like the common case. Companies using cloud computing internally include Intel, Pixar, Boeing, Airbus, GM, Ford, Chrysler, Honda, Toyota, Caterpillar, NASA, NOAA, University of California, Oracle, Apple, Nike, Adidas, CERN and many more. Cloud vendors use their own clouds too including Yahoo!, Amazon, IBM, Microsoft, Salesforce, EMC/VMWare and Netsuite. All use large internal clouds that are flexibly applied in a scalable fashion and accessed remotely using standardized protocols on the company's internal network by a variety of groups using web browsers within the company. While their infrastructure is not accessible from the public Internet, the employees still benefit from cloud computing. The lede should include cloud computing regardless of whether its on an internal network or the Internet. Replacing "Internet" in the lede with "network" would capture that. I do think that the "Internet" is the largest and most commonly referenced of all networks.

Re: my latest edits: they suffered from an "edito" (a sub-species of typo) and I've repaired it. Pre-edit mentioned worldwide, but network access is required and that only reaches about a quarter of the world's population. sn‾uǝɹɹɐʍɯ (talk) 02:32, 4 March 2009 (UTC)

Sorry I can't accept your assertion that customers of a virtualisation vendor who has just announced some tech bearing the "cloud computing" badge suddenly become "cloud computing" customers, especially when exactly zero of them are listed in the provided reference and I can't find any links between most of them and cloud computing (after dropping the cloud vendors) - this looks like a vendor prospects/pilot list.

It is questionable anyway that this use case goes even close to meeting the existing definitions (off-site, third party, accessed via internet, minimal skills required, web/web service based, minimal capex, reduced opex, 'infinite' scalability, etc.). Granted to some extent such vendors will be able to co-opt the definition (see updates in grey) but that doesn't mean we have to treat it as anything other than an alternative PoV or indeed fringe theory (e.g. "ideas that depart significantly from the prevailing or mainstream view"), especially when most of the noise is coming from said vendors whose existing businesses are threatened by the new model.

It's also questionable as to whether cloud vendors "eating their own dogfood" is at all notable - the car manufacturer down the road happens to have one brand in its employee car park too and guess who Fedex use to send packages? (incidentally the examples you cite *are* Internet accessible, and indeed accessed via the Internet). In any case, would your proposed redefinition be any different from, say, client-server? Letdorf and I have already been working towards a sensible definition that keeps all players happy and it's not so easy.

Your own unreliable sources that you used to equate "network computing" (an Oracle-ism) to "cloud computing" yesterday also concur with the mainstream view:

"The term ‘cloud computing’ encompasses many areas of tech, including software as a service, a software distribution method pioneered by Salesforce.com about a decade ago. It also includes newer avenues such as hardware as a service, a way to order storage and server capacity on demand from Amazon and others. What all these cloud computing services have in common, though, is that they’re all delivered over the Internet, on demand, from massive data centers."

Given your insistence despite protests from two editors, consensus from various earlier discussions, lack of consensus here and my suggestion to "see where it goes", now would be a very good time for you to reveal any potential conflict of interest. I may have found some reliable sources to justify a mention but don't forget the law of unexpected consequences. -- samj _in^out 09:09, 4 March 2009 (UTC)

Conflict of Interest

The result of this discussion was that User:Mwarren_us eventually admitted that they are an Oracle employee when called to disclose and created a new account, User:MarkWarren. According to LinkedIn Mark's a Senior Software Architect in the bay area though I'm not sure what his focus is (I guess it's related to Oracle's cloud computing efforts).

I explained to him that I believed this conflict "gives rise to WP:NPOV violations in cloud computing and related topics (for example, attempting to push a WP:FRINGE view as mainstream) as well as WP:V violations in trying to equate it to network computing using inappropriate, unreliable sources" and suggest that he treads very carefully, for example using the {{request edit}} template.

I'm still somewhat concerned about edits like this (which used inappropriate, unreliable sources) and this (which had failed to get consensus above) as well as the fact that this is going on in business hours (I would hate to think that "fixing" this article was an assigned task), but at least now we know what to keep an eye out for. -- samj _in^out 04:33, 6 March 2009 (UTC)

First use of a cloud picture

The first time I saw a picture of a "cloud" being used as an abstraction for a complex networked system was for the Public switched telephone network, so the concept of the picture of a "cloud" for an abstract network is far older than many might think. Mahjongg (talk) 23:34, 5 March 2009 (UTC)

Yeah it's been around for a while. I tried to find a good network diagram with a cloud to illustrate this point a while back, but there were none. As you can see I've finally created something a bit more suitable than the infamous "The Cloud" placeholder image. -- samj _in^out 04:36, 6 March 2009 (UTC)

Risk mitigation undue weight flag

Opened this section for discussion of the undue weight flag added to the Risk mitigation section. sn‾uǝɹɹɐʍɯ (talk) 03:57, 6 March 2009 (UTC)

Cool, so the point was that the other sections are derived from almost 100 reasonably reliable sources and are somewhat of a consensus. This entire section is derived from one report/article. It's not to say we shouldn't have a section dealing with risks... just that we should try to find some consensus on the top issues e.g. privacy and security. I think I saw something like this in one of the IDG reports. There was talk of a criticism section (and I think one may have briefly existed) but we generally don't like {{criticism-section}}s. -- samj _in^out 03:53, 6 March 2009 (UTC)

OK, so it's mostly a request to insure that all of the risks are covered. While Wikipedia doesn't have two-sources rule like the NY Times, that sounds like good request. sn‾uǝɹɹɐʍɯ (talk) 04:05, 6 March 2009 (UTC)

Essentially yes. Let's kick off a list of things to hunt for references for. -- samj _in^out 04:53, 6 March 2009 (UTC)

• Security (including availability?)
• Performance (including availability?)
• Privacy
• Governance and Regulatory Compliance

I am in a panel discussion on cloud computing security at this moment, and the Gartner study was cited by one of the panelists. I have also seen it cited elsewhere. So, I believe it is a trustworthy source. I do not think undue weight is given to this section. User:Jeremycec