Doctorow On What Cloud Computing Is Really For 348
Diabolus Advocatus alerts us to an article Cory Doctorow has up on guardian.co.uk, addressing what cloud computing really means for the average consumer: "The tech press is full of people who want to tell you how completely awesome life is going to be when everything moves to 'the cloud' — that is, when all your important storage, processing and other needs are handled by vast, professionally managed data-centers. Here's something you won't see mentioned, though: the main attraction of the cloud to investors and entrepreneurs is the idea of making money from you, on a recurring, perpetual basis, for something you currently get for a flat rate or for free without having to give up the money or privacy that cloud companies hope to leverage into fortunes."
Rent Seeking (Score:0, Interesting)
This is what is broadly defined as "Rent Seeking"; extracting more revenue from customers without delivering additional value.
SaaS != Cloud Computing (Score:5, Interesting)
Doctorow's gripe is NOT about cloud computing, but Software as a Service setups, where the software is externally hosted.
"Cloud Computing" is a very nebulous term, ranging from online apps in the browser (Google Apps) to high level compute APIs (Map-Reduce etc) to low level VM hosting and storage (Amazon EC2/S3).
The interesting things, IMO, from the cloud point of view are the compute side, which is a windfall (we used EC2 to great effect with Netalyzr), and the reliability/infrastructure offloading.
And let's do a puzzle here. Yes, a cheap computer is just that, CHEAP, which implies unreliable. Gmail, for all its griping, has pretty much 99.99% uptime. Does Doctorow realize how much even that level of reliability costs when done in-house?
Re:What are they talking about? (Score:2, Interesting)
Well you get a flat rate for that disk and CPU sitting next to your desk that will be worth nothing in 3 years.
Or you can spend $15-20 a month and get a constantly refreshed and updated/upgraded system every time you turn it on.
Do you mean the client terminal will be part of the deal and be rented too ?
Otherwise it won't be updated/upgraded each time I turn it on.
Well, there are other things, too... (Score:5, Interesting)
Moving people from their own computing resources to yours is about one fundamental: control. I control my PC in ways that I normally have a great deal of say about (sure, "regular people" may have to hire consultants or expert systems to regain control of their systems, but at least the potential is there).
The recurring payment model is the modern gold rush... companies are willing to give you "free" satellite STBs, cell phones, etc. in return for knowing they're getting your $50-$100 back on a regular basis. This also moves to an interesting market model. With regular purchases, you probably have to convince me that you're the best for my needs, if I'm a well informed consumer. With contracts, once I've bought in, you need to finr the minimal amount of satisfaction that keep the vast majority of your customers "hooked". So people love and defend their choice of Nikon over Canon, or Sony over Panasonic, for the most part. But everyone complains about their cable company, their cellular provider, etc. And yet, those are the guys making the Big Bucks.
So it's inevitable that web services will go in that direction, at least some of the time. There's currently little precedent for getting consumers to pay, but "cloud" subscriptions are at the same time being sold to business as an alternative to expensive desktop tools (even when free desktop tools are also available). For some business use, it's not going to be about the money, per se. They might actually prefer a subscription to a lump payment... that makes expenses predictable... the same reason many businesses lease equipment, rather than buy, even though the long-term expense is greater.
But what they'll really be buying is control. Many companies work hard to keep workers from installing "unapproved" software applications. Move everyone to the cloud, and they lose the ability to customize anything you don't want customized. This is probably the engine that'll push business into the cloud, and get them to pay.
For consumers, follow the cell/cable model... if you sign up for two years of Bubba Jones' computing services, we'll send you a netbook (running a ChromeOS style OS that puts everything under control of the cloud services, even though some local storage will still be possible). There are enough people unconcerned about "real" desktop computing that this will probably seem like a good deal. Particularly if they're unable to do the real math. Which many won't... ask any iPhone toting friend what they paid for their iPhone.. they'll usually say "$200" or some such. When in fact, they're probably paying a total of something like $2000-$3000 over the course of two years, once you factor in the contract costs. But if it's a slow enough bleed, and you keep them happy enough, folks don't notice.
Confusion of terms. (Score:3, Interesting)
On the one hand, you have something like Gmail: Basically everything there(your data, all the code, etc.) is on their systems and under their control. On the other hand, you have something like EC2, which is basically just VPS hosting with higher-than-traditional provisioning speed.
The first type creates real risk(particularly for more unsophisticated users) of the expensive longterm rental replacing ownership problem we've seen with other industries. (Consider poor old Grandma, still renting a phone from AT&T decades after 3rd party devices were allowed, cable box rental fees, and all the other attempts to tie individuals to a recurring charge setup). The situation isn't all bad; but there is real room for concern.
The second type seems much less threatening. First, it'll be aimed largely at more sophisticated users, who will have more options and negotiating room. Second, the potential for easier migration will presumably keep costs down and service relatively high. Something like EC2 is largely standard(the compute VMs you are allocated) or fairly simple(the mechanism for requesting/provisioning more) and available in independent implementation. Amazon can still crush the little guys through scale and efficiency; but there is nothing stopping you from going somewhere else, or running your own, if they decide to abuse the power.
Given that Doctorow is writing for a popular publication, about the impact on joe user, I'd say his warnings are justified. They may well not be justified for you but all the whinging in the world about how simple it is(for you) to just run your own server won't change the fact that you'll be surrounded by people paying more than they expected every month for the cloud(just like they do for all the other "services" in their lives). However, it isn't at all clear that his warnings usefully apply to the commercial sense of "cloud computing" where it basically just means hosting.
The Personal Cloud (Score:3, Interesting)
The web was supposed to be a cloud to begin with. I think services like Opera Unite are pulling in the opposite direction and reinforcing what the web was supposed to be like to begin with.
Did you know that the HTTP protocol has PUT and DELETE commands? As far as I can tell no browser implements them. It does explain why we have primitive authentication.
I call services like Opera Unite and Mozilla Weave a personal cloud because they can be hosted yourselves. The Opera servers only provide hole punching between unite users.
This is an example of what I want to see http://jkontherun.com/2009/06/16/opera-unite/ [jkontherun.com]
and my here. [inforumal.com]
It's sad that our society's photographs are on Facebook in low quality. The big tech companies want to make us powerless over our data and retain control of them.
Subscriptions have always been more profitable than actual game sales. Blizzard is laughing its way to the bank after selling the game and then asking for more money to play the game you already paid for.
It's not a commodity! (Score:2, Interesting)
Re:Ars Technica Already noted and responeded ... (Score:2, Interesting)
One example of how I use the cloud is Evernote, which I use on my Mac and on my Pre. If I'm going to an off-site event or meeting, then I'll open up Evernote and create a new note in my "Events" folder. I'll paste into that note the meeting location and time information (usually from an email), any related instructions or information (e.g., "go to the front table of Building D and pick up your guest pass from Susie"), and part of a screen grab of a Google map of the location and maybe even a grab of the street view.
I use my cell phone for that, and the data are stored on the phone itself.
When I get to the area, I pull up Evernote on my Pre and this entry appears on my phone, complete with the Google map clip
If I need a map, I can get to Google with my phone.
Another example is flight tracking apps for the Pre and the iPhone; these let you put in your flight information, and they give you real-time updates of gate changes, departure times, and even live maps of the flight.
You're already on the way to the airport, so what good is that information?
A final example is my mobile address book, which, as I pointed out Part II of my Pre review, is a bundle of services that I subscribe to, not a static repository. If I call up a friend's mobile number on the Pre, I know that this is the latest number that they've put into Facebook or Google, and not some potentially out-of-date number that they once sent me in a vCard.
How often does that happen? If I change my number I tell people. At least, the ones I didn't change the number to avoid.
My point is that in all of these instances I'm not just doing the same thing that I previously did but in a browser window--I'm actually having a computing experience that I couldn't previously have sans cloud.
It would have been nice if he's pointed out something that was actually useful.
But the same dynamic doesn't necessarily hold true when giant, faceless corporations deal with each other, and this is why metered cloud service models work in the enterprise.
BINGO! If you find a worthwhile use for it, use it. But the "cloud computing" people want ME to use it; they're going to have to make it useful to ME.
Doctorow is right: it's just so much vapor.
That was the bottom line.
Re:I'm not sure I understand (Score:5, Interesting)
Do I use gmail ? certainly not ! Ewww !
And yes.. I do have a single server somewhere I use to handle my e-mail.. and my DNS.. (only thing is I have to hire the service of a registrar to write stuff in the ICANN db.. but I can live with that)
Do I use Google Docs ? You've got to be kidding right ?
Do I use collaborative solutions to author documents.. sure.. e-mails, mailing lists (which I can eventually host by myself should it become necessary) and a couple tools I host on the aforementioned server
Online backup service ? YUCK ! I have a few machines here and there and cross backup (ok.. so it IS Online Backup.. but I *know* were my stuff is located).
And I don't even have (or want) an iPhone !
So ! there !
(well... you weren't actually asking ME the question were you ?)
--Ivan
Re:I'm not sure I understand (Score:4, Interesting)
Yea, no, I'm not doing that. I'd rather keep my computer and not run a thin client and "trust" that the company isn't monitoring what I'm using "their" server cluster for. The exception I have to that rule is Google docs.
Agreed. Sorry, but when I read that the first thing that occurred to me was "all of this because the average person thinks Windows is too hard, or otherwise refuses to get a clue." What concerns me is that buzzword-ridden ideas like cloud computing will probably appeal to the non-technical masses (addicts to convenience that they are), to the point that the rest of us may be forced to partially or wholly accept them. I really don't care to give up even a small fraction of my privacy merely because Joe Sixpack couldn't be bothered to read a book or two. There's no justice in it.
This reminds me of the more asinine software EULAs which not only state the standard fact that you don't really own anything despite having paid for it, but also state that the vendor has no liability no matter what happens, not even when the software fails to perform as advertised (I think they call it "suitability for purpose" and expressly disclaim it). If the cloud computing vendors decide to implement a TOS like that, then your data is effectively held hostage and you have no recourse if something happens to it. What would be their real incentive not to do things that way? An informed, technically literate public which fully understands all of these issues? Yeah, right.
Like any and all proposals to do for you what you can easily do for yourself while charging you for the privilege, this has "bad idea" written all over it. As though all of the buzzwords didn't tip you off...
Re:I'm not sure I understand (Score:4, Interesting)
You missed the lock in model of being forced to work with the applications that the cloud provider supports.
Just like with a desktop app, if the provider chooses to lock you in, you're locked in. If they let you export to a standards based format, you're fine.
Any time I choose to export my messages out of GMail, I can do so (of course, due to volume, it may take some time).
Re:I'm not sure I understand (Score:4, Interesting)
There are definitely reasons not to use clouds, but lock-in isn't one of them.
Re:Jeez, the clould is not a conspiracy (Score:3, Interesting)
It is a just a fucking technology to make managing those backend servers easier. Not good or not bad.
The most informative post here is yours, AC. Cloud computing is about businesses reducing their IT costs by using other companies' servers. It's not about Joe Schmoe getting rid of his cheap hard drive and putting all his information online. Cory's example of S3 is an example of the relatively small amount of overlap, and it's presumably not too evil for he himself to use.
Speaking as a software-as-a-service provider... (Score:2, Interesting)
OK, I'll bite. As someone who runs a SaaS product (http://gimlet.us [gimlet.us], in case you care), I can assure you that we're not trying to nickel-and-dime our customers. We're trying to provide useful software at a reasonable price — nothing more, nothing less.
I've run a very similar open-source project, and found that by far, the most frequent question from people was "how do I get this running?" I talked to many people who wanted to try it, only to find that their IT department was an obstacle. One person told me — no lie — that their IT staff would charge $26,000 to install a small PHP/MySQL app.
Offering our software as a hosted service means we can provide it to nontechnical users without needing the help or approval of their sysadmins. It means that deploying patches is relatively straightforward, and that installers and packaging are things we just don't need to worry about. Instead, we can spend our (limited!) development time making our app better.
Will we, at some point, offer our code "for sale" as an installable, locally-run product? Almost certainly. However, the demand hasn't been there so far, so our efforts have been focused elsewhere.
Yes, there are some real concerns about putting your data up in someone else's cloud. But the idea that we're offering our app as a service to fleece people is simply not accurate.
Re:I'm not sure I understand (Score:3, Interesting)
Sure I do..
But this is pretty much *ONE* way.. (except for any coordinate or mouse movement)..
And of course, I also (extensively) use google's search service (which could probable have been a much better example !)
But I don't *store* anything on google maps.. or the search engine. If it were to fail.. to fold.. or whatever, I'd still have basic service. I wouldn't have lost 10 years of documents.. google (or whoever comes next) wouldn't be scanning my documents for every pesky little detail about my life.. and I wouldn't be - all of a sudden - asked to dish out 50% of my income so that I can retrieve that information (not saying they will.. but with the current state of affair.. so probably *could*)
I'm not saying we should *never* rely on external services.. I'm just saying I'd rather (while I can) keep control over whatever I can (ok.. call me a control freak.. maybe I am !)
--Ivan
Re:I'm not sure I understand (Score:3, Interesting)
Because I don't trust Microsoft, Google, or whoever with my data? They could lose it, data mine it, or sell it to my competitors to maximize their profit. If you think paying someone a monthly fee is better than hosting your own data and apps, well then your data must not be worth much in the first place. Not to mention the fact that distributed apps running over the network will usually (although not always) be slower than running them locally.
By the way, isn't anybody who uses Flickr already using "cloud computing"?
Re:Clouds are not solid enough (Score:1, Interesting)
Cloud computing would make sense if bandwidth were unlimited and highly reliable, while storage, memory, and CPU were expensive and unreliable. Unfortunately for the cloud enthusiasts, the opposite is true.
The only advantage I see in the cloud is backups are essentially outsourced, plus you get an easy way to ditch Windows. After all, if the universal/thin client is all you need, the local OS becomes irrelevant.
The 1996 version of cloud computing the "Network Computer" [wikipedia.org]. Benevolent companies (like Oracle) wanted to sell us dumbed-down computers at dirt cheap prices so that nobody would need a hard drive anymore. Applications could become the software equivalent of "Pay Per View" television.
Two major causes of failure: Plummeting storage prices wiped out the savings of a diskless machine, and the architecture was closed. These factors combined to kill the NC. It was such an obviously bad deal for customers, even the CIOs addicted to perks could not be persuaded into a cheerleading role.
So here we are, deja vu all over again. Except now the sales pitch is backups.