Unintentionally hilarious given the way the FSF worked to make itself irrelevant by ignoring hosted apps for decades. I mailed them in the late 90s asking why it was that running an app on my computer meant I should get the source code but running the exact same app via telnet meant I shouldn't. At the time I was writing a MUD server I wanted to open source, but I was also thinking of websites. The only answer I got was that, yes, the GPL doesn't cover this and they have no interest in extending it to do so.
Stallman can whine about it if he likes, but he can't pretend this is some sudden terrible attack on Free Software - it's been a problem for a long time, and it's been ignored for a long time.
(Yeah, I'm aware of the Affero GPL. AGPLv2 was just bad. AGPLv3 is decent but way late and inexplicably separate from the GPL.)
I remember reading some Eben Moglen & i think he mentioned that they didn't hardcode this into GPLv3 because the rights of someone using an ATM are (should be?) different from those of someone owning the machine with the code.
When trying to grok the GPL, it helps to keep rms' MIT printer story in mind. It's on wiki.
The 55-year-old New Yorker said that computer users should be keen to keep their information in their own hands, rather than hand it over to a third party.
"users should be keen to keep their information in their own hands"
That's Zero Knowledge. Like the recently unfunded fantastic Clipperz library. I can't believe no smart investor found a way to make money with that. It torpedoed my respect for angel investment. Anything "social" gets money, something that can revolutionize the whole Internet experience, zip.
I honestly thought that was a Douglas Adam-style parody (from one of the later books). Holding all your passwords in one place (no matter how secure) creates a single point of vulnerability.
I do think that it's (theoretically) possible to safely store (in some respects) your data online, through encryption... but this doesn't protect you from innocent data loss (e.g. act of god, bankruptcy, incompetence, bad luck etc).
Why not store all your data locally as well, so it's distributed git-style, and only use the cloud for crunching/archive? This is the model that Google (google gears) and Adobe are moving towards (though I thought their main motivation was so you could work off-line). HDD are unbelievably cheap, and getting cheaper and cheaper... Gmail gives you 7GB for free - wow, that's so much! - but how much do you have at home already?
There's no shortage; it's just convenience.
That startup's PR connection with Stallman (that they use GPLv3, they talked to Stallman and are concerned about the same problems) seems tenuous and disingenuous to me (though I didn't investigate in depth).
Please try it before bad mouthing a very clever and working project.
They don't forbid you from storing things things locally. And it isn't just for passwords. This is for anything really. Watch their presentations at least.
I know something about public/private key encryption (or once did). I'm not criticizing that, but their application of it. In their presentation, they say that they picked passwords as the hardest possible test of their idea (i.e. the one people are least likely to accept). I think this is where the "parody" aspect comes in, with their professed motive of proving a point, rather than picking the most likely candidate for adoption.
I wasn't bad-mouthing them, but listing definite concerns. You might see the Douglas Adams reference as "bad-mouthing", except that what I'm referring to is his fictional idea of storing all your identification/passwords as one - i.e. what these guys are doing. I also don't think it's bad-mouthing to state my opinion that their PR connection to Stallman is tenuous, since they promote online storage in his name - which he appears to oppose (in the very article we are commenting on, I might add).
But of course, they still could change the world through their ideas, ideology or technology - everyone has that chance if only they take it. I strongly believe in encouraging anyone who takes action (so it would be really cool if they proved me wrong - and I'd have something to learn which would be a bonus for me). But I also think it's important to discuss ideas with intellectual honestly. Those two views are at odds sometimes, and I don't always know how to resolve them.
If you disagree with my criticism, you could focus on refuting it instead of labeling it as "bad-mouthing".
I honestly thought that was a Douglas Adam-style parody (from one of the later books).
I wasn't trying to put the project down with that, but reporting my reaction to the site - I actually was laughing along with (what I thought was) their parody for a few minutes (it seemed, though maybe shorter). I don't recall when I started to think they might be serious, but it might been when I saw their page "How dumb are we?", and they were defending their position...
It's true that it would have been more diplomatic to conceal my reaction.
Also, I searched around for a reference to the Douglas Adams passage, but couldn't find it (I couldn't find my copy at home - all my books are in boxes). I'm pretty sure it's near the start of "Mostly Harmless", as a product of the "Infinidim" corporation, and something to do with a multi-dimensional bird... I would like you to see it, because then I think then you'd have a clear sense of my point of view (independent of your dis/agreement with it)
Incidentally, labeling a comment as "childish" and "inflammatory" provides little information about the subject, and more about the emotional state of the labeler (I am sorry to have upset you). Giving the reasons for that emotional state are often a more compelling refutation - that is, if you can see the reasons (I find that very hard myself).
Only if you don't know the book. Even then, I explained it in the next line (single point of vulnerability).
But I can see that if you read the first sentence in isolation, and didn't know about the book, it could be interpreted as rude. It wasn't meant as rude. I'm sorry that it came across as offensive to you, and so I clarified in my follow up.
BTW: They're excellent books (the second in the series is best IMO).
I've been thinking about this lately - about giving all of your data to a third party and hoping they don't go out of business or erase your account (for whatever reason).
Issues of software freedom aside, I think the key to making people trust the cloud is to make the data contained in the services more transitive. Essentially, creating a standard Import/Export for the web. i.e. - If I decide to switch to vimeo for my web video needs, I'd like to be able to import all of my videos from youtube without much fuss (or make a local backup). That's not a great example, but it's the general idea.
Hmmm...after I think about it, this pretty much boils down to: "let me get at my data in a format that I can read."
I've been thinking about this lately - about giving all of your data to a third party and hoping they don't go out of business or erase your account (for whatever reason).
Or be forced to give up access statistics via Pen Register.
Or provide governments the ability to access your data via CALEA compliance hardware/processes.
I value the privacy of my data as much as I value the integrity of it. I don't see myself doing much cloud computing in the near future.
What's wrong with having the company charge you (and all users) enough to make their service profitable, and having a contract between you that they wont erase your data without sending you a copy first?
Keeping your personal data on a free account might be a bit of a crapshoot, but keeping your company data on a paid for service account shouldn't be.
Nothing wrong with that at all - that actually sounds really reasonable. Is anybody doing this? I'm getting to the point where I'd pay for most of the stuff that I'm getting for free now - if I had some sort of contract that would give me some recourse in case things go awry.
I've thought some about what it would take to set up a system for failed startups so that customers could be assured that their data / services don't disappear when a company goes bust -- something similar to the Free Qt Foundation that KDE has setup with Trolltech:
Here at salesforce.com (the company that does cloud computing/saas for businesses,) we have the data loader tool that customers can use to download all of their data locally as often as they wish. Also, there is an API so people can write their own backup/restore/integration code.
Hmmm...after I think about it, this pretty much boils down to: "let me get at my data in a format that I can read."
Basically what you're talking about is standards - but I think the SaaS/Cloud space is too immature to be able to define standards at this stage... In fact, whilst it would give consumers extra security, it would very likely slow down innovation.
You can see the trap he's talking about as one possible future on the horizon--I don't think we're there yet, and it will take continuing to be vocal about our concerns to keep the benefits from turning sour.
The 2 main potential problems with your data being in the cloud versus on your own computer seem to be 1) privacy and 2) portability. Privacy long-term seems to be a tough nut to crack, and seems to depend mostly on choosing companies we feel we can trust. Data portability will only win out against walled gardens when users scream for it loud enough rather than go along using whatever service (such as Facebook) despite your data being locked in.
But the benefits we get from using the cloud for storage and computation are going to be difficult to ignore, despite the above concerns, and big brother is going to be constantly creeping in as long as the people continue to be scared into letting him.
I can see this being a problem for consumers long-term. But it doesn't do any good to hear that Larry Ellison has come out strongly in favor of keeping your data in-house.
Its like saying "Lets not use banks cause they can just take your money, profit from it, and may not even give it back to you." Which, ironically, is true because of the financial crisis.
The solution though is not to "use in-house computers" (or store your money under a mattress). I think it would be to build your own cloud computer services based on a free software model.
Easier said than done. Like building your own currency or financial system, it is easy to create a toy and very difficult to scale this out where you have to deal with trust issues, counter-party risk, and transactional overhead. Free software is easy, free hardware and bandwidth is not.
It is not necessary to use open source software to be able to protect your data (provided the applications have an API or some method of exporting your data in a parsable format). It's also not difficult to build your own secure, replicable network in the Cloud. Get a Linode and a Slicehost, install a Notes server on them both, keep your data locally and in two secure, encrypted automated backups in the Cloud. Whenever you don't have access to your local Notes, you can even have elected to make all or just portions of your data available via a browser and https.
addendum:
Forgot to mention, total cost for the Notes client and server licenses: about $150 for 1 user but any number of servers. And Notes will run on Windows, Linux, OS X, ...
Unfortunately they're using the term "Cloud computing" when they really mean "Software as a service". I could be wrong, but I think RMS would have no problem with people using open-source software to manage their data on a rented slice at Slicehost or Amazon.
It really just depends on if there will be market competition or not. Desktop processing could suffer the same closed system-high-price fate if there weren't any competition in the hardware market (IE: Intel vs. AMD). The question isn't really cloud vs. clientside, it's Company A's cloud Vs. company B or C or D. As he said, it's mostly branding and the underlying technology exists and is pretty freely available so my sense is that cloud computing could be a more competition-rich marketplace than clientside processing is today since it's pretty dang tough to build chips.
He's generalizing. Cloud computing must involve locking away files in the cloud? Depends entirely on the company/organization you're doing business with. There's no reason a cloud app can't provide export features as well as a local Gears-like presence. There could even be an open source app operating in the cloud (like a free App Engine app) with full data dumps to your local computer.
His comments echo those made last week by Larry Ellison, the founder of Oracle, who criticised the rash of cloud computing announcements as "fashion-driven" and "complete gibberish".
I never expected Richard Stallman and Larry Ellison to agree on something involving the software industry.
The majority of users out there do not have the computer savvy that Richard Stallman has. What about my mother who's hard drive is always on the brink of failure and doesn't care to know what's involved in properly backing up her photos or documents? I encourage her to use services like flickr and Gdocs for this reason. The benefits far outweigh the drawbacks like privacy or portability. Anything that is so sensitive or important is on paper and in a fireproof safe like it should be. Don't get me started on Privacy.. what are you people doing that's so sensitive??
Until every computer user has sys-admin capabilities, making use of the "cloud" makes perfect sense in my opinion. That being said, I think its important to pick services you trust and who make it easy to export your data at will.
data loss and temporary outages at those big companies are not unheard of.
using flickr to share your photos with everyone is not the same as using google docs to store all of your personal and private documents.
but why not simply set your mother up with an automatic online backup system to store a copy of her files in the cloud instead of requiring that the master copies be there? that way she retains privacy (assuming the encryption of the provider is trusted) but gains portability by having a copy available online. something like dropbox comes to mind here.
Should we also home-school our children, keep our money under our mattress, grow our own food, give ourselves medical exams? There are companies/service providers out there who we outsource those things to because we don't have the time or the expertise. Seriously people have been using web-based email for over a decade. I've yet to hear of any doomsday failures but I do know I've had a couple hard drive failures during that time span.
> Should we also home-school our children, keep our money under our mattress, grow our own food, give ourselves medical exams? There are companies/service providers out there
grow some salad greens and tomatoes and eat those instead of McDonalds?
self check for breast or testicular cancer (whichever applies)?
take an active role in your child's education? I should hope so.
I'll give you the money under the mattress point. No return on investment. However you should be managing it yourself in some form.
Should we also home-school our children, keep our money under our mattress, grow our own food, give ourselves medical exams?
Hmm, my family does all of the above as a supplement of commercial/public versions all the time. (In the case of home-schooling, we plan to once our child arrives.)
I can't remember ever losing an email or photo from gmail or flickr. Yes, there's some downtimes (one big one with gmail and a few small ones spread out with flickr), but nothing show-stopping (note: you could still access emails with html-only gmail if you needed)
Experts in specific areas develop blindness to problems in certain areas of their solution. Usually the framework they use to approach the problems makes certain simplifying assumptions, and these assumptions become their blind spots. These frameworks solve the hard problems they are designed to solve, so the biggest remaining problems remain hidden under these assumptions.
Examples:
Economists assume rationality. Java developers don't notice all the boilerplate. Lisp developers don't notice the brackets). Stallman doesn't notice all the sysadmin burden.
But that's not true in Stallman's case. He openly acknowledges it by stating that we should choose freedom over convenience. He recognizes the burden and encourages people to choose it anyway because freedom is worth fighting for. That's at the core of his message. Its not in the article unfortunately but he does say this at his talks.
Is this a true choice? I am choosing convenience, if I need freedom later I can pay the price and get it.
For example, I am developing on the closed Google App Engine. If what I'm doing becomes successful and I hit limitations of the platform, I can reimplement the whole thing on an open source stack. But I don't want to pay that cost up front.
You have the choice - Stallman is just encouraging you to chose the bunch of admin work that you don't want, arguing that the freedom is worth more than the admin work costs.
Well... it is impossible to deny that the whole 2.0 social/sharing/blabla is possible with resident applications, distributed by definition in users PCs, faster, without javascript magics.
It is equally impossible to deny the fact that it is much harder to get a user to download and install software than it is to offer the same features and services via the browser. Unless you are interacting with the user/OS/hardware in a way that cannot be duplicated via the browser a web-based solution will always have a larger audience than a resident application.
I don't see a problem with cloud computing, as long as the software that runs the cloud is Free. That way, when my provider dies, I can easily host it myself or get another provider.
Electricity is a commodity. There's no inherent "freedom" lost in paying someone to give it to you, any more than there is for buying your bread from a market instead of growing your own wheat.
The proper analogy would be not to electricity in general, but to (in the US) 220V 60Hz two-phase alternating current. What if you could only buy that power format from one vendor, who then decided to increase the price or change the format such that you'd have to replace all your existing electrical hardware or be at their mercy? Then you might feel that your freedom to operate your equipment was being infringed, no?
Ya, the vendors of alternating current are the electrical utilities. And the issues your talking about happened in the 1880s, between Edison's DC vs Teslas's AC. Obviously 110V AC won (not 220V) and market forces and regulation have kept that standard and the price stable.
The point is that if we feel okay paying for electricity as a service, why not computing? Data ownership and security is one reasonable concern but they're not show stoppers.
There is nothing wrong with centralized models, it is when there is a center that is not open source. Anything that is programmable (multiple times, not write once mediums) should be ran on open source software. The electrical utilities are immorally not sharing the source code that controls them.
Yeah, there's a strong economic case for cloud computing. I'm not worried though. If control & source code availability become important, I'm sure people will switch again to the AGPL as they once switched to using GPL.
Stallman is a commie crackpot and everybody knows it. No respectable business person of note takes him seriously.
You can't have Falstaff, and have him thin.
Meaning, you can't have superior computing power like the "cloud" provides and expect to retain total transparency/privacy. Something has to give. I think it is inevitable in the long run that business will decide that the help to their bottom line that the cheap resources of the "cloud" provides offsets any theoretical leftist concern about transparency, privacy, etc.
Asking everyone to do their own computing is like going back to an agrarian culture where everyone grows their own food. Seem to recall Pol-Pot tried that, and I don't think history looks upon him too favorably. 'Nuff said.
'Respectable business people of note' have just overcooked the world economy and driven it up the embankment, through the hedge and are currently rolling it ass-over-teakettle through the nearest field, hoping desperately that someone built some bloody airbags into it somewhere.
Anyway, if people want transparency/privacy enough to pay for it then 'theoretical leftist concerns' will suddenly become noteworthy and start appearing. As they already do with online ordering and banking and virtual private networks and so on.
It is a common misconception to call Linux-based systems “Linux”. Linux is the kernel, the piece that manages devices, processes, and memory. GNU is a collection of tools (GNU ls, GNU Bash, GNU C Compiler, etc.).
What the user sees is the set of GNU and non-GNU tools running on top of the Linux kernel. This is the reason why rms insists on people calling their system “GNU/Linux” and not only “Linux”.
The Linux kernel is not the entire environment.
The set of GNU licensed tools and applications is not the entire environment.
Gnu + Linux is not the entire environment.
Those who care are aware of the distinction between "Linux" the kernel and "Linux" the environment built upon a Linux kernel. Those who don't are unlikely to be aware of the derivation or importance of either term. Of those who care few are willing to use his terminology.
This is not what I would consider a "misconception."
In my opinion Stallman does his cause damage by tilting at this particular windmill.
The same personality traits that cause him to spend inordinate amounts of time focusing on pedantic bits of terminology are the same ones that allow him to persevere in hammering home the same basic message over and over for 25 years in the face of setback after setback.
How many times do we have to go through this? You could have just used your favorite search engine to find the answer and the rest could just ignored you. By the way am I being too anti-social? Excuse me, please.
Seriously, what a drama queen. This week it's cloud computing.
The article seems to confuse software freedom with control over personal information --- two distinctly different concerns. The latter actually relevant.
Stallman can whine about it if he likes, but he can't pretend this is some sudden terrible attack on Free Software - it's been a problem for a long time, and it's been ignored for a long time.
(Yeah, I'm aware of the Affero GPL. AGPLv2 was just bad. AGPLv3 is decent but way late and inexplicably separate from the GPL.)