Hi everyone
Lately I'm reading lots of things about "open source cloud computing". In press releases of press articles, no one really ever explain what it means, and why it is important.
Do you know more about it? Do you have concrete examples? What can be the advantages in term of freedom (could those "clouds" be interoperable for example?) Is there already some standardisation going on? The normal problems of SaaS remain, don't they?
Thanks for sharing your insights
Lucile
Hi Lucile,
I think "open source cloud computing" is used, at least in most of the cases, as a short way to say "open source software for building private and public clouds". OpenStack (http://www.openstack.org/) is, maybe, the main example of that. Do you think there is any other meaning for that expression?
Best regards
Vicen
El mié, 12-02-2014 a las 15:13 +0100, Lucile Falgueyrac escribió:
Hi everyone
Lately I'm reading lots of things about "open source cloud computing". In press releases of press articles, no one really ever explain what it means, and why it is important.
Do you know more about it? Do you have concrete examples? What can be the advantages in term of freedom (could those "clouds" be interoperable for example?) Is there already some standardisation going on? The normal problems of SaaS remain, don't they?
Thanks for sharing your insights
Lucile
Discussion mailing list Discussion@fsfeurope.org https://mail.fsfeurope.org/mailman/listinfo/discussion
Lucile Falgueyrac wrote:
Lately I'm reading lots of things about "open source cloud computing". In press releases of press articles, no one really ever explain what it means, and why it is important.
That's because what's being offered is often purposefully unclear so the organization making the offer can change service terms without tying themselves down to something one can hold against them later. The usual framing of the issue is narrowed so user convenience is paramount and user's software freedom is left out entirely. There's typically no mention of the freedoms would-be users must give up in exchange. What the services seem to have in common is using programs that qualify as "open source" to provide a service as a substitute for software one probably ought to run themselves instead. But you'll never get a copy of the complete corresponding source code for those programs, only those within the service provider's organization will.
Do you know more about it? Do you have concrete examples?
I think the FSF's many papers and remarks on both "open source" and "cloud computing" do a good job of explaining both terms.
https://www.gnu.org/philosophy/open-source-misses-the-point.html
https://www.gnu.org/philosophy/words-to-avoid.html#CloudComputing
Any remotely hosted service on a computer one does not own would serve as an example here: email hosting service, calendar sharing service, web hosting service, any file hosting service, the list goes on and on. Of course, the details of what one intends to host determine the relevant threats; not all hosting services pose problems for all uses.
For a few years Richard Stallman has been giving talks about how "digital inclusion" can be a bad thing. In these talks he outlines classes of major problems quite well and his classification of threats address your questions. I recall one such RMS talk from 11 October 2011 -- http://audio-video.gnu.org/video/stallman-sciencespo-freesociety.webm -- I suggest finding this and other talks on audio-video.gnu.org.
What can be the advantages in term of freedom (could those "clouds" be interoperable for example?)
The service owner's software freedom might be respected, depending on the details, but:
1. The open source movement doesn't talk about software freedom. That's part of the reason why that movement exists -- to not bring up software freedom as an ethical issue because that movement's proponents think such talk interferes with talking to businesses. See https://www.gnu.org/philosophy/open-source-misses-the-point.html for more on this.
2. It's unlikely that any remote service will respect a user's software freedom or give the users a chance to know exactly which threats they face in using the service. Such users are essentially handing the service data without knowing how that data will be processed. Perhaps the data will be shared without the user's knowledge? Perhaps the service will misrepresent the user in some way? Perhaps information about the use of the service will be shared in a way the user would not approve of?
Is there already some standardisation going on?
Any standards I've read about focus on technical issues and on helping service providers make it difficult for users to migrate off of the service with their data intact.
The normal problems of SaaS remain, don't they?
Yes, all the problems of giving one's data to be handled by someone else's computer (software as a service substitute -- https://www.gnu.org/philosophy/who-does-that-server-really-serve.html) remain including: - one can't be sure what happens to the data, regardless of what the terms of service claim. - one can't be sure what happens to the data (or records describing the data) after the service relationship ends. - one has (quite rightly) no control over the service provider's computer -- there is no way to give multiple people complete and exclusive control over someone else's computer.
The concept of owner's rights work against those who seek to have someone else do their computing for them while retaining control of the other person's computer. One can't justly demand control of someone else's computer without tacitly supporting losing control of one's own computer as well.