RMS Says "Software As a Service" Is Non-free
BillyG noted an RMS interview where he says "'Software as a service' means that you think of a particular server as doing your computing for you. If that's what the server does, you must not use it! If you do your computing on someone else's server, you hand over control of your computing to whoever controls the server. It is like running binary-only software, only worse: it's even harder for you to patch the program that's running on someone else's server than it is to patch a binary copy of a program running on your own computer. Just like non-free software, 'software as a service' is incompatible with your freedom."
...you can't use it when you don't have an internet connection. Why doesn't anyone think about this?
http://pinopsida.com
no it's not.
I had a customer just last week ask for us to get his backups from carbonite. I was confused and he said," I stopped paying for it a month ago, I want the copies of my backups from them."
I had to explain to him that you cant go to the car wash and demand the dirt off your car given to you after the wash cycle. It's gone, they delete all of it when you stop paying them.
He still did not fully understand it. And this is a college educated business owner.
"that's unprofessional of them to delete MY data."
Do not look at laser with remaining good eye.
I'm no RMS fan (GPL2 all the way) but isn't this shit obvious?
Do my rms-ian freedoms include deciding to use a website I know doesn't release the source code? Or is that more like the BSD freedom?
The problem is when you crack open your daily life and look at computing services you can't crack open and look at.
Every minor, hidden process is suddenly game. Go to a grocery store and use your credit card? You are using the credit card company's servers to take care of the work of moving your money from account to account. Your data. Their software. Cut up your cards. Heck, stow your money in a mattress. You don't want the bank to be liable for doing account computing when you can't get to the software.
I mean, I'd like to see the underlying process and know how their software works, up-to-the-code-level, but realistically? I don't have the time or interest. I'll trust the bank to keep my money safe and they get to enjoy the benefits of that trust.
import system.cool.Sig;
The message if you Read The Lengthy Article, is that if they don't have and open license to the server code, don't use them. He seems OK with the idea that you use a server based application if they are covered by the GNU Affero GPL.
If you are reading this, you have a perfect example of software as a service, in an open fashion. If you want to make your own /. go download the slashcode and set it up.
The correct direction to charge with pitchforks and torches would seem to be pressuring the Gmail team for a G-Code release, or making SquirrelMail (or your favorite server-based e-mail) as robust and reliable and Gmail.
That won't be easy. Does anyone here have a good suggestion for a starting point? What's the best FOSS ServerSide E-mail server?
Nobody has any intentions of stopping you from doing so. RMS merely recommends that you don't.
Really? He should read RFC 2119.
"Software as a service" means that you think of a particular server as doing your computing for you. If that's what the server does, you must not use it!
Sure sounds like an imperative to me.
Exactly. I don't dry clean my own clothes. Theoretically, I could. The methods and chemicals aren't a secret. Instead I turn over control of my cleaning to a third party.
And RMS has no problem with this. Would you, however, buy a new suit which could only be cleaned by the manufacturer, with an undocumented chemical formula? Would you encourage your organisation to adopt a policy that only people who wear these suits are allowed to attend important meetings?
I am TheRaven on Soylent News
If you agreed with RMS in principle, and wanted to create a SAAS business that is ethical, how would you do it?
Because really, it's the entrapment that makes it non-ethical, not the collaboration.
Making sure your users are able to get copies of their data in a useful format that are complete enough for them to walk away from you is an obvious one. Using an entirely open source stack and releasing any changes and improvements you make back to the community is another, more indirect one.
What other steps might you take?
Seems to me, releasing your entire source tree wouldn't necessarily be relevant for a lot of web apps, because they're more about representing network effects and business relationships on a grand scale, and are only useful if you wish to also be a service provider. Giving someone the source code that makes eBay run isn't going to be particularly useful if all they want to do is sell used merchandise.
Anyone got any clever ideas?
-1 Uncomfortable Truth
If the undocumented formula produced much better results, I might adopt such a policy.
If we finally get him the source for that printer driver will SFU?
This is my stance. When I'm programing software, I want the best tools possible to do my job. I like to use open standards and libraries when possible because they tend to foster a solid product. I'd like to make my code open source, but in most cases I'm not allowed by my employer (besides it's all for our internal staff).
When I'm using software, I just want it to work. I don't want to fuck with it, or tweak it, or read the source code. I don't want to modify it, edit it, hell I really don't want to even open a properties dialog if I can help it.
This is how I ended up moving away from linux. I was a big linux buff for years. I have taught classes at a community college on linux. I have gone to events and handed out ubuntu CD's by the dozens. But in the end, I wanted my computer to let me do work, not to let me work on the computer.
So I settled on a middle ground. I stopped being a hard line must be open source advocate. I use a mac (the most unfree of unfree) because it is solid, unix, and has a good selection of software that I need to do work.
So I use open source software if it meets my needs. But I'll also use non-free software without any hesitation. Who cares if I can't edit the code to my liking. I can't think of a single time I've done that in an open source product. Who cares if it can go away. I keep good backups of my important data. I can move to something else.
In the end I just want to get the task done as efficiently as possible so I an move on to having fun.
One single question of about twenty five regarding 'software as a service'. Your article title and selective quote does tend to give an erroneous impression as to what the article was about.
Maybe so, but you're also guilty of some selective quoting. Shortly after what you've chosen to include is...
Having an aim to actually remove consumers' choice as to whether they are allowed to choose proprietary software is indeed a little on the fringe-y side, you have to admit. I think that, based on this quote, the summary is fairly fair...
The CB App. What's your 20?
So if I remotely log into a linux server running 100% GPL software, and use that software to crunch data, it's non-free and I must not use it, because the server is owned and controlled by someone else.
So now software isn't free by it's license, it's free only if it's got a free license and it's on your personal box.
Ironic, because I was introduced to free software on my universities mainframe (e.g. emacs, LaTeX) and now I find out that wasn't free at all because I didn't have the money to buy a computer that could run it locally.
The GNU system works equally well today with Linux and BSD. Debian has released a BSD version, and a Linux version, both with essentially the same software from GNU LIBC up, just a different kernel.
What is it then? Surely not a "Linux" system when it's using the BSD kernel but is otherwise identical. RMS has always called this a "GNU system", and he had a point.
Bruce Perens.
If server-side processing is a bad thing, shouldn't he also be against X, SSH, VNC, and HTTP?
And if you consider the CPU as the "Client" then server microcode not on-die must also be remote, such as coprocessors, daughter boards, and peripherals.
Not to mention Beowulf clusters of anything.
Well, the fix is to periodically sync your data to your own machines and have code that could replace the SAAS if you wanted it to. At that point the SAAS vendor is just providing you with convenient, quick access to the cloud and if they go away, you can just buy some hardware and rapidly be back in business for your own stuff at a degraded speed.
But nobody does this which is why SAAS is a bad idea for anything other than 'nice to have' uses that you could live without.
Having the expertise is a whole different question. Not everyone can debug a program, but that doesn't mean that it isn't useful to have the source code of a program for those people that can.
If the SaaS you are using is open source, then you have the _option_ to set up your own version (or hire someone to do it for you) and modify it in any way you please, if you want to. You don't have that option with a traditional SaaS-provider. That is the freedom RMS wants.
This sig is intentionally left blank