Thinking about it though, a lot of it is a question of surprise and unknowns. I would find this message to be a lot better - "We see that you've taken advantage of the Ubuntu image that Canonical provide in the Azure Marketplace. I am available to you for (etc.)".
No. That's not better at all.
The mere fact that Canonical has specific information to reach me when I am not a direct customer of Canonical is a complete violation of my privacy.
Ubuntu is a free product. Canonical should not be able to find out if I (specifically me or my organization) allocates or runs 1 or 10000 instances of Ubuntu.
I agree with the message behind this and obviously Canonical and Microsoft are both being extremely gross.
But Ubuntu as a binary image (or source code) is a very different product than a VM with Ubuntu pre-installed and pre-configured, which is what you paid for (and is why you got ensnared by their horrible anti-user license).
How? Why? If it's different in any meaningful way from just clicking "next" on the installer then it's no longer Ubuntu, and certainly not Canonical Ubuntu, that's pre-installed. It's become, at best, Microsoft-Ubuntu-Because-Microsoft-Added-Telemetry-For-Azure. Or it's Canoncical-Ubuntu-Configured-By-Microsoft-With-Azure-CLI-Preinstalled.
It's not "Ubuntu" any more.
When I'm paying for an official Azure version of Ubuntu on Azure, I darn well expect there will be a closer support relationship than the free desktop version.
Okay, but maybe other people don't want that if it entails their information being shared with a company they haven't initiated a business relationship with?
> haven't initiated a business relationship
????
To say that you have no business relationship with Canonical while paying Canonical to use Canonical software with official Canonical technical support is absurd to the highest degree.
You deciding to resurrect the comment because you happened to see it before I deleted it is really not OK. It's the exact kind of toxic hostile, creepy interaction I was trying to avoid from you by deleting the comment!
I thought your comment was interesting and merited a reply for others to see and discuss. But I see you disagree so I've removed the content of my reply.
Feel free to flag any comments you find particularly toxic or hostile. You can do that by clicking on the timestamp of the comment and clicking the `flag` link.
Or even better, let me know (like you have done so here). I can't improve myself if I don't know there's a problem.
Posting something and deleting it after it has been seen is basically gaslighting. Imagine the kinds of harassment people could get away with if they said rude things to coworkers on chat, then edited the messages to appear benign after the coworker responded to their hostility.
That is why people quote the text of comments to which they want to reply.
Furthermore is strenuous disagreement now toxic and hostile?
Wouldn't it be more trivial to say I do not wish to engage and leave it at that? Ironically calling someone toxic hostile and creepy is... pretty toxic.
I think someone has the right to change their mind about something they've said. That's why I edited my comment to remove it.
> Furthermore is strenuous disagreement now toxic and hostile?
I don't think so. But I know that I sometimes get passionate about my opinions. I welcome someone's input to keep me friendly.
> Wouldn't it be more trivial to say I do not wish to engage and leave it at that? Ironically calling someone toxic hostile and creepy is... pretty toxic.
I would like to think better than that. I think it was good of @ojnabieoot to let me know that they thought I'd wronged them.
Some people can feel very anxious or awkward to conversation for very good reasons. They can state opinions and then choose to retract their opinions for any reason -- even if the opinion is held but they choose to remove themselves from the conversation. I think that's a good thing to discuss but this isn't the venue to.
Ditto the Ubuntu images on Docker Hub.
This is a big misstep for Microsoft, from my point of view. I think it's less a reflection on Canonical, because once they have the information, it's ultimately going to be used. Microsoft just should not have agreed to the arrangement at all.
To quote the old native american (?) fable: You knew what I was when you picked me up.
I don't think that most of the people have a problem with that. The problem is being sucked-in to something without ever agreeing into.
In the era of privacy sensitivity (which I think is healthy), being watched in a place and prodded from a different channel is disturbing.
I don't mind people trying to reach me with the hope of sales based on information I've provided to them, but this is too far.
Also it removes two veils from both companies at once:
1. It seems Microsoft still has sneaky tactics, but they're more invisible.
2. Canonical is somewhat more aggressive and greedy than it seems, and Ubuntu desktop is just a freemium product, or another capturing device for further vendor lock-in.So MS sharing "their" customer details with the image provider seems more generous than evil. Provided there's a "Do not share" config option somewhere.
If I download packages and Ubuntu, and assemble my own image, or use one assembled by another org, probably not.
I think the disconnect is that for me, image packaging and updating is work, and that work has an author, and the author is deserving of certain rights others are not.
If Azure is auto-pulling Ubuntu images, building containers, and publishing themselves, then that's a different story.
Ubuntu is gratis, so Canonical can't have coerced Microsoft into doing so; it is quite probable that one approached the other to make a deal, and that Canonical is paying a certain fee for this information.
We're talking about a curated, supported, official image here, right?
If folks want to use a "MyUbuntuImage" they or someone else packaged and uploaded, more power to them.
But by pulling a Canonical image, you have a relationship with Canonical. Expecting that relationship not to exist "because open source" seems to be misunderstanding who does what work.
As to whether this should be opt-in, done, etc. is another matter entirely. But the fact that it exists at all doesn't feel particular shocking.
It's not like we're talking about everyone who pulls a RedHat image's info being sent to Canonical!
I don't expect an OS based on an OS based on an OS based on a half-finished OS based on free software principles to have shady data-dealing attached, yet hidden from the people whose data is being dealt.
Privacy protection is not an obligation, but transparency and openness is. Yes, you're not contractually required to not make a separate computer system that's proprietary and closed and disempowering, but that's so pedantic as to be malicious.
I mean, it's kind of ridiculous to think that you could do anything in a cloud environment system and not have your actions tracked. Hell, with automated load balancing and load-based billing, that's literally what you're signing up for.
In this case, the license is the GPL, none of which has anything to say about privacy. Maybe this is a failure of the Free Software Foundation's to not include privacy protection in the GPL. Though even if they were to create a GPLv4, the Linux Kernel is still only licensed under v2, so distro implementors have no obligation to use a more restrictive license.
AKA, "the cat is already out of the bag".
In the OP's case, they additionally are are customer of Microsoft's, who explicitly stated they share this kind of information with their vendors.
Debian Free software guideline does not allow discriminate against using debian for evil.
Citation needed. RMS, the FSF and many other orgs made public statements around privacy many times.
Now, I can't exhaustively prove a negative, but I think I can easily demonstrate that the FSF has never meaningfully expressed an opinion on privacy. Go to https://www.gnu.org/philosophy/philosophy.html, open every single page it links to in the body of the text, and search for the word "privacy". It does not show up in the body text of any of those documents. It shows up once in a footnote that mentions a change that Samsung made had that "caused privacy concerns".
The closest they get to even mentioning the concept of privacy is when they talk about the right to modify software and use those modifications "privately", which clearly does not mean anything about user privacy.
If privacy were so big of a concern for the FSF, you'd think they'd talk about it in their official documentation on their philosophy, or put something about it in the ONE tool they have to have power over anyone: the GPL.
The anti-patent-trolling, anti-tivoization and copyleft provisions are there to protect developers and users.
Additional clauses around privacy and security would be very nice.
Unfortunately, corporate-sponsored FUD made a lot of people wary of the GPL - which is ironic, given its protective features.
There is a certain level of reasoning where one might say that, if the software were truly libre, you could "just" fork it and rip out the parts you don't like. But because you clearly can't "just" do that, then the software must not be free.
Yes. The software is not Libre.
But it's not clear to me that this is the case because the system is hosted on Azure or the distro is Ubuntu. Your rights within a marketplace go only so far as you can throw your alternatives. Software, especially operating systems, are just too complex to expect the concept of Free Software to be sufficient to protect user privacy.
Another interesting question: aren't you a direct customer of Canonical here? When you buy stuff off of any marketplace or though a reseller, it seems to me you are a customer for multiple companies. Examples: buying an iPhone from AT&T, buying a laptop from Amazon, buying a Subaru through a dealer.
When you get it a certain way through Azure you both enter a contractual agreement with each other, and that does make you a customer.