Hacker News Reader

Comment by bdcravens | Hacker News Reader

So how many developers here use Google Analytics, Intercom, Segment, error logging like BugSnag or Sentry, etc?

Wait until the BBC finds out how many of us are giving Amazon user data. (I mean, it's s3 and RDS, but that clarification would be overly pedantic)

View on HN

The difference is that Amazon isn’t an advertising company and has little incentive going through user data on their systems. In addition, the data isn’t in a standardised format so they would have to spend considerable efforts parsing the data first.

Facebook and Google on the other hand make their money from stalking people and developers are giving them data in a nice standardised format.

Is there a difference here between "an advertising company" and "does a lot of advertising"?

Amazon is very much the latter and that would seem to provide plenty of incentive to do dig through data / recognize the value / use it as they wish.

I'm sure they do plenty of analytics on data from their own platforms and that considerable effort could easily be extended to include any other data that they have access to.

Amazon's advertising platform now serves 13% of US search ad revenue [1]. They're definitely working on getting more revenue from advertising.

[1] https://www.geekwire.com/2019/amazon-gaining-google-search-a...

> Facebook and Google on the other hand make their money from stalking people and developers

Yet React and Angular are quite popular

I fail to see the correlation. You can be an asshole on one side and still make a great product on the other side.

You actually quoted the part about tracking developers.

Amazon is working their way towards vertical integration across as many industries as possible. Since effective advertising is critical for some of those integration steps, Amazon is 100% incentivized to hoard and process user data.

AWS is the only thing that keeps Amazon alive; if they lose their clients’ trust they will go down the drain in no time so there’s no way they’ll risk it.

i have no doubt that if you have a ballooning SaaS business hosted on AWS there is a capture team lead investigating what your tool does and whether or not it is worth amazon creating a team to re-implement in order to crush you.

they definitely do this for physical goods so i'm not naive enough to think they aren't talented enough to do it digitally.

They'll analyze your company from the outside, as a black box. They might even try to do business with you, so they can do a grey box analysis.

But I don't think Amazon will try to dumpster dive your encrypted S3 buckets in the process of doing the above. At least, not yet.

But Amazon is an advertising company, capturing 9% of the digital ad market in 2019.

Their main cash cow is AWS. There’s no way they’ll sacrifice it for a short-term gain on the advertising side.

Google and Facebook only have advertising; they have nothing to lose by being unethical and/or breaking privacy laws like they do with the GDPR.

How would they be "sacrificing it"? Where are people going to move to? AWS is the industry standard for public clouds, Azure's feature set is worse, and GCP is basically an also-ran.

Amazon is an advertising company, and has been doubling down on that market despite users not being happy with it:

https://apnews.com/a5ad925b06b7454ea746764399e1a096

This is a fair point. However isn’t amazon investing very heavily in advertising and supposed to compete Facebook and google. https://www.forbes.com/sites/kirimasters/2019/07/26/whats-dr...

> Facebook and Google on the other hand make their money from stalking people and developers are giving them data in a nice standardised format.

Do we know that the privacy policy/terms of service for these services allow Google/Facebook to use the data in the way that the BBC article seems worried about?

And before you say that "it doesn't matter what the ToS says", it does very much matter, breaking a ToS would paint a very big target no these companies, there are tons of lawyers out there that would love to catch companies at this, easy money.

Facebook claimed 2FA phone numbers would not be used for advertising. They eventually broke that claim.

There are thousands of companies breaking the GDPR (Facebook and Google included) and yet I have yet to see the tons of lawyers going after the easy money. Companies keep doing it because they know the regulation isn’t enforced.

It's interesting, given the reputation of ambulance chasers, that attorneys are deciding to ignore easy money. Why do you think this is?

There's no easy money for lawyers in GDPR - the big financial 'teeth' of GDPR are in the form of fines enforced by regulators to the state, not (as often in USA) in the form of huge civil lawsuits from which private law firms could get a share.

There's a big difference between Google Analytics and Sentry. Putting those in the same list is so reductive that it undermines your argument.

Tracking everything you can about your users so advertisers can better target you is evil.

Capturing all the local variables during some unexpected Exception (that might happen to include some user data) for the purpose of debugging is not even remotely evil.

User-data is totally fine to have, it's what you _do_ with it that matters

>user-data is totally fine to have, it's what you _do_ with it that matters

Unless the user explicitly opts into having that data recorded, it isn't ok to stockpile it, regardless if the intent behind it.

The path to hell is paved with good intentions.

I disagree. By interacting with a web server, you are inherently sharing some data with the operator like your ip address (so you can get a response back) or any request headers you choose to set (User-agent for mobile vs desktop sites, for instance). Nobody is forcing you to make requests to any particular site, and nobody is forcing to you include all this info. If you don't want your data spread around, don't do all that spreading.

I agree that using user-data outside of some agreement is bad (and illegal under the GDPR), but I believe that an implicit agreement exists between web-server provider and user that their data will be used for the mechanical operation of the website, including logging stack-traces. Otherwise TCP/IP wouldn't work.

Reply on news.ycombinator.com