3 ms·
> I really thought this misinformation video people found a loophole to not have to pay for what bandwidth actually costs. I really thought that Google had offe
by smichel17 6y ago
> I really thought this misinformation video people found a loophole to not have to pay for what bandwidth actually costs. I really thought that Google had offered a free CDN for everyone, limitless, when they (and Amazon, and Microsoft, and Akamai) charge huge amounts for this as a cloud service.
> Yeah, that was sarcasm.
Maybe so, but I came across a streaming site (of questionable legality) that did exactly this. Loaded an iframe with what looked like a random domain (always two first names, like "PedroMaria.xyz"). Then the iframe made a bunch of XHRs to load chunks of movie data from Google Drive. It was really quite fascinating.
Aside, this is one of my favorite things about using uMatrix with aggressive defaults -- as you unbreak sites, you learn a lot about what services are commonly used to build websites.
- knorker 6y agoYeah, I'm not surprised. Having worked for multiple service and content providers I would assume that abuse like this happens. One notable case was when a deal to license a database fell through when the would-be client said "I've found a different free way to solve this". Turns out that "different way" was to scrape our website in real time and present it to their clients as their own, without attribution or payment. Abuse happens. Shoplifting happens. That doesn't mean that the shoplifter "found a loophole", nor that it's morally or legally right. And it's not unfair if the selective enforcement is based on "bandwidth actually used" or "in addition to violating our policies you're also hurting the world" (or even "we just plain don't like you"). I.e. in the case above about scraping our data, we just dropped it. They weren't going to be our customer, clearly, and suing (they were in a different country too) or technically blocking them would not be worth the cost. And they didn't consume that many resources at our scale, so whatever. But if that site had been a porn site or something then the cost of doing nothing would include potential PR issues of being associated with them[1]. So in short: This (CDN from article, and using google drive like this) is abuse. A proof of concept likely won't get shut down, but it's not something that'll work in the long run. [1] I'm not against porn as such, but let's say our database was one of children's toys. That kind of company would understandably be hurt by apparently making deals with porn sites to show up there.