4 ms·
Xfinity is most likely committing data usage measurement fraud due to their implementation of this banner. I asked the person responsible for the banner if it
by Sephr 7y ago
Xfinity is most likely committing data usage measurement fraud due to their implementation of this banner.
I asked the person responsible for the banner if it counted towards data caps and was ignored. https://twitter.com/sephr/status/941067958096244741 https://twitter.com/sephr/status/941067958096244741
- grepsedawk 7y agoRT'd
- WillPostForFood 7y agoThey typically don't count traffic within their network towards the data cap (e.g. streaming Xfinity TV), so I don't think it is safe to make an assumption about the banner one way or another.
- grepsedawk 7y agoIt's delivered alongside the webpage so there's no technical way to split the data logically... Unless they're using "rough math".
- diminoten 7y agoYes there is, it has to come from somewhere, and we know it's not coming from the actual site being visited...
- Havoc 7y ago>it has to come from somewhere Very much doubt they count their traffic by adding up all the origin addresses contributions individually
- diminoten 7y agoYeah, I doubt any of it counts at all, and honestly this is the least compelling part of this argument... It's kb vs. 100GB, hardly matters.
- grepsedawk 7y agoNope... The packets need to be sealed back together, at that point the data is denormalized and you can't measure how much it "actually" counted.
- diminoten 7y agoYes? You can see how much you added, the size before and the size after. This is not a technical impossiblity, I'm sorry if you're invested emotionally in it being impossible.
- tedunangst 7y agoHuh? I happen to run my own intercepting proxy which rewrites HTML pages, and if I wanted to, I could easily track exactly how many bytes came in and how many bytes went out.
- cortesoft 7y agoWhat? They just have to measure bandwidth from the other side of whatever is modifying the content.
- milankragujevic 7y agoThey could measure the size of the JS payload, and subtract that from the size of the web page, before adding that number to the number of bytes used in a billing period. That way they could "more-fairly" measure bandwidth usage even with their MITM "value-add" and "informative" content. Though, seriously, I have a hard time understanding the reasoning for data caps on DOCSIS infrastructure. On LTE, yes. WISP, yeah - kinda. DOCSIS, DSL and GPON? Absolutely no! Not to mention the horrifying realization of most uninformed people that their ISP can and will intercept, log, modify or restrict access to content that the user has requested, even though the user has the right to such content, having paid the monthly subscription fee for the connection. But hey, I don't work at a large cable ISP, I couldn't possibly understand their reasoning and advanced calculations. /s
- Sohcahtoa82 7y ago> Though, seriously, I have a hard time understanding the reasoning for data caps on DOCSIS infrastructure. It really $houldn't be that hard to under$tand.
- guntars 7y agoIt’s my impression that the bottleneck isn’t the last mile technology (wireless, copper or fiber), but the interconnects between ISPs and the rest of the internet. This is why Netflix can be crawling in the evening while you’re getting the full speed on a speedtest (that’s not fast.com)