Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Except fast.com doesn't as many locations as speedtest.net; so the best I can get with fast.com is ~250 Mbps, while speedtest.net measures 900 Mbps.


Idk if you noticed, but with Speedtest.net often times you’re testing against an ISP-run or sponsored server so it’s hardly an accurate portrait of speeds with a typical service.


But you can select servers yourself. So I usually use the auto-selected server (sitting at my ISP or similar) to check local conditions such as is my cable modem or router/firewall doing fine or not. If I want to test "usable" bandwidth, I'll select a few servers in other, far-away cities like Amsterdam, Miami etc.


But that depends on what you’re trying to test. If you want to test your DSL/cable/fiber connection (and nothing else), it’s actually good if the ISP is hosting the test server.

If you want to measure the speed you’ll realistically experience in the wild, including your ISP’s peerings and whatnot, that’s another question.


Doesn’t it measure from the same locations as Netflix is delivered? I think that was the whole point of the service?


The main problem seems to be it’s HTML/JS based and just can’t keep up with high download speeds. Same goes for speedtest.net. Speedtest has a desktop app which doesn’t suffer from this issue.


I have well over a gigabit download speed on fast.com just now. I think it can keep up just fine.


It craps out at around 500Mbit on Safari/MacOS. Same on Chrome. The stand alone speedtest app goes up to >900 Mbit just fine. It randomly freezes during the test for seconds at a time, presumably the JS GC is triggered.

To be fair, I do have about 150 tabs open.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: