Speedtests from Ookla becoming increasingly useful

Forums › LAN (ethernet/Wifi/routers/Bluetooth) › Speedtests from Ookla becoming increasingly useful

jonherries

1433 posts

Uber Geek
+1 received by user: 316

Trusted

Subscriber

#61936 26-May-2010 11:38

This was posted on Slashdot

http://tech.slashdot.org/story/10/05/25/1924235/Global-Last-Mile-Performance-Stats-Going-Public

Here is the NZ data

http://www.netindex.com/download/2,5/New-Zealand/

Might be an interesting step to see what you have compared to the averages...

Jon

Website: herri.es

Linkedin: jonherries

LennonNZ

2459 posts

Uber Geek
+1 received by user: 411

ID Verified

Trusted

#334999 26-May-2010 11:53

I wouldn't trust the Location Details at all.
Taupu being the fastest place in NZ?
Hamilton being one of the slowest?

Epitiro had Hamilton being one of the fastest places.. so you can see someone (or both) are wrong.

Thanks

jonherries

1433 posts

Uber Geek
+1 received by user: 316

Trusted

Subscriber

#335004 26-May-2010 12:08

Hmmm, I noticed that but also checked the sample size for Taupu (sic), and there are only 282 tests in the index vs Hamilton at ~4000.

I suppose the value in this site is that it reflects real world testing by end users. Not sure about Epitiros measurements (they may use similar data).

Jon

Website: herri.es

Linkedin: jonherries

vexxxboy

4336 posts

Uber Geek
+1 received by user: 2072

#335007 26-May-2010 12:17

LennonNZ: I wouldn't trust the Location Details at all.
Taupu being the fastest place in NZ?
Hamilton being one of the slowest?

Epitiro had Hamilton being one of the fastest places.. so you can see someone (or both) are wrong.

Thanks

i thought the same thing , i would have thought Rotorua would have been there, i cant see them been lower than Wanganui, as at least a third of Rotorua has ASDL 2

Common sense is not as common as you think.

LennonNZ

2459 posts

Uber Geek
+1 received by user: 411

ID Verified

Trusted

#335009 26-May-2010 12:20

jonherries: Hmmm, I noticed that but also checked the sample size for Taupu (sic), and there are only 282 tests in the index vs Hamilton at ~4000.

I suppose the value in this site is that it reflects real world testing by end users. Not sure about Epitiros measurements (they may use similar data).

Jon

The problem is how do speedtest.net know where someone is? ISP's don't give this away. User A with IP address x.x.x.1 user may be in Auckland and user B with IP address x.x.x.2 may be in Christchurch so you cannot base it on IP Ranges.

Epitiro have physical DSL connections in the Areas they test so they know where they are when they report it.

Thanks

Ragnor

8279 posts

Uber Geek
+1 received by user: 585

Trusted

#335041 26-May-2010 13:40

Geo location by ip address is horribly inaccurate, useful only as a rough guide.

jonherries

1433 posts

Uber Geek
+1 received by user: 316

Trusted

Subscriber

#335045 26-May-2010 13:52

Unless you have mapped your ip address to skyhook, or confirm your location via locations services in your browser (think iphone and gps).

Jon

Geekzone

Want to support Geekzone and browse the site without the ads? Subscribe to Geekzone now (monthly, annual and lifetime options).

raytaylor

4076 posts

Uber Geek
+1 received by user: 1296

Trusted

#335980 28-May-2010 23:43

Whats even more interesting
Recently in my router, i have found that if i redirect any random250x250.jpg and various other files it uses to my internal lan http server, it reports the speed at 30 to 70mbits, and seems to also send this result back to the speedtest server as an auctual result for using in their statistics.

So each time someone on my lan does a speed test, call plus slowly climbs the chain towards NZ's fastest ISP. Upload speed however still stays the same because it is still going up the DSL.

Funny that

Ray Taylor

There is no place like localhost

Spreadsheet for Comparing Electricity Plans Here