HTML5 Speed Test - check the 'Details' button when you are done.

[GreyGeek]

https://sourceforge.net/speedtest/

[oshunluvr]

Not sure what this actually means, but I'll bite:

Vivaldi with about 50 tabs open:


Pale Moon with no tabs open:



No appreciable difference here.

[GreyGeek]

Did you click the "Details" button and view the detailed report?

[oshunluvr]

I did, but everything looks real good - nothing but green checkmarks. Kind of interesting, but I don't really know much about what I'm looking at.

I did forget that I had my VPN up so here's a retest with it off:



Quite a bit faster!

But it illustrates how little speed matters for most internet access functions - once you're above a certain level, at least. Even with the VPN running I had enough bandwidth for any service I need. Of course, add-in 4 other people using my connection adn a movie remotely streaming somewhere and the extra capacity might come into play.

[whatthefunk]

Same here. Nothing but green checks in the details.

[vinnywright]

yup ,,,nothing but green ...



VINNY

[Goeroeboeroe]

This is a very strange test. To be honest: I think it's mostly meant to sell licenses.

Oh, I didn't see the price they ask for te test when I wrote the line above. If you want a licence to use this 'test' on your site, there's an annual fee of 980 dollar (that's no typo: ninehundred and eighty dollar every year).And a one time fee for set-up, also 980 dollar.
(You can see the price if you click on the question mark in the testing square.)
That's really ridiculous! Internet is full of free connection testers. 980 dollar each year. unbelievable. And 980 dollar for set up.
Okay, that's the price.

They call it a html5 speed test. 'html5' is a real buzzword. But speed has nothing, absolutely nothing, to do with html5, html4.01 strict, xhtml, css3 or whatever technique you use. html5 is simply a language to code a website, it has nothing to do with the speed of your connection. A bad site can be (very) slow, but that's not the connection: that's a bad coder writing bad code.

They write almost every speedtest uses Java or flash. Not true. Most tests never used flash, and today probably no test at all, because flash is rapidly disappearing everywhere. Flash is simply not the right language for a speedtest.

Java was never used for speedtests, because it's not very usable for a speedtest. And today it's impossible to use it, because most browsers don't support it any more. Most people don't know the difference between JavaScript and Java. JavaScript is used for this kind of tests (among other techniques), but this site also uses JavaScript. If you disable JavaScript, you can't run the test.
(JavaScript has NOTHING to do with Java. The 'Java' in the name is there because of historical marketing reasons. 'Java' was popular, so Netscape renamed it from Mocha via LiveScript into JavaScript. But only because 'java' was popular, it's really a completely different language.)
Why they use this confusion between Java and JavaScript is unclear to me, but it doesn't make them trustworthy.

All together: using the buzzword html5 for a test that has nothing to do with html5, suggesting it's one of the few test not using Java or flash (not true) and asking 980 dollar YEARLY (and 980 for setting up) looks to me like a scam to let people pay for something that's available for free in lots of places.

CORRECTION:
After a quick check if what I wrote was accurate (did it out of my head): I was wrong about the flash. There were indeed in the past some speedtests with flash. I forgot, because those tests are simple completely unreliable.

[vinnywright]

This is a very strange test. To be honest: I think it's mostly meant to sell licenses.

Oh, I didn't see the price they ask for te test when I wrote the line above. If you want a licence to use this 'test' on your site, there's an annual fee of 980 dollar (that's no typo: ninehundred and eighty dollar every year).And a one time fee for set-up, also 980 dollar.
(You can see the price if you click on the question mark in the testing square.)
That's really ridiculous! Internet is full of free connection testers. 980 dollar each year. unbelievable. And 980 dollar for set up.
Okay, that's the price.

They call it a html5 speed test. 'html5' is a real buzzword. But speed has nothing, absolutely nothing, to do with html5, html4.01 strict, xhtml, css3 or whatever technique you use. html5 is simply a language to code a website, it has nothing to do with the speed of your connection. A bad site can be (very) slow, but that's not the connection: that's a bad coder writing bad code.

They write almost every speedtest uses Java or flash. Not true. Most tests never used flash, and today probably no test at all, because flash is rapidly disappearing everywhere. Flash is simply not the right language for a speedtest.

Java was never used for speedtests, because it's not very usable for a speedtest. And today it's impossible to use it, because most browsers don't support it any more. Most people don't know the difference between JavaScript and Java. JavaScript is used for this kind of tests (among other techniques), but this site also uses JavaScript. If you disable JavaScript, you can't run the test.
(JavaScript has NOTHING to do with Java. The 'Java' in the name is there because of historical marketing reasons. 'Java' was popular, so Netscape renamed it from Mocha via LiveScript into JavaScript. But only because 'java' was popular, it's really a completely different language.)
Why they use this confusion between Java and JavaScript is unclear to me, but it doesn't make them trustworthy.

All together: using the buzzword html5 for a test that has nothing to do with html5, suggesting it's one of the few test not using Java or flash (not true) and asking 980 dollar YEARLY (and 980 for setting up) looks to me like a scam to let people pay for something that's available for free in lots of places.

CORRECTION:
After a quick check if what I wrote was accurate (did it out of my head): I was wrong about the flash. There were indeed in the past some speedtests with flash. I forgot, because those tests are simple completely unreliable.

when I tryed to retest ,,,it crashed ,,,closed the page and reopend it ,,,it crashed again ,,,,,I ran 2 other tests ,,,thay ran fine ,,,came back to this one ,,,,and it ran again .

the other 2 tests





VINNY

[Goeroeboeroe]

Hmmm, I haven't looked for some time at speedtests.
I see that's lots of them call themselves 'html5 test' today. That's really utterly nonsense, because there's simply nothing in html5 that can measure any kind of speed. it simply doesn't exist.
They all use JavaScript to measure the time. And JavaScript is in no way part of html5. You can use it together, but you can use PHP and html5 together too.
So it's not only this test that uses the buzzword 'html5'. It probably IS the only one asking 980 dollar per year to put their test on your site...

[WaltS48]

I had one red x for the Live Streaming upload speed test

The minimum upload speed threshold for Live Streaming is 5.00 Mbps and your average upload speed is 1.28 Mbps.

[wizard10000]

We're moving from central Michigan to Chattanooga, TN this week. Chattanooga has fiber available to about 95% of the area so when we called Comcast to move service there our account got reprovisioned from 50Mb to 100Mb before we moved

Gigabit internet in the area costs about $70 a month. I think Comcast gave us the speed bump to keep us from bailing on them

[GreyGeek]

Shouldn't you be getting 100Mbps download?
Too bad it is asymmetrical.


What does the DL bandwidth look like. Does it build to max almost immediately or does it do an exponential buildup? The latter means, in affect, that your actual DL speed is about 1/3rd of the promised speed.
Spectrum is playing that game around here. That's why I'm moving to 100Mbps on Allo, which is symmetrical and instant max bandwidth.

[wizard10000]

Shouldn't you be getting 100Mbps download?
Too bad it is asymmetrical.


What does the DL bandwidth look like. Does it build to max almost immediately or does it do an exponential buildup? The latter means, in affect, that your actual DL speed is about 1/3rd of the promised speed.
Spectrum is playing that game around here. That's why I'm moving to 100Mbps on Allo, which is symmetrical and instant max bandwidth.

Couple thoughts -

I ran that speedtest over a wireless connection with about 60% signal strength; that missing 10Mb could have come from my end. The reason I still use that Ookla site is that it's one that lets me choose the target server; since I'm on Comcast's network in central Michigan I use Comcast's POP in Detroit as the target location. Also, Comcast says "up to" in all their speed ratings, and 100Mbps isn't even available in my area

DL (and UL) speed don't ramp up, they're available immediately. Takes 2-3 seconds for a speedtest to stabilize, but they don't appear to throttle anything.

I do use OpenDNS (hardcoded into my router) so Comcast doesn't get quite so much of my information, but TBH I've been happy with their internet service but slightly less than enthusiastic about their TV service when compared to DirecTV or Dish, but the spousal unit switches our TV provider when each contract ends

Will probably ditch Comcast and go fiber when contract is up next year - the local power company offers a TV and Gigabit internet package that has a better TV selection for less than I'm paying now.

[GreyGeek]

Vinny had an image of his Spectrum download test in another thread. It shows the classic exponential build up in speed they use to avoid full speed from the get go. It takes about 5-10 seconds to get to advertised bandwidth and by the fifth second the page is usually loaded, which averages out to around 5-10Mbps, so the user doesn't benefit from the full bandwidth speed.

[GreyGeek]

Well, the Allo tech just left and I've been playing with my shiny new 100Mbps Internet connection. Here is the speed test: