[Update] ALL 20>30Mb upgrade discussion

[Peter_]

Originally Posted by qasdfdsaq

Now you've had two people clearly prove you wrong with evidence, I think you're the one who needs to rethink how the test works.

They will not balance out as they will be run over different times, which are not completely overlapping. Time is completely relevant.

I like how you've completely ignored how I've already explained this and shown you it in practice.

Then why do they balance out when running at the same time on mine and most other machines and the test is fairly similar in results to most other test out there and as I said earlier when we involve 2nd line they request you to run these files so they can check on the uBR.

Remember I have not set these tests, they are set by our IT department because they know exactly what to expect from this test and if you or anyone else feels it is incorrect then you need to take it up with them.

So as it stands you ring in with a speed issue, we run Gamefiles and ascertain you are running slow, we contact 2nd line and they have us get you to run some more files.

They logon to your uBR and check your download and upload plus available bandwidth, they then inform us of the result, if the is an issue they advise us of what needs to be done such as raising it as an area fault or booking an engineer sometimes with a request to call 2nd line while on site.

The is little point in arguing how a company checks its customers speed because that is the process we must use, if you refuse to download any files then the is little we can do to help as we cannot book an engineer without evidence.

[qasdfdsaq]

Originally Posted by Ignitionnet

See above - it's supposed to be a snapshot of when all files are downloading simultaneously and have had the chance to ramp up.

Obviously not ideal but you are getting the wrong impression if you are looking at speed at the end of the download.

I am taking a snapshot while they're all downloading. The problem is IE takes an average of speed throughout the whole download, including at the beginning when only one download is running. Averaging it out over 30-45 seconds is not enough to remove this effect (as shown above, starting the downloads as quickly as I can still results in speeds of 60-70mbps measured).

[Peter_]

Originally Posted by Ignitionnet

You're supposed to have a snapshot of the speed of all of them, once settled down, when all 4 are downloading simultaneously not their average speeds when all done. That would indeed be a futile measure.

Again all 4 must be running at the time you check their speeds, and all 4 must have had time to ramp up.

---------- Post added at 12:35 ---------- Previous post was at 12:34 ----------



See above - it's supposed to be a snapshot of when all files are downloading simultaneously and have had the chance to ramp up.

Obviously not ideal but you are getting the wrong impression if you are looking at speed at the end of the download.

That is it in a nutshell.

[qasdfdsaq]

Originally Posted by Masque

Then why do they balance out when running at the same time on mine and most other machines and the test is fairly similar in results to most other test out there and as I said earlier when we involve 2nd line they request you to run these files so they can check on the uBR.

You have not stated how you do the test or what browser you are using. I've already told you these are major factors that wildly affect the results. I've already told you certain versions of Firefox works, IE doesn't. But again you seem to be in complete denial of the facts and just be rambling on about "it works for me it must work for everyone".

Remember I have not set these tests, they are set by our IT department because they know exactly what to expect from this test and if you or anyone else feels it is incorrect then you need to take it up with them.

You're continuing to advocate a test that has been scientifically proven to be absolute rubbish and in complete denial of the facts because "it works for you".

So as it stands you ring in with a speed issue, we run Gamefiles and ascertain you are running slow, we contact 2nd line and they have us get you to run some more files.

Or you ring in with a speed issue, your flawed test says I'm getting 50mbps when I'm actually getting 20, you deny there's a speed issue, and the customer hangs up disappointed after being lied to.

The is little point in arguing how a company checks its customers speed because that is the process we must use, if you refuse to download any files then the is little we can do to help as we cannot book an engineer without evidence.

Yet in another thread you argue the complete opposite, it's all about what you can do, not the company. Quit selectively ignoring facts when it selfishly suits you.

The test quoted is rubbish, and (I would hope) you have sufficient free will and brain power of your own to interpret speed test results yourself. If speed tests suggest a problem, you can ask second line to confirm. Telling a customer to perform this flawed and inaccurate test accomplishes nothing, other than deceiving them to get them off the phone and hiding the problem.

"It's company procedure" to deceive and lie to customers does not cut it. Please. You are not on this forum as a representative of the company. You have your own mind and I certainly hope you'd have a basic grasp of math and computing fundamentals.

---------- Post added at 13:03 ---------- Previous post was at 12:59 ----------

Originally Posted by Masque

That is it in a nutshell.

This is assuming both me and Chrys are incapable of following basic instructions.

Now there's two ways to interpret this
a) We're both too dumb to follow instructions and are doing the test wrong.
This makes the test flawed, as in if we're both too dumb to get it right, how on earth do you expect your average joe to do it right?

b) We're doing it right and Ignition is incorrectly assuming we're doing it wrong.
This makes the test flawed, as we've demonstrated correctly that the numbers are far off.

In either case, the test is useless.

Originally Posted by Ignitionnet

See above - it's supposed to be a snapshot of when all files are downloading simultaneously and have had the chance to ramp up.

Obviously not ideal but you are getting the wrong impression if you are looking at speed at the end of the download.

Ignition and Masque, please read:

Originally Posted by qasdfdsaq

[Edit]
Just did it again, IE8, following instructions to the letter, but left downloads for 45 seconds instead of 30. Cleared browser cache, started 4 downloads of >400MB files, left downloads "to settle" for 45 seconds and took a screenshot to snapshot speeds. 57.84mbps "data related to the files being transferred". Add on 10% as they instruct, that'd be 63.62mbps.

[Chrysalis]

ignition yes but all cannot be started at the same time, there is human reaction delay between starting them. (they were all running at once by the way just not started at exact same time). The 10 second delay probably emulating a typical end user who is probably unsure of what they doing whilst talking to an agent on phone.

also browsers do report average speeds, nothing I or any other end user can do about that its how they work. Unless of course using something like dumeter to monitor live bandwidth, but VM dont tell people to do that.

[Peter_]

Originally Posted by qasdfdsaq

You have not stated how you do the test or what browser you are using. I've already told you these are major factors that wildly affect the results. I've already told you certain versions of Firefox works, IE doesn't. But again you seem to be in complete denial of the facts and just be rambling on about "it works for me it must work for everyone".

The test works for the majority and it is the test that we must use and as you may have read even Ignitionet has told you how it works and I respect his knowledge of our systems over virtually anyone else on this forum.

Originally Posted by qasdfdsaq

You're continuing to advocate a test that has been scientifically proven to be absolute rubbish and in complete denial of the facts because "it works for you".

By you I think not.

Originally Posted by qasdfdsaq

Or you ring in with a speed issue, your flawed test says I'm getting 50mbps when I'm actually getting 20, you deny there's a speed issue, and the customer hangs up disappointed after being lied to.

If they have an issue it will flag up on my systems and the speedtest is not the only tool at my disposal.

Originally Posted by qasdfdsaq

The test quoted is rubbish, and (I would hope) you have sufficient free will and brain power of your own to interpret speed test results yourself. If speed tests suggest a problem, you can ask second line to confirm. Telling a customer to perform this flawed and inaccurate test accomplishes nothing, other than deceiving them to get them off the phone and hiding the problem.

As already stated 2nd line require large files to be downloaded so they can check the uBR while they are running, did you miss that bit.

Also the is little point in just getting a customer off the phone as that just causes a repeat.

[Chrysalis]

Originally Posted by Masque

That is it in a nutshell.

dont get too excited, all 4 of mine were running.

if they wasnt I would have been posting 4x 3.8meg/sec.

[qasdfdsaq]

Originally Posted by Chrysalis

ignition yes but all cannot be started at the same time, there is human reaction delay between starting them. (they were all running at once by the way just not started at exact same time). The 10 second delay probably emulating a typical end user who is probably unsure of what they doing whilst talking to an agent on phone.
.

This is it in a nutshell.

If human reaction time can skew the results by 15-20mbps when done by a technically competent user clicking quickly and starting the downloads as fast as the server will allow, then for a slower user the effect can only be worse.

Additionally if the results can differ by 20mbps between different browsers (and is worst in the most commonly used browser) while actually transferring at the same speed then clearly there is something wrong with it.

This appears to be going round in circles - how about agreeing to disagree?

[Peter_]

Originally Posted by Chrysalis

ignition yes but all cannot be started at the same time, there is human reaction delay between starting them. (they were all running at once by the way just not started at exact same time). The 10 second delay probably emulating a typical end user who is probably unsure of what they doing whilst talking to an agent on phone.

also browsers do report average speeds, nothing I or any other end user can do about that its how they work. Unless of course using something like dumeter to monitor live bandwidth, but VM dont tell people to do that.

They do not have to be started at the same time ever! as all we want is for them to level out to similar speeds at which point we multiply them together and ascertain their speed.

[Chrysalis]

Originally Posted by qasdfdsaq

I am taking a snapshot while they're all downloading. The problem is IE takes an average of speed throughout the whole download, including at the beginning when only one download is running. Averaging it out over 30-45 seconds is not enough to remove this effect (as shown above, starting the downloads as quickly as I can still results in speeds of 60-70mbps measured).

correct, many of those files in truth are way too small to be testing 30mbit and higher in a browser. I would say dvd iso images are more suited but then it gets a bit heavy on bandwidth.

---------- Post added at 13:16 ---------- Previous post was at 13:15 ----------

Originally Posted by Masque

They do not have to be started at the same time ever! as all we want is for them to level out to similar speeds at which point we multiply them together and ascertain their speed.

they wont level at true speeds tho.

you simply fail to understand that there will be inflated speeds for the first 3 files as they are contending with more than 1/4 of the bandwidth. its actually reasonably diffilcult to get 4 files going before the first finishes as well. unless I pick one of the larger files deliberate to start with.

as hugh has asked I will agree we disagree. Shame yet again tho VM staff never ever backing down on anything, their process is always right etc.

[Peter_]

Originally Posted by Chrysalis

they wont level at true speeds tho.

you simply fail to understand that there will be inflated speeds for the first 3 files as they care contending with more than 1/4 of the bandwidth. its actually reasonably diffilcult to get 4 files going before the first finishes as well. unless I pick one of the larger files deliberate to start with.

as hugh has asked I will agree we disagree. Shame yet again tho VM staff never ever backing down on anything, their process is always right etc.

No it is just you failing to understand even when ignitionet tells you the same thing.

[qasdfdsaq]

Originally Posted by Hugh

This appears to be going round in circles - how about agreeing to disagree?

Agreed.

I've shown the test to be inaccurate in many scenarios

Masque has shown the test to be accurate in many scenarios

I put this down to difference in browsers and difference in understanding of how the test works. I know for a fact that different browsers *do* report wildly different results using this methodology.

Ignition has wrongly assumed we're doing the test incorrectly whereas I have clearly already stated I'm doing what he assumed I'm not doing. His point is therefore not valid in this context. It is not incorrect, but based on a wrong assumption that does not apply here.

Personally I do not trust this test and will not trust anyone who tells me to do this test. Any test that gives me a result of 70mbps on a 50mbps connection is deeply flawed. No, I am not secretly on 100mb, as much as I wish I was.

I could explain this fully and provide proof in terms of mathematical formulas graphs and spreadsheets but if you want that... we should start another thread.

[Peter_]

Originally Posted by qasdfdsaq

Agreed.

I've shown the test to be inaccurate in many scenarios

Masque has shown the test to be accurate in many scenarios

All that really matters to be honest is that it provides large files to download so our 2nd line guys can logon to the uBR and check what is actually going on with your connection.

Also it is best to use Internet Explorer or Firefox is checking your own speed using gamefiles as they download showing the transfer rate, but it does not matter what browser you use if 2nd line are checking as they just see the live results from the uBR.

Really, guys - get a room (or another specific thread)......