Pgoodwin1

This chart is from this morning This one is from mid November on the Central Server As you can see (upper photo),when on Auto, the results are consistently lower than typical past Central server results (lower photo). Also in the upper photo, not that the East Coast server speeds are double the Automatic server mode speeds, and about 50% higher than the typical in the lower photo. Both photos represent typical -upper photo from today Jan 10, 2013, lower photo from Nov 19-20, 2012

Thanks TriRan. Is there a way to determine which test server is being used when in Auto Route?

I see a temporary banner pop up that says it's using 23 different data centers. I used to use a default Central server, and have about a year's worth of test data averaged so that I know what "Normal" looks like. It seems like I'm getting about 1/4 to 1/3 slower times using the Automatic Routing now than in the past on both the East Coast and Central servers. I've been having problems with my iPad Safari browser not finding the Central server. I wondered if the Automatic Routing results are getting skewed down somehow because of that issue. I would just like to understand at a high level what the Automatic Routing is doing. What's strange too is that wen I choose the East Coast server lately, I get download speeds that lately have averaged 50-60% higher than my past experience. Did they rebuild the entire East coast Internet since Sandy? haha.

CA3LE. There's something odd/different about the Dallas server from a couple of weeks ago. I mentioned above that my iPad Safari couldn't find the server when I selected Central. when I select East Coast, and use the Test Now button on the testmy nome page, it runs fine. I actually am getting speeds using the East Coast server that are consistently about 50% to 90% higher than I've typically gotten on the Central server over the past year. As for the Central server, after rebooting my cable modem, and my Router, and after flushing Cookies and Data on the iPad, I was able to connect to the Central server. I ran a few combined tests, and the speeds were about 25% lower than what I used to see on that server. Then I went back to the East Coast server and got the higher speeds I was seeing earlier. Then I selected the Central server again. And Safari again said the server couldn't be found. I flushed the cookies and data again but that didn't fix it. I've never had this happen in the past. Any thoughts? above I called the Home page test button the "Test Now" button. It's actually called the "Test My Speed" Button. I use that, then use the Re-Test button in the current test results page for another test.

Welcome from SW Ohio

Thanks.

CA3LE....do you recommend that we put the test server location on "Automatic" now to get the most realistic results? This evening (and a few evenings ago) I was getting significantly higher download speeds using the Washington DC server. I'm in Cincinnati, OH. I used to leave it on the Central (Dallas) server - it was called my default. This evening, when I chose the Central server, the page wouldn't open (got a server not found error). don't know if that was just a temporary outage or not. It was working on Central earlier today. I'll try clearing my browser History and try again. Just a note...Clearing the iPad Safari history didn't bring the Central server back.

Test results today on East Coast and Midwest servers: I live in the suburbs of Cincinnati, OH. My default testmy.net server is the one in Dallas. I tried the East Coast server today for grins. It's averaging about 60% faster on downloads today (12.7 Mbps vs about 8.0 ). Uploads were about the same (I'm Time Warner RoadRunner limited at 1.0 Mbps and typically get about 0.8 ). I don't remember ever having tried the East Coast server before. Distance wise, the two locations are similar from where I am-maybe slightly closer to the East Coast. I've never gotten a single download datapoint on the Midwest server above about 12 Mbps, and ones near 12 are rare. On the East Coast server this morning I hit a little over 15 Mbps (almost double my typical Midwest throughput). And it was very rarely below about 9 Mbps. I guess I'll leave it on the East Coast server and see if results over time are consistent. I normally use results to look for dips in performance so using the Midwest server is still useful for identifying Time Warner problems. Is the default server chosen by testmy.net or is that a personal preference? Is there any reason I should leave it on the Midwest server.

Hi from SW ohio

From CA3LE " .... When I've seen this it never pauses, then turns into a fluid progress bar (meaning that it moves smoothly). It will pause then JUMP... as if the browser couldn't keep up with the java used to update the progress. " My iPad behaves like that. The progress bar updates roughly every 2 seconds or so. It doesn't bother me. I have a new talent....counting and calculating Mbps. My download speeds normally vary between 6 and 10 Mbps. My average over a year is about 8 Mbps, and the test sizes are usually about 8 MB. When I hit the start button, I count one thousand, two thousand...until the progress bar completes. During the upload test, I've gotten good enough that I can now calculate the download test result in my head to within 10%. The mind is a terrible thing.

Welcome for SW Ohio

As Ninja states, they all over sell. As I've stated elsewhere in TestMy, all of them do it so they've collectively used overstated specs on their speeds. They're not afraid of getting sued, because it has become the norm, and on their instruments, that's what you get, regardless of the reality of it. I typically get only about 80% (8Mbps as measured here) of what they tell me I'm supposed to be getting for download speed (10 Mbps min), and only about 1/3 what their speed measuring tools show (>20Mbps).

Virtually anything (no how bad the design) can be gotten used to. I have only Macs at home but at GE before I retired, I had to use a PC. One thing that drove me nuts was that it seemed like no matter what I was trying to do, it took twice as many mouse clicks as it did on the Mac. And the thing that drove me the most nuts was that whenever I was finally able to find something command related, that I needed to use often, was always buried deep down in some ill-organized hierarchy of menus or buried deep down in some control panel that only had the loosest connection with the function you wanted to command. The MS organizational brain is not like mine. Maybe it's me. But the Mac OS and menu structures are more intuitive - to me anyway. Like I said, if you use something enough, you learn it and it isn't as painful. But when using the PC, after finally finding what I was looking for, my comment to myself was always "why in the hell would ANYBODY ever put it there?", and "what design review committee decided THAT was where it should go?" MS ruined Office with their whole menu redesign. It's an atrocity. I will never get used to that piece of crap.

HAHAHAHA mudmanc4. There are no real zombies. But there is MS. It eats a little bit of your brain every time you use it.

My real download and upload speeds (measured here) are always about 1/3 of what the Time Warner Roadrunner speed test shows. Their results at times have told me my service was fine when in fact there were signal problems at their amps outside. They and most ISPs advertise falsely. The only thing you can say is that they typically deliver their promised rates as measured on their instruments. When you read the TestMy explanation, all this you will have a better understanding. The reality among many ISPs is that they (as a business) know they are advertising falsely, but each one is advertising not any differently than most other ISPs, so they feel there's a built in immunity to legal action because everyone is doing it. Some are undoubtedly worse than others. In your case, if they "guaranteed " you 20MBPS down, and you're getting 10, you're better off than I am. The only way to fight all of this would be a class action lawsuit. The only solution would be to create a set of specifications for speed measuring (similar to what they did with audio power amplifier power measurements). Then an ISP could either be compliant to the spec or standard or not, and the buyer could decide if he wanted to go with one that didn't claim compliance, or one that did. Right now it's a mess. I don't know if the typical support person you talk to knows anything other than the tools he was given to measure speed at that company. If he/she does know, they're between a rock and a hard place, because they can't tell a customer that their tools are wrong.

Hahahaha

I wonder what Bill Gates really thinks of all this.

Steve Ballmer has gone insane. He's telling his stockholders that the demand for Win8 is "fantastic". I'm not a MS basher at all. I think their Office was one of the greatest pieces of SW ever created (until they screwed up the toolbars). An although I use Macs exclusively, I have great respect for what Windows has been. But MS seems to be floundering badly, kind of like Apple was when Steve Jobs was gone. I can't vouch for how accurate this article is, but it's an interesting one: http://www.bloomberg.com/news/2012-12-04/microsoft-intel-push-to-combat-apple-in-tablets-sputtering-tech.html

I still have my original iPhone 1....every once in a while I take it out and just look at it and hold it. It was a marvel.

The picture isn't clear enough for me to read it.

Are those speeds average results or one time test results? I typically get speed test results that vary by 20% to 100% from test to test. Not because of the testmy site or my hardware/OS but because of what's going on out on the Internet. The right 1/2 of the plot above was taken over a period of about 10 minutes. Some points were using Safari, and some were using Mercury (Firefox on iPad). Time Warner Cincinnati speeds are always all over the place from minute to minute but generally my averages are 8 down, and 0.8 up. This morning Time Warner's loading seems to be pretty varied. Speeds in the middle of the night usually vary by 20% or less from minute to minute. And I get similar results to this iPad on my iMac running 10.8, my wife's MacBook Pro running 10.6.8

Hi mikec.....from SW Ohio. I did some coding on the 6502 also...in college. And got to do some on HP's early desktop computers....in HPL.....OMG what a language.

Hi and welcome from SW Ohio

Happy belated birthday. 31 is a great age to be.

Welcome from SW Ohio