CA3LE

With a title of simply, "yup" people who are searching have a hard time finding this. It was changed so that more people will read your post and you'll get more response. I only very rarely do that... but I will especially change the topic title if it contains only 3 letters. "Yup" doesn't provide users or search engines with any relevant information as to what this topic is about... you do want people to read and respond to this, don't you?

Here's a Steve Job's rank on the whole Fragmented vs Integrated issue. I personally feel that both platforms are awesome. My girlfriends dad got an iPhone 4 and after playing with it I'm very impressed. I have a Motorola DriodX and love everything about it. If I were offered a trade, even if the iPhone was on Verizon's network... I'm not sure if I would do it because of how happy I am with my phone. Now, one thing to understand is that Google has left their product open to competitive distribution. Buying an apple product is cool because you know what you're getting... you know that your getting quality hardware backed by quality software. But that doesn't mean that the same can't be true for Android. If you buy a quality model you're going to be happy. I have seen people with problems with Android but it's usually on lesser models, I'm sorry.. but you get what you pay for. Having said that, I think the approach that Google has taken is cool. They leave the public open to options... and the market open to competition. Apple on the other hand... well... your only real option is how much storage space you want. BUT, as I said... it's a quality piece of hardware. For me... Android is the best option. How does it impact users... well, for starters on iPhone.. when you download an app chances are that it will work as intended because there are far less variables. But on Android the same can be true... just make sure you get a good phone, lol. Like I said, I have the DroidX, I haven't seen any apps that didn't work properly. My girlfriend has the HTC Droid Incredible... her phone also works perfectly. Just don't buy a piece of junk and you should be happy with it. -- READ REVIEWS BEFORE YOU PURCHASE. -- An iPhone on the other hand, for the most part, is an iPhone... Android or iPhone... whatever you buy will be outdated in a week anyways

You have to realize that their test is on their internal network. So if there is congestion along Comcast's route to TestMy.net you won't be able to pull those speeds. Also, our testing method can help detect browser and OS issues that most other tests take a path around. For instance, my Chrome install is completely bugged right now, my scores here reflect the issue BUT testing at other sites it's not noticeable (I'm using Firefox until I figure out what's up with Chrome... it was fine a week ago)... One thing to keep in mind is that as connections get faster it's becoming more apparent that TestMy.net isn't just an internet connection test. Yes, it tests your internet but because of the method I employ if other things aren't running correctly it will be reflected in your speeds here. Another example of this that I like to bring up is in the case of my girlfriends laptop... she has a 2006 Macbook, her scores here were <8000 Kbps. I realized that her HDD was the component that was slowing her down. Swapped that out and immediately after she was able to max our connection out. All other speed tests didn't catch that. I don't know... nor do I really want to know how they all do it... all I know is that I seem to have inadvertently created more than just an internet speed test over the years. You may want to check your Rwin and MTU settings. I had a friend the other day who said, "Man, I can't pull more than ~20Mbps on your site... everywhere else shows nearly 60Mbps... And I can pull my full connection on Usenet..." -- he's on the same host as me and in the same city so I was pretty sure his route wasn't an issue. I let him know that just because you can get your full connection on Usenet doesn't mean that there isn't a problem... multi-threaded downloads like that can be unaffected but there can still be an underlying issue. He then remembered that he just recently reinstalled his ethernet drivers... further investigation showed that his receive buffer was set too low... changed it and was immediately able to max out. After I release v11 I want to work on giving the end user a better idea of what their problem might be... it's thousands of programming hours away however... and a ton of reading and research needs to be done on my end to better understand exactly why my tests detect the things they do (when nobody else does). So, I'm sorry that I don't have an exact reason for you right now. ... And it could always come down to Comcast's route to TestMy.net. Comcast did have a trunk outage a couple of weeks ago (see https://testmy.net/ipb/topic/27915-comcast-trunk-outage/) but I was informed that it was resolved. Let me know if you find some underlying problem. Hopefully some of the examples I sighted may help you. Here's suddenlink.net's averages >> https://testmy.net/hoststats/suddenlink.net They're currently ranked #3 for average download speeds here.

Is Sudden Link a cable provider... where are you at. Nice speeds.

I haven't had a chance to read over your whole post... about to leave my house. But DAMN, 7Mbps off satellite is awesome. Can you sustain that over a larger download though? I saw a test of yours that was 25MB @ ~5Mbps... but I don't know what the conditions outside were. Post up results from a larger test with clear skies if you could please. - CA3LE

How do you become a BETA team member?

For now, you could just send them here and have them test off our servers... you said it yourself... ... Wouldn't the same be true for your customers.

Well thank you. Yes, this is all my creation. All written from scratch... the forum software is the only thing that I didn't design. I use templates when I create sites for other companies but never with TestMy.net. TestMy.net is the only fully PHP based test of it's kind. The software on the server-side is a trade secret so I have to re-create a license version that will allow people to host it on their own server while still protecting TestMy.net's interests. Once this is online people will be able to sign-up for it 100% free of charge and have a clone test running on their server within minutes. These cloned tests will have their own separate database logging as well as all other TestMy.net features. A version with full control over the ads will be available for a monthly fee soon after it's release. I have half the code written in my head I just need to finish the prerequisites before I can dig into it. What ISP do you work for? What are the server specs? If you'd like I can contact you as soon as it's available. Just shoot me a PM with "Clone Test" in the subject. - CA3LE

Yes, the software here is unique to TestMy.net. I used to offer remote testing but before the rebuild it was taken down for redesign of that system. This is planned for return either with version 11 or shortly after. Thank you for your intrest and kind words. Sorry that I don't currently have a solution for that but there are steps in the development process that are prerequisites for that system that aren't yet complete. -CA3LE

Check this out https://www.speedguide.net/read_articles.php?id=2574 And you may want to try the TCP Optimizer from SpeedGuide.net. I think an RWIN of 256960 is what you need. Try an MTU of 1460... then 1500 and keep the one that works best for you. I would think that 1500 would be best but I guess it depends on your satellite system really. Let me know if this helps you. And if it does help could you please explain the exact steps and settings that you took to arrive at better performance... this will help users in the future who may have a similar issue. Good luck, - CA3LE

First, here's a post I made about satellite latency last month >> Now, your latency seems quite high even for satellite. But really... I have NO first hand experience with satellite so I'm really not the best person to give you advice on improving it's performance... in terms of latency issues like you're having. I do however know of many users that have seen marked improvements with their satellite by using the .ccs files for Cablenut that were created by Vanburen. He made specific settings just for satellite connections. cablenut408.zip VanBurens_cablenut_settings_all_versions.zip ... I'm unsure if that will help your issue with latency but it doesn't hurt to give it a try. Here's another post that speaks of satellite latency that I found >> Does anyone else know of anything that can help? rayj0054, if you find a solution OR if those settings help you... please let us know.

Cool, thanks for sharing that info with our community.

A screenshot isn't necessary. You posted the information I needed. Your latency is massive.... really bad. 1.4 - 2.1 (1.6 avg) seconds is really bad. Unfortunately it's a reality of satellite. I'm pretty sure this is why Google Talk won't work for you. That may be a service that isn't very satellite friendly. You are running faster than 56K... but with that much lag in your transmission your conversation would be more like a walkie talkie than a telephone. ... Google Talk may be set to simply not work in that situation. Although, I'm surprised that the application wouldn't notify the end user of the problem.

I wonder if latency is causing your issue. What is your ping to Google, for instance. To ping a host... Windows Start > Run > type CMD [enter] > from the command prompt type ping google.com and/or ping testmy.net Linux / Mac Under Applications/Utilities open 'Terminal' > type ping google.com and/or ping testmy.net Paste those results here.

Very nice, looks like they're moving in the right direction.

Niiiice, we've got a contender

They still are, poor satellite users.

One of the things that's often times overlook with ISPs Quality Of Service is their ability to maintain good peering relationships outside of their network. Sending users to an internal speed test is a cruel trick that unfortunately many people fall for. "Testing off a server that is right next door does not give you a real world results. How could it? When you use the internet how often are the servers right next to you... rarely." I believe that there are many ISPs these days that cover their problems up with their internal speed tests. A user sees a problem, the ISP sends them to their speed test that is usually hosted within their own network or on the edge of their network. The user sees no problem and the ISP blames the other end. Now, don't blame the phone support guys... they're just doing what they've been told. But the people above them that established the practice know exactly what's up... they just won't admit it. Say you have a 100 Mbps line, most likely unless their is a internal problem with your ISP you'll get that speed until you leave their network. And if the ISP has quality peering then you can maintain that speed for thousands of miles (I'm well over a thousand miles from my server and I more than max out my Cox.net 50/5 connection). Take the same scenario, 100 Mbps... but the ISP has little or no good peering relationships. You go through your ISP's internal network out to the internet... You then get routed through Jim Bob's over congested trunk... and then through Billy Bob's congested trunk. It's like trying to stuff a watermelon through a garden hose. You will eventually hop onto some quality providers if you're going to testmy.net because of the great peering that my host has... but by then the damage is done. Now, your ISP is going to blame the hops in between ('Ol Jimbo and Billy, lol)... and they SAY that it's not their problem. BUT IT IS! They're friends with those guys, they have a working relationship. Your ISP doesn't own or control those lines BUT they DO control who they peer with. Something to think about. Post up a traceroute. Although a bottleneck along your route doesn't always show up on a trace.

Cheers, hope to see you around. Thanks for adding that... that's actually my preferred method on Windows machines. Dr TCP from DSLreports is another good one. Whatever the method the idea is the same. Optimize the TCP Stack.

First, thank you for the kind words k2000day. Secondly, as tdawnaz said... read the page I wrote about the legitimacy of our tests (link again is https://testmy.net/legit-speed-test.php'>https://testmy.net/legit-speed-test.php). I can't speak on behalf of any other site out there. I also don't like to talk negatively about my competition. But I can tell you this, I've heard countless times about the results at that site being way off... either too high or too low. People get confused about the inconsistency with the results. One of the reasons I no longer allow linking to them, and if you read the legitimacy page you'll understand better why. I don't know the programming behind their site... and I don't want to know the programming behind their site. I wouldn't think that they would falsely inflate results like that... but really who knows. What I do know is that the tests here use a time tested method that's been developed over a long time. And they use no third party apps that may throw off results. I employ very unbiased methods. I tell my users the truth based on the numbers. Many of them still do. It's just that STdN is a more popular site that TMN now. The people 'in the know' still come here In fact, someone I was talking to the other day was telling me how they just got installed with Cox.net and they told me how the tech who installed them went here as soon as the connection was live. They also have ways to run speed tests and QOS tests directly from their meters these days... but that only tells them that the modem into their network is okay. There is a billion other factors that could cause problems that kind of direct test wouldn't see. If anything... this site makes them look good. As long as the user is actually getting what they're paying for. :::.. testmy.net test results ..::: Download Connection is:: 52116 Kbps about 52.12 Mbps (tested with 51200 kB) Download Speed is:: 6362 kB/s or 6.4 MB/s Upload Connection is:: 4915 Kbps about 4.9 Mbps (tested with 5983 kB) Upload Speed is:: 600 kB/s Tested From:: https://testmy.net (Dallas, TX USA) Test Time:: 2010-10-13 11:43:16 GMT D-Validation Link:: https://testmy.net/db/CKSE7MQP9 U-Validation Link:: https://testmy.net/db/ZO1LXRCMU User Agent:: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.10) Gecko/20100914 Firefox/3.6.10 [!]

Update on this

Welcome to the site! I just wanted to let you know that I'm working on getting test servers in the UK as well as other locations. However, I want to finish my site restructure before I get into that. The numbers you get here are still of use to you. My test server is setup in a way that enables it to push data over distance better than most servers. If you configure your end correctly while using testmy.net as your test bench you can see a huge increase in transfer speeds to the US and other locals that are far from you. I've seen people in the past 2X, 3X, 5X their connection speeds over seas by tweaking their settings. Soon I'll be writing a updated tweaking guide that covers many of the new Operating Systems but for now if you'd like to look at a member submitted tweaking solution here are the files as submitted by VanBuren. cablenut408.zip VanBurens_cablenut_settings_all_versions.zip You may also want to do a search here of all forums for "cablenut" or something along those lines. Trust me though... distance doesn't always have to screw up your speeds. Especially with a pipe like yours. Vanburen for instance pulled like 16 Mbps all the way to Sweden, I'm pretty sure that he only had like a 20 Mbps package or something... although we can't really ask him anymore since he hasn't been an active member of the site for a few years now Anyways, good luck... and if you do figure out how to pull more from your connection over distance please post some instructions for other testmy.net users who are in your situation to read. - CA3LE

I thought you were The King of Winnebago, MN?!?!

First off, what is the upload speed that your supposed to have... what package do you pay for. We'll go from there.

very nice