nanobot

Hah, forgot to post this here. Yeah, they asked me about it on Twitter a week or so ago.
 
It's an interesting topic, we're discussing things that should never have to be worried about, no one should fear that their ISP can censor what they can see based on the ISP's own political motives. (Hell, just look at the Comcast v. Netflix issue from a couple years ago: Netflix traffic was throttled to unbelievable slow speeds as a result of Comcast wanting to extort them for money.)
 
Personally, I am anti-government regulation, but in this case it's a necessity.
 
One of the arguments I have recently heard in favor of revocation of the net–neutrality law was:
 
 
This is a false premise, if the ISP's customers want traffic from that source, the ISP should not be throttling it in either direction. If your customers want to watch Netflix all day, you don't get to make additional profit off of that. The customer already pays for a broadband connection, it is the ISP's job to deliver that broadband service in an unbiased manner.
 
Thanks,
EBrown

Well with a switch between them, it got up to about 1.1Gbps, which is statistically irrelevant. Without the switch (crossover) it got up to 1.5Gbps.
I think the bottleneck was HDD transfer speed.

And I'm not at all surprised by my results - Full-Duplex should always be capable of this.

Thanks,
EBrown

This site has no such app, you can use the browser on any smart-mobile for it though. Chrome, Safari, Firefox, Opera. All of them work.
 
Thanks,
EBrown

I foresaw this problem with a project for work, as I was using Base64 encoding on them.

So, I wrote my own Base32 implementation pattern:


Public Shared Function ToBase32String(ByVal Bytes As Byte()) As String Dim Arr As String = "0123456789bcdfhjkmnpqrtvxyDFGHJLMNPQRTVXY" Dim Result As String = "" Dim BT As Byte() = New Byte(Bytes.Length + (5 - (Bytes.Length Mod 5) - 1)) {} For I As Integer = 0 To Bytes.Length - 1 BT(I) = Bytes(I) Next For Group As Integer = 0 To BT.Length / 5 - 1 Result += Arr((BT((Group * 5)) And &HF4) >> 3) Result += Arr(((BT((Group * 5)) And &H3) << 2) Or ((BT((Group * 5) + 1) And &HC0) >> 6)) Result += Arr(((BT((Group * 5) + 1) And &H3E) >> 1)) Result += Arr(((BT((Group * 5) + 1) And &H1) << 4) Or ((BT((Group * 5) + 2) And &HF0) >> 4)) Result += Arr(((BT((Group * 5) + 2) And &HF) << 1) Or ((BT((Group * 5) + 3) And &H80) >> 7)) Result += Arr(((BT((Group * 5) + 3) And &H7C) >> 2)) Result += Arr(((BT((Group * 5) + 3) And &H3) << 3) Or ((BT((Group * 5) + 4) And &HE0) >> 5)) Result += Arr((BT((Group * 5) + 4) And &H1F) >> 0) Next Return Result.Substring(0, (Bytes.Length * 8 / 5)) End Function(Note: the only usable portion of the Arr is "0123456789bcdfhjkmnpqrtvxyDFGHJL", the rest is there for completeness.)
I designed this pattern to avoid the chance for certain, offensive or obscene terms, and avoid characters that are visually similar, as I was generating Base64 from a direct Byte-Array. Now I generate it as a custom Base32 set so that I don't have those issues.

Thanks,
EBrown

Oh, also, a nice little chart to explain: http://www.ntia.doc.gov/files/ntia/publications/spectrum_wall_chart_aug2011.pdf

And the FCC one: https://transition.fcc.gov/oet/spectrum/table/fcctable.pdf

http://gizmodo.com/5629814/giz-explains-why-everything-wireless-is-24ghz

Thanks,
EBrown

Well with a switch between them, it got up to about 1.1Gbps, which is statistically irrelevant. Without the switch (crossover) it got up to 1.5Gbps.
I think the bottleneck was HDD transfer speed.

And I'm not at all surprised by my results - Full-Duplex should always be capable of this.

Thanks,
EBrown

You guys will love this:

I was farting around with my network just now, and I was testing my Gigabit throughput between my desktop and server, and I have some pretty good evidence of having achieved MORE than 1Gbps on the link. (Not in one direction - mind you.)

The actual throughput (send + receive) was stable between 550Mbps (each) to 750Mbps (each). Essentially, I achieved up to 1.5Gbps total throughput.

As is evident in the images, the speeds exceeded 1Gbps total. (And not just by a tiny bit - either.)

Thanks,
EBrown

http://gizmodo.com/5629814/giz-explains-why-everything-wireless-is-24ghz

Thanks,
EBrown

You guys will love this:

I was farting around with my network just now, and I was testing my Gigabit throughput between my desktop and server, and I have some pretty good evidence of having achieved MORE than 1Gbps on the link. (Not in one direction - mind you.)

The actual throughput (send + receive) was stable between 550Mbps (each) to 750Mbps (each). Essentially, I achieved up to 1.5Gbps total throughput.

As is evident in the images, the speeds exceeded 1Gbps total. (And not just by a tiny bit - either.)

Thanks,
EBrown

You guys will love this:

I was farting around with my network just now, and I was testing my Gigabit throughput between my desktop and server, and I have some pretty good evidence of having achieved MORE than 1Gbps on the link. (Not in one direction - mind you.)

The actual throughput (send + receive) was stable between 550Mbps (each) to 750Mbps (each). Essentially, I achieved up to 1.5Gbps total throughput.

As is evident in the images, the speeds exceeded 1Gbps total. (And not just by a tiny bit - either.)

Thanks,
EBrown

Unfortunately PowerShell does not display transfer speed, which is why you use WGET on a *Nix platform to determine it.

Thanks,
EBrown

Unfortunately, Geo-Location by IP address is not an exact science - or a science at all, really. 
(Click the spoiler for the boring comparison part)




An IP Address is exactly like a Phone Number, given my IP 10.0.0.5, my physical neighbour is NOT guaranteed to have IP Address 10.0.0.4, nor 10.0.0.6. So how do we find out where you are located by it? We attempt to determine where the home-network for that IP claims it is "physically located." We use WHOIS API's (ARIN, RIPE NCC, AFRINIC, APNIC, LACNIC) to determine what the "location" of the IP address is.

To give you an example, my IP Address is in the following block: 68.37.128.0 - 68.37.255.255. Testmy.net thinks my IP Address is from Jersey City, NJ. I will tell you for a fact, I do not live in Jersey. I live in the Detroit Metro Area, Michigan. Unfortunately, ARIN does not realize that COMCAST gave me an IP Address that is from New Jersey, so Testmy.net thinks I am currently in New Jersey.

This issue is by design, the IPv4/IPv6 protocols were never designed to allow geo-locating a user to any degree of certainty. They were always designed to allow anonymity.

That being said, there are ways to more accurately pinpoint a users location. If I do a tracert to Testmy.net, I find that my data goes from 10.0.0.1 (my gateway) to *.hsd1.mi.comcast.net, allowing me to surmise that I am SOMEWHERE in Michigan.

As the tracert continues, my data goes through several more nodes, to the node just before TMN: *.dal05.networklayer.com, which gives me the impression that Testmy.net is located in or near Dallas, TX. (It is in fact in Dallas, we should all know this for fact.)

So, as for why you are 2000% above your city average today, but maybe not tomorrow: when you reset your modem, if your ISP gives you a new IP address, it may show you as being in another city. Regardless if you are or not. I implore you to click "my city average" and compare what city it lists to what city you actually live in. If you look at this list (https://testmy.net/city/jersey_city_nj -> Member Rank) you will see myself (nanobot) at the top of the list for Jersey City, NJ. Unfortunately, I got my gas bill this week: right-proper addressed to E. Brown, New Baltimore, MI.

And believe when I say, this is not a TMN issue, this is intentional. If you use any proxy servers you can even more-so assuredly bet that your city will be even more wrong.

Thanks,
EBrown

Have you run TCP Optimizer? I had the same issue when I switched network providers and my IE speeds were horrible but my Chrome speeds were fairly good.

Thanks,
EBrown

Where are you at? We used to have Hughes Net at my dad's place but we ditched it because another ISP had recently shown up. (We didn't get out of the contract, but we did get service with the new ISP.)

I'd check that there aren't any small, local providers first. If there are, look into getting one of them and then start with the BBB about Hughes.

And check if you get 3G/4G in your area. I think Tommie Gorman uses EVDO for his internet, and he seems satisfied with it.

Thanks,
EBrown

The difference of the two you say?

Speedtest.net uses a pure flash-based approach, and drops the top and bottom 5% of your tested speed. Why? Because the first instant of testing is typically slower due to the requirement to establish a connection. Then there's often a sudden burst which is typically irrelevant to your speed. Speedtest.net is more useful for a total-throughput test. It's often handy for gamers to test what the typical speed of their connection is. (After all, they are using UDP mostly which is connectionless, so there is no establishment of a connection.)

Testmy.net uses a pure HTTP based approach, and does not remove any data. Why? Because almost all of the typical users network traffic is over HTTP, which requires the establishment of a connection. (Downloads, webpages, style-sheets, images, etc.) Testmy.net is much more relevant to testing your total connection throughput. Basically giving you a true establishment of how websites should download.

Thanks,
EBrown

It's listed by username, not display name. So whatever you use to login is what would be there.

Thanks,
EBrown

Well it really gets you when you visit pages that have 2/3 of the browser view covered in ads.

I have never once used advertising on any of my sites. (And with them being so small, it wouldn't benefit me anyway.)

I don't ever browse to unknown or untrusted sites on servers and such that I run. I think the only site outside my own that I view from my servers is TMN, oddly enough.

Thanks,
EBrown

Well it really gets you when you visit pages that have 2/3 of the browser view covered in ads.

I have never once used advertising on any of my sites. (And with them being so small, it wouldn't benefit me anyway.)

I don't ever browse to unknown or untrusted sites on servers and such that I run. I think the only site outside my own that I view from my servers is TMN, oddly enough.

Thanks,
EBrown

I've always used the Dallas server. (And I don't use multi-threaded either. ) Yeah, the D.C. server is probably closer to me, but I prefer having the large route. Gives me a better overall figure that way.

Thanks,
EBrown

I've always used the Dallas server. (And I don't use multi-threaded either. ) Yeah, the D.C. server is probably closer to me, but I prefer having the large route. Gives me a better overall figure that way.

Thanks,
EBrown

The difference of the two you say?

Speedtest.net uses a pure flash-based approach, and drops the top and bottom 5% of your tested speed. Why? Because the first instant of testing is typically slower due to the requirement to establish a connection. Then there's often a sudden burst which is typically irrelevant to your speed. Speedtest.net is more useful for a total-throughput test. It's often handy for gamers to test what the typical speed of their connection is. (After all, they are using UDP mostly which is connectionless, so there is no establishment of a connection.)

Testmy.net uses a pure HTTP based approach, and does not remove any data. Why? Because almost all of the typical users network traffic is over HTTP, which requires the establishment of a connection. (Downloads, webpages, style-sheets, images, etc.) Testmy.net is much more relevant to testing your total connection throughput. Basically giving you a true establishment of how websites should download.

Thanks,
EBrown

Considering you are located in the UK, I would recommend you use the London server. (Mostly because the large majority of .uk sites are located within the UK, so it represents the typical speed from you to them.)

Other than that, the choice really doesn't matter except to change what routing paths are used and such. If you manually decide to select the San Jose server, you will likely choose a significantly different route than if you were to select the Dallas server, or the Washington D.C. server.

Thanks,
EBrown

The difference of the two you say?

Speedtest.net uses a pure flash-based approach, and drops the top and bottom 5% of your tested speed. Why? Because the first instant of testing is typically slower due to the requirement to establish a connection. Then there's often a sudden burst which is typically irrelevant to your speed. Speedtest.net is more useful for a total-throughput test. It's often handy for gamers to test what the typical speed of their connection is. (After all, they are using UDP mostly which is connectionless, so there is no establishment of a connection.)

Testmy.net uses a pure HTTP based approach, and does not remove any data. Why? Because almost all of the typical users network traffic is over HTTP, which requires the establishment of a connection. (Downloads, webpages, style-sheets, images, etc.) Testmy.net is much more relevant to testing your total connection throughput. Basically giving you a true establishment of how websites should download.

Thanks,
EBrown

Indeed, happy birthday man.

Thanks,
EBrown