Slashdot Mirror

← Back to Stories (view on slashdot.org)

FCC App Lets Android Users Measure Mobile Broadband Speed

Posted by Soulskill on from the fast-enough-except-when-it's-not dept.

itwbennett writes "The FCC's new Android app will allow users to measure the speed of their mobile broadband connection, while providing aggregate data to the agency for measuring nationwide mobile broadband network performance. Released as open-source software on Thursday, the free FCC Speed Test App will test network performance for parameters such as upload and download speed, latency and packet loss. An iPhone version of the app is in the works."

72 of 93 comments (clear)

  1. Government efficiencies by Anonymous Coward · · Score: 1

    Can't they just ask the NSA?

    1. Re: Government efficiencies by iamhassi · · Score: 1

      Hasn't the speed test app been doing this for many years already?

      --
      my karma will be here long after I'm gone
  2. Tempting... but no thanks. by scottbomb · · Score: 1

    If every 3-letter agency in the federal government wasn't so busy spying on us, I might get it. But nowadays, I don't trust anyone from Washington.

    1. Re:Tempting... but no thanks. by aitikin · · Score: 3, Insightful

      Being that it's open sourced, I'll wait 2 months and someone will audit the code, then I'll consider it. (yes, I don't know enought o do that myself).

      --
      "Don't meddle in the affairs of a patent dragon, for thou art tasty and good with ketchup." ~ohcrapitssteve
    2. Re:Tempting... but no thanks. by Anonymous Coward · · Score: 1

      I'm much more concerned about the telecos identifying this software and gaming the results.

    3. Re:Tempting... but no thanks. by simonbp · · Score: 2

      What precisely are you afraid of? And I'm being serious, what could the app access that you find worrying? It does grab your location, but that is trivial for any law enforcement agency these days.

      If this app helps the FCC ensure that wireless companies are honest, I'm all for it.

    4. Re:Tempting... but no thanks. by Anonymous Coward · · Score: 2, Funny

      I'm much more concerned about the telecos identifying this software and gaming the results.

      Good! Then I'll write my Apps to mimic the requests sent out by the FCC App.
      Once the teleco switches my connection into "high speed for great rankings!" mode, my App will switch into "all your bandwidth are belong to us" mode!

    5. Re:Tempting... but no thanks. by icebike · · Score: 3, Insightful

      Won't be useful, unless you want to do nothing but download the same chunk of data over and over again.

      We can only hope that the authors were wise enough to request randomly named data blocks from randomly selected data servers, because otherwise we will be measuring the effectiveness of the carriers cache at the nearest tower base.

      --
      Sig Battery depleted. Reverting to safe mode.
    6. Re:Tempting... but no thanks. by Charliemopps · · Score: 4, Funny

      What precisely are you afraid of?

      I'm sorry, but that's none of your business.

    7. Re:Tempting... but no thanks. by artor3 · · Score: 2

      Why are you dodging the question? The GP wasn't making one of those fallacious "nothing to hide" arguments. He was correctly pointing out that this test doesn't provide the government with any useful spying information.

      It collects your location at the time you run the test (you can uninstall it afterwards), your phone model & carrier info (which they already have easy access to), and... that's it.

      So again, what exactly are you afraid of?

    8. Re:Tempting... but no thanks. by Charliemopps · · Score: 2, Interesting

      So again, what exactly are you afraid of?

      And again, it's none of your fucking business. I can chose not to use an application written by an entity that have absolutely no trust in. What I have to hide is everything. They have no right to know anything about me that they do not explicitly need to know. By asking me "What are you afraid of?" you're indicating that I couldn't possibly be afraid of this innocuous app and you're asking me to defend my abstinence with an explanation of how the app could be used against me or to incriminate me. The point of fact is, it's none of your business why I chose not to use the app. I do not have to explain myself. My choices and my reasons for them are my own and not subject to peer review. This fact is what needs to change about modern society. You have no right to know why I chose not to trust my government.

    9. Re:Tempting... but no thanks. by icebike · · Score: 1

      what could the app access that you find worrying? It does grab your location, but that is trivial for any law enforcement agency these days.

      And app running in your phone could grab anything it wants from the phone, and send it anywhere it wants.
      Its called malware, and no phone is totally immune from it, not Android, not IOS. There are ways.

      This is why it matters, ESPECIALLY when the application is provided by the government.

      In case you've been on vacation on Mars for the last several months, you missed the big
      foo-fa-rah where we all discovered that the government was in fact reading all of our email
      and monitoring all of our phone calls, even though we thought we had passed laws to forbid that.

      Once bitten twice shy.

      --
      Sig Battery depleted. Reverting to safe mode.
    10. Re:Tempting... but no thanks. by DigitAl56K · · Score: 1

      As long as the versions on the Play store are identical to the audited version and you don't update until the updates are audited either..

    11. Re:Tempting... but no thanks. by icebike · · Score: 1

      He was correctly pointing out that this test doesn't provide the government with any useful spying information.

      From your mouth to god's ears.

      But If there was nothing to be afraid of, do you think the FCC would have bothered to released the source code?

      THEY understand that there is a lot to be afraid of, and in an attempt to build confidence and distance themselves
      from other branches of the government, they release the entire code-base AHEAD of time. You have to give them
      credit for that, and credit for understanding the level of distrust that exists between the government and the citizens.

      --
      Sig Battery depleted. Reverting to safe mode.
    12. Re:Tempting... but no thanks. by aitikin · · Score: 1

      I'd be surprised if someone didn't fork it. Not like it'd be difficult to do a fork, audit, and update as needed.

      --
      "Don't meddle in the affairs of a patent dragon, for thou art tasty and good with ketchup." ~ohcrapitssteve
    13. Re:Tempting... but no thanks. by Anonymous Coward · · Score: 1

      no, cockhead, he's asking if there's something real that he needs to be concerned about. He's not saying you're a cunt (I am); he's trying to get some useful information from the (misguided) assumption that you are aware of a real reason to be concerned.

    14. Re:Tempting... but no thanks. by Gravis+Zero · · Score: 1

      What precisely are you afraid of?

      I'm sorry, but that's none of your business.

      that's fine. we already know.

      --
      Anons need not reply. Questions end with a question mark.
    15. Re:Tempting... but no thanks. by Lodlaiden · · Score: 1

      Why is it when people have something worth saying, they do it as anonymous coward?

      --
      Suborbital [spaceflight] is the special olympics of spaceflight. - Rei
    16. Re:Tempting... but no thanks. by FlyHelicopters · · Score: 2

      I can chose not to use an application written by an entity that have absolutely no trust in. What I have to hide is everything.

      That sounds reasonable, except that you're using a phone in the US that is, by design, not secure to those three letter agencies.

      So... they don't need you to run this app to get access to your phone, it is hard-wired into the chips (installing a clean version of Android doesn't help with that).

      You can't turn off the location, you can't turn off the camera and sound, and you can't disable it, unless you remove the battery. (turning off the camera in options disables it for apps that you install, not to the chips themselves. You'd have to rip the GPS chip out, for example, to actually disable location monitoring.)

      So, if you really feel that way, why would you have a phone that they have access to?

      Note: I'm not knocking your position or saying that you're wrong, maybe you have a reason to care, maybe you simply choose to because you believe in the 4th amendment and your privacy. More power to you. But you give that up by using a phone that is already open to them.

    17. Re:Tempting... but no thanks. by FlyHelicopters · · Score: 2
      Yep, releasing the code says: "We don't care about you or your contacts, your location, or anything else, we just want to see how fast the cell network is running and gather data on that."

      Frankly, it is part of the FCC's job, this is something they *should* be doing, so more power to them.

      The FCC is charged with regulating the airwaves of the United States, so this is part of their publicly stated function.

    18. Re:Tempting... but no thanks. by FlyHelicopters · · Score: 1
      Yes, all true...

      However...

      You do know they can access your phone already, this isn't even a secret. The GPS chip is always tracking your location, it has to provide the location for E911 services regardless of the "setting" in Android. The feature is hardwired into the chips, turning it on and off in software doesn't actually do anything other than block your location from installed apps.

      The camera and mic can be turned on remotely, and even the phone can be turned on remotely, if there is power from the battery, so it is never "really" off.

      If you are paranoid about the three letter agencies, a cell phone with your name on it is the last thing you should have.

      Just saying...

    19. Re:Tempting... but no thanks. by icebike · · Score: 1

      Mostly a collection of myths and half truths.

      All allegations of turned off smartphones being remotely turned on have been traced to phones to which the FBI had access, and upon which they inserted software to make it LOOK LIKE it was off, when it in fact was on. The same for turning on your mic and camera. Every instance traced to phones handled by police or FBI, and some malware infected phones.

      There is no other cases of this in the wild. Lots of tinfoil hat types posting the same nonsense, but when you track it down it all goes to exactly ONE court transcript concerning a phone that came into FBI hands for a while, and they didn't fool anyone because they drained the battery so fast the owners knew something was wrong.

      Further, if you call 911, and IF your 911 call center is so equipped, they can send a signal to turn on your GPS chipset (on some models of phones).
      Other phone models turn on GPS the instant you dial 911.

      Not all 911 call centers have the technology to even read your phone's location. Big cities do, but it still hasn't rolled out to West Podunk yet.

      One thing that is true, is dialing 911 on a GMS phone with a dead or missing SIM card WILL often get you connected to 911. Even when your carrier doesn't love you any more, your State and County government may.

      --
      Sig Battery depleted. Reverting to safe mode.
    20. Re:Tempting... but no thanks. by artor3 · · Score: 1

      Yeesh, calm down there little guy. Obviously, you can choose not to use an application for any reason. You can choose not to use the app because squirrels ate the moon, if that happens to be what makes sense in your own mind. But if you're going to post about your decision in public, people might like to know your reasoning. And if your reasoning doesn't hold up to scrutiny, people will say so.

      And, by the way, your reasoning doesn't hold up to scrutiny.

    21. Re:Tempting... but no thanks. by binaryhermit · · Score: 1

      You'd have to rip the GPS chip out, for example, to actually disable location monitoring.

      (Tin foil hat on) And even then, they can still triangulate your location from the cell phone towers. Also, it's possible in many cases to get your location from comparing the WiFi access points that your phone sees to a database compiled by wardriving or from users using something along the lines of http://opensignal.com/. (Tin foil hat off)

    22. Re:Tempting... but no thanks. by binaryhermit · · Score: 1

      My understanding is that it's federal law that all cell phone networks must route 911 calls from compatible cell phones regardless of service status.

    23. Re:Tempting... but no thanks. by ooshna · · Score: 1

      People that have something worth saying usually end up gagged, blindfolded, thrown in the back of a van, and end up in some secret prison.

    24. Re:Tempting... but no thanks. by HiThere · · Score: 1

      Actually, it *IS* providing them with useful information, your location. The only thing it, they could already get that if you're carrying a phone. All recent models have GPS tracking, and even my antique can be located via cell tower triangulation.

      P.S.: The cell towers NEED to know roughly where the phone is if it is to be able to be used. If you don't want that, just put it in some sort of Faraday cage. (Metalized plastic envelopes should work, but I don't remember whether you need to ground them.)

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    25. Re:Tempting... but no thanks. by icebike · · Score: 2

      P.S.: The cell towers NEED to know roughly where the phone is if it is to be able to be used.

      Don't anthropomorphize cell towers, they hate that.

      They don't NEED to know anything. Your phone is programmed to find all the cell towers it can "hear" to and connect to the one with the strongest signal from its preferred roaming list. It is strictly controlled by the phone, not the tower. Once your phone connects, the tower updates a database so that calls can be routed to it when your number is called.

      If you are moving, your phone will periodically find stronger towers, and register to those towers, while keeping the channel on the old tower. As soon as the tower you are leaving reaches a low enough signal, the call will jump seamlessly to the new tower.

      Towers themselves are really rather dumb about which phones are in their area, its up to the phone to connect to the tower, not the other way around.

      --
      Sig Battery depleted. Reverting to safe mode.
    26. Re:Tempting... but no thanks. by HiThere · · Score: 1

      You have demonstrated deeper mechanical insight to the process than I carry in ready memory. I anthromorphize (to a limited extent) to have a simpler model to use. For the purpose of "can you be identified by cell tower triangulation" it works sufficiently. Your answer was, however, obviously more technically correct...but it doesn't change the assertion.

      OTOH, whether they DO locate you by your phone is another matter. I suspect that (for most people) they don't, to avoid drowning in information. OTOH, I never would have guessed that they were tracking everyone's email, so I don't have a great deal of certainty about that.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
  3. ISPs: new TODO list item by Anonymous Coward · · Score: 1

    Memo to ISP network engineers: identify servers used by FCC and add QoS rules to prioritize the relevant traffic.

    1. Re:ISPs: new TODO list item by icebike · · Score: 1

      You mean cache everything they fetch on every tower?

      --
      Sig Battery depleted. Reverting to safe mode.
    2. Re:ISPs: new TODO list item by SeaFox · · Score: 1

      They don't need to.

      ISP's have "official" speed testing sites generally, either ones they themselves operate, or ones they bless as being the official test site. You can call them up and say "I'm only getting blahblah megabits per second when I test my speed at whatever.net" and they wont care unless you're getting similar results at the official test.

  4. By mobile broadband they mean.... by icebike · · Score: 2

    I have this installed, and it keeps separate track of BOTH wifi measurements and cellular network measurements.
    But it measures both, and allows you to swipe left and right to see each measurement it took.

    What I've learned: My carrier is pretty pathetic.

    Note
    Being Open source, you can see exactly what is being reported, but I predict that won't stop the tinfoil hat crowd from claiming the binary does not reflect the source code.

    --
    Sig Battery depleted. Reverting to safe mode.
    1. Re:By mobile broadband they mean.... by girlintraining · · Score: 2

      What I've learned: My carrier is pretty pathetic.

      They're all very pathetic. They're oversubscribed by many thousands to one; Your shittiest cable provider doesn't hold a candle to how pathetically oversubscribed the average mobile provider is. These towers typically only have a T1 backhaul... it only takes a couple of phones to saturate those links. You will never, ever, get the full-rated OTA speed. Anywhere.

      And they employ super-massive buffers; They're the reason buffer-bloat has become a problem. Latencies far above what even 90s-era modems provide -- 500, 800ms easy. Bandwidth is irregular and employs highly manipulated QoS to allow access to a select few websites at full speed, while taking the piss out of the rest of them -- there's a reason Facebook loads quick, while a site like, say, Slashdot, takes 30 seconds or more.

      The FCC needs to not just run bandwidth tests, but suss out their QoS; People need to show that anything but the top 50 websites give absolutely terrible performance. You can get your google results in seconds, but it might take several minutes to load up the homepage of the restaurant you were searching for.

      --
      #fuckbeta #iamslashdot #dicemustdie
    2. Re:By mobile broadband they mean.... by binaryhermit · · Score: 1
      No offense, but I have to call BS on this:

      These towers typically only have a T1 backhaul.

      Since T1s are a little bit over 1.5 mbps symmetrical, (1.54 mbps IIRC) this result: https://pbs.twimg.com/media/BZJiaGpCcAAonC9.png:large seems to disagree with your statement. If I understand Verizon's network setup correctly, I'd guess that they're using at least something like a OC-3c.
      Now, if you'll excuse me for a second, a bit of a related rant.
      The only problems with my experience with mobile internet are that:
      a) Verizon's LTE network is nowhere near as fast as that result in some places where I regularly use my phone.
      and
      b) By my calculations. I could blow through my 2 GB data allowance in under 36 minutes just by maxing out my down speed.

    3. Re:By mobile broadband they mean.... by Shakrai · · Score: 1

      These towers typically only have a T1 backhaul

      That's a pretty impressive T1, since my LTE speed record with Verizon Wireless is 49mbit/s, with typical day-to-day speeds of 8mbit/s to 12mbit/s. Have you actually talked to a network engineer for any modern cellular provider or are you just making assumptions? The few times I've worked directly with the wireless carriers (on behalf of two clients with buildings the carriers wished to install base stations in) they've pulled fiber in to feed their equipment. T1s might still be used as a last resort for 2G/3G-only rural base stations, but they are not being utilized to feed modern base stations. Verizon's 2x10mhz LTE deployment provides a theoretical maximum of 75mbit/s downstream and ~18mbit/s upstream. They aren't feeding that with T1s....

      Latencies far above what even 90s-era modems provide -- 500, 800ms easy.

      My latency (tested via pings to 8.8.8.8) on Verizon's LTE network is around 80ms, on the EVDO network it's about 200ms. The only way I can spike a wireless ping test to the numbers you're talking about is to saturate the link beforehand with a huge download.

      --
      I want peace on earth and goodwill toward man.
      We are the United States Government! We don't do that sort of thing.
    4. Re:By mobile broadband they mean.... by girlintraining · · Score: 1

      No offense, but I have to call BS on this:

      These towers typically only have a T1 backhaul.

      Well, It's true. They're starting to integrate into other assets, as Time Warner points out: Many cell phone providers are hooking cable modems up to their towers to boost speeds. Some towers, where regulations permit, and where sufficiently high enough to avoid a safety hazard, also use microwave links to nearby central offices. But the majority of towers being deployed only have a T1 or equivalent for the backhaul.

      If I understand Verizon's network setup correctly, I'd guess that they're using at least something like a OC-3c.

      Except you aren't. Parts of their network do, sure, but a lot of towers don't. And you don't seem to understand how these cells mesh together. Your cell phone can, in a typical urban environment, probably talk to over a dozen towers. But it doesn't. It usually talks to the nearest one; To keep transmitter power low and keep wireless "slots" free in adjacent cells. But there may only be 1 or 2 4G towers, but maybe 5 3G, or phone-only towers.. or whatever. My point is that it's a mixed environment. They can even have you talking on one tower while making a data connection on another, and all of this is being handed off all the time when you're mobile. Sometimes a tower oversaturates and hands traffic off to another one, or forces your phone to downgrade; It'll say 4G but it's only talking on 3G, for example.

      b) By my calculations. I could blow through my 2 GB data allowance in under 36 minutes just by maxing out my down speed.

      Yeah. Why do you think the data allowances are so low, while believing the network capacity to be so great? It strikes me as a big flaw in your line of reasoning.

      --
      #fuckbeta #iamslashdot #dicemustdie
    5. Re:By mobile broadband they mean.... by Shakrai · · Score: 3, Informative

      Verizon pulls fiber to their base stations wherever possible.

      T1s would be a nightmare. The current 2x10mhz LTE network tops out at about 75mbit/s down and 18mbit/s up. Multiply that by three (most base stations have three sectors), then add more bandwidth to account for each of the 3G Ev-DO channels (3mbit/s down and about 1.8mbit/s up) provided, multiply that number by three, then add a non zero amount for voice service (9.6kbit/s per call with current CDMA codecs, 13kbit/s for older codecs), SS7 signaling and other overhead.....

      T1s barely scaled to meet demand for the Ev-DO network, where you had to contend with a demand for at least 9mbit/s (3mbit/s times three sectors) of data, in reality more than that since they typically allocate at least two channels for Ev-DO service, and you still need to have bandwidth for voice and signaling service. To meet that sort of demand you're talking about twelve or more bonded T-1s, and at that point you may as well just use a microwave repeater to reach those rural base stations where it's cost prohibitive to pull fiber.

      --
      I want peace on earth and goodwill toward man.
      We are the United States Government! We don't do that sort of thing.
    6. Re:By mobile broadband they mean.... by Shakrai · · Score: 1

      I won't even bother replying to the rest.

      In other words you got called out on your BS and refuse to back it up. Gotcha. I think you're the kind of person who knows just enough to be dangerous. Case in point:

      Some towers, where regulations permit, and where sufficiently high enough to avoid a safety hazard, also use microwave links to nearby central offices.

      "Safety hazard"?!? How much power do you think microwave links need? Hint: With good antennas and a clear LOS you can go miles with a power output measured in milliwatts. TV stations broadcast with power levels measured in kilowatts, and people live right next door to the transmitters while suffering no health effects whatsoever. You might want to familiarize yourself with the inverse square law and the basic fundamentals of RF engineering....

      BTW, that link you gave the other poster doesn't even mention the word "T1" in any context, never mind validating your absurd claim that they're feeding 4G base stations with T1s.

      --
      I want peace on earth and goodwill toward man.
      We are the United States Government! We don't do that sort of thing.
    7. Re:By mobile broadband they mean.... by thejynxed · · Score: 1

      I can confirm the T1s where for I live. While somewhat rural, even the nearest 100K+ population city doesn't have (and probably won't have) anything 4G/LTE in the foreseeable future. Maybe by 2024. Maybe.

      Let me tell you, they roll those things out in very select and specific areas to make it appear they have great coverage with this, when in fact they do not, and aren't even close to covering the numbers they are claiming on those maps.

      Put it this way - if it isn't going to be a population center of at least 500k or more, it won't happen anytime soon, and even then it will be as cheaply done as possible to save on the fiber rollouts and lease fees.

      --
      @Mindless Drivel: 100% of Twitter posts ever Tweeted.
    8. Re:By mobile broadband they mean.... by icebike · · Score: 1

      He's not using any "personal assessment". Your claim is simply ludicrous. Oh and you want a link? Try this one where Verizon talks about using fiber optic solutions from FiberTower to build out their LTE network. Verizon could not run an LTE network off a T1s to their towers. You're simply an idiot.

      Exactly, anyone thinking that an tower is still served by only T1 is hopelessly behind the times.

      FiberTower isn't a Verizon entity, it services the entire industry.

      FiberTower Corporation is a provider of facilities-based backhaul services to wireless carriers. As of December 31, 2010, the Company provided services to 6,400 billing customer locations at 3,276 billing sites in 13 markets throughout the United States; had master service agreements with nine wireless carriers in the United States; had relationships with fiber service providers giving the Company access to over 1,000 mobile switching centers (MSCs) and 125,000 fiber-based aggregation points, and owned a national spectrum portfolio of 24 gigahertz (GHz) and 39 gigahertz wide-area spectrum licenses, including over 740 megahertz in the United States metropolitan areas and, in the aggregate, approximately 1.55 billion channel pops calculated as the number of channels in a given area multiplied by the population, as measured in the census, covered by these channels.

      Crown Castle. As of December 31, 2012, it owned, leased, or managed approximately 31,500 towers, and they just bought 4 billion worth of towers (9,700) from AT&T. The US has 4 or 5 if these giant tower companies, and all of them are served by much larger than T1.

      --
      Sig Battery depleted. Reverting to safe mode.
    9. Re:By mobile broadband they mean.... by binaryhermit · · Score: 1

      I did the math, and in an area where Verizon is only deploying on the 22 MHz they have in the C block, it comes to 152.1 T1s per cell site, not counting voice , SMS, and overhead. They probably could cut back on the data backhaul and just deliver slow data speeds, but I can't imagine over a hundred T1s being more feasible than some sort of fiber-based solution.

    10. Re:By mobile broadband they mean.... by binaryhermit · · Score: 1

      Why do you think the data allowances are so low, while believing the network capacity to be so great?

      All the analysis I've seen says that it's nothing but a cash grab. No more, no less.

    11. Re:By mobile broadband they mean.... by pspahn · · Score: 1

      You have responded with ad hominid attacks and strawmans.

      Ad hominid isn't a real thing. Your entire argument is therefore flawed.

      --
      Someone flopped a steamer in the gene pool.
    12. Re:By mobile broadband they mean.... by faedle · · Score: 1

      I can tell you for a fact that at least one of the "big four" is buying way more than a single T1 for their towers, at least where I live. They buy a fairly large amount of dark fiber from the company I work for to connect a number of towers on the outskirts of town.

      Granted, they could be running a single T1 over every fiber. But I somehow doubt that.

    13. Re:By mobile broadband they mean.... by Shakrai · · Score: 1

      Meanwhile, the rest of us have figured out that if sticking your head in a home microwave is dangerous

      Home microwave: 0.8 to 1.5 kilowatts (59dBm to 62dBm)
      Microwave repeater: No more than a few watts, typically less. (30dBm for 1 watt, 37dBm for 5 watts)

      If you can't see the difference between 1.5 kilowatts and something typically measured in milliwatts then I don't know how to help you. Besides, I never advocated climbing the tower and sitting in front of the transmitter. I was responding to your absurd (and unsourced) claim that they can only use microwave repeaters "Where the tower is tall enough" because of "safety concerns". Your microwave leaks more RF than you'd be exposed to living under a transmission tower hosting repeaters.

      Are you also afraid of wi-fi networks? Same frequency range as your microwave, with power outputs that aren't much lower than the microwave repeaters you're trying to claim raise safety issues.

      --
      I want peace on earth and goodwill toward man.
      We are the United States Government! We don't do that sort of thing.
  5. Speedtest by x0ra · · Score: 1

    It would seem that Speedtest already has app for most major mobile players. What is so what is different about that one, beside the fact that all the result is centralized by the FCC (with the associated risks) ?

    1. Re:Speedtest by icebike · · Score: 1

      It would seem that Speedtest already has app for most major mobile players. What is so what is different about that one, beside the fact that all the result is centralized by the FCC (with the associated risks) ?

      Speeddtest.net uses the data for their own purposes. Using them enriches them.

      Supposedly this is strictly for the FCC, by the FCC.

      --
      Sig Battery depleted. Reverting to safe mode.
    2. Re:Speedtest by binaryhermit · · Score: 1

      I've seen some pretty suspicious (as in higher than the speed of the connection that you're paying for) numbers come out of the speedtest app.

    3. Re:Speedtest by leuk_he · · Score: 1

      According to the speedtest app, our wifi network @home is the limitiation, and on android the phone is limited more.
      -cabled pc show full network speed.
      -modern laptop show almost full speed of wifi network
      -android shows less accurate full speed of wifinetwork with higer ping.

      Cheapest internet i can get here is the fiber :)

  6. We already know by gelfling · · Score: 1

    It's shit megabytes. This is America where we're 15 years behind everyone else and it costs more.

    1. Re:We already know by tepples · · Score: 1

      But the real reason I posted was - Android app out before iPhone app? What is the world coming to?

      For one thing, the app is distributed under a free software license, and Android by and large is friendlier to free software than iOS. For another, Android users outnumber iPhone users.

  7. Re:Who wrote the App by icebike · · Score: 1

    There appears to be no magic new technology here.

    It measures upload and download speed and packet loss and latency.
    The target server from which upload and download are measured is not clear.
    Carriers prefer you only measure against a server on their own network, and will blame the upstream for any measurements to other locations.

    --
    Sig Battery depleted. Reverting to safe mode.
  8. Easily gamed by gweilo8888 · · Score: 1

    All of these speed tests are ludicrously easily-gamed, and are thus of next to no value in the real world. They don't tell you what speed you're getting on real-world websites, they tell you what speed you could theoretically get when your internet provider lifts caps on bandwidth, prioritizes your traffic over those of other users on the same cell tower / network for the duration of the test, etc.

    And you're naive if you think some or all of the above doesn't already happen.

    1. Re:Easily gamed by Em+Adespoton · · Score: 1

      All of these speed tests are ludicrously easily-gamed, and are thus of next to no value in the real world. They don't tell you what speed you're getting on real-world websites, they tell you what speed you could theoretically get when your internet provider lifts caps on bandwidth, prioritizes your traffic over those of other users on the same cell tower / network for the duration of the test, etc.

      And you're naive if you think some or all of the above doesn't already happen.

      The crazy thing is, even with all of the above being true, the speed tests still suck. That means that while GAMING the tests, they still provide sub-par performance. Think about that for a moment.

    2. Re:Easily gamed by icebike · · Score: 1

      All of these speed tests are ludicrously easily-gamed, and are thus of next to no value in the real world.

      Looking at the code, and a log of the application I can see that it first looks up the closest server from some list, mine ended up hitting samknows1.sjo1.level3.net, port 8080, this appears to be location dependent, then it proceeds to down load (HTTPGET)a file called 100MB.bin for the download test.

      It then turns around and uploads that same file (HTTPPOST) to the same address. It appears to ignore the speed of the first 2621440 bytes.

      Then it starts to measure latency and packet loss against port 6000 of the same server.

      So If I wanted to game that, as a carrier, the best I could do is make sure anything named 100MB.bin got a lot of priority.
      Since the 100MB.bin download doesn't change between consecutive fetches (tested with a browser) its quite possible you could cache both ends of this.
      This of course would require the carriers to rush into production some level of deep packet inspection to determine exactly what is being fetched. And if that inspection doesn't happen in the tower it would be too late to adjust the allotted bandwidth. I seriously don't think the carriers are competent enough to rush such a system into production. (Besides, they rent tower space and transmitters anyway these days, which means they don't necessarily have end to end control).

      (On the other hand, I can't be sure that each requesting IP will get the exact same copy of 100MB.bin, as I could only capture if from my computer's connection.)

      So I can't dismiss the results out of hand, because they do seem to measure my purchased bandwidth when on WIFI, and also my meager bandwidth when on cellular. At this time, I doubt they are being gamed, and probably do measure real world performance.

      --
      Sig Battery depleted. Reverting to safe mode.
    3. Re:Easily gamed by gweilo8888 · · Score: 1

      You're thinking to hard. All you need to do is build a list of all the servers used for the speed test, and then prioritize traffic to those servers. No packet inspection required. No filenames required. Just simplicity.

      Do it for five or six of the most popular speed tests, and you're golden.

  9. Data plan? by BigT · · Score: 1

    How much of my data plan is running this speed test for the FCC going to eat up?

    --
    Is it weird in here, or is it just me?
    1. Re:Data plan? by binaryhermit · · Score: 1

      You can set that in the app, the default setting is no more than 100 MB/month.

    2. Re:Data plan? by Impy+the+Impiuos+Imp · · Score: 1

      That's 5% of my awesome Verizon monthly data cap!

      I mean **only** a **mere** 5%.

      --
      (-1: Post disagrees with my already-settled worldview) is not a valid mod option.
    3. Re:Data plan? by artor3 · · Score: 1

      If you download it, run it once, and delete it: a little under 30 MB.

    4. Re:Data plan? by faedle · · Score: 1

      This is sort of my pet peeve with this whole thing.

      It's not the SPEEDS that suck, dear FCC, it's the stingy caps. I'd be happy with 1-2Mb down if I wasn't hard capped at 1.5GB per month.

    5. Re:Data plan? by coolsnowmen · · Score: 1

      You aren't hard capped; they'll gladly charge you a fee for each MB over :-)

  10. Re: deja vu... by Nkwe · · Score: 1

    didn't the fcc do something very similar with home broadband routers?

    They gave out routers to selected volunteers around the country. The routers measured and sent data automatically. May be if they were to give out cheap Android phones, this would probably be more successful

    I have one of these and the program is still running. Details of the program can be found here. As stated above, they gave out routers that perform intensive testing on your broadband connection over time. Looking at the website they are announcing a new mobile application, so I assume that it is an extension of the same overall program.

  11. Open Source on iPhone will be fun ... by Alain+Williams · · Score: 1

    Given Apple's stance on Open Source - will they refuse permission to have it in their catalogue ? Then what happens if the FCC insists that they list it ? Once it is there someone will ask for the source of the Apple libraries that it is linked to ....

    Time to get some popcorn and watch the fun!

    1. Re:Open Source on iPhone will be fun ... by apenzott · · Score: 1

      The fun begins when the iDevice app is rejected, and Apple has to explain itself to the FCC their decision.

      Remember that the FCC came down on AT&T, Apple, and Verizon on the absence of tethering apps and tethering functionality that was carrier crippled and ORDERED them to explain themselves.

      --
      The Roman Rule: The one who says it cannot be done shall not interrupt the one who is doing it.
  12. Re:Left out in the cold, again by ArbitraryName · · Score: 1

    Anyone know where to download this directly as an APK file?

    From the github link provided in the summary.

  13. Public key by tepples · · Score: 1

    We can only hope that the authors were wise enough to request randomly named data blocks

    Or better yet, use some sort of public-key crypto so that nothing can imitate the FCC servers. In any case, it's supposed to be free software; you can verify that the source does what it says it does.

  14. Re:Who wrote the App by tepples · · Score: 1

    The upstream is exactly what tests like this are supposed to test: connectivity to the Internet itself.

  15. And, of course, because it's the FCC... by EmagGeek · · Score: 1

    ... the app will contact a single FCC-hosted server sitting on a 10mbit Ethernet connection in some office somewhere on 12th st.

  16. Cue the paranoia by nurb432 · · Score: 1

    Ooo its the feds. cant install this.. etc.

    Really, its the FCC, they are the least of our concerns. the other agencies already have your content, so why bother worrying about these guys.

    --
    ---- Booth was a patriot ----
  17. Accuracy not to be trusted. by Revek · · Score: 1

    I can't get any information on the endpoint or their testing methodologies. It shows a 2% packet loss on every test but i've been running winmtr continuously to a ip in the dallas area and have received no packet loss and my tests on speakeasy test to dallas are at the qos i'm paying for. All in all I say this is just another half assed government money pit.