Jan

29

Measurement Lab Launches!!! Sascha's Inbox, Server, Head Explode.

sascha's picture

Well if you've been following tech news today, you probably already know -- after month (years, really) of work, the MeasurementLab.net (M-Lab) initiative has now officially launched. Given how many people have been working on this project, I'm amazed it didn't leak. Having just launched, I've been stunned by the immediate outpouring of interest -- not only did it peg our servers, taking a couple of the research tools momentarily offline, but my inbox has been entirely flooded by correspondences.

The news coverage has been pretty phenomenal:

But it's the stories that didn't make the press that intrigue me the most. So here's three quick bits of info that you won't read that's part of the real history and success of the M-Lab project:

1. Derek Slater -- my colleague and co-conspirator on the M-Lab project. Without his organizing skill, dedication, and all-around fantastic leadership, M-Lab wouldn't have gotten off the ground. Derek's kept a low profile on this project, but deserves to be publicly outed as the Wizard of M-Lab who kept everything on the rails. My hat's off to him.

2. kc claffy runs the Cooperative Association for Internet Data Analysis (CAIDA), who I consulted with for a couple years before moving to Washington, DC. Her freakishly and frenetically brilliant assessments of how much we don't know made me (and many others) a true believer in the need for maximum network research as soon as humanly possible. When things go awry with the Internet, people are going to be saying, "Didn't kc write a paper predicting this problem about ten or twelve years back?"

3. Awesomely cool researchers! M-Lab's afforded me with an opportunity to work with some remarkably cool folks whose life work is to help ensure that you're able to read this praise of their research. But two dozen folks have come forward to provide the brains behind M-Lab and they've been unbelievably helpful -- cramming to get servers set up, troubleshooting when the inevitable ghosts in the machine pop up, and keeping the back end of M-Lab humming along.

More when I get caught up a bit more...

Tags:
»
  1. Brett Glass (not verified) on Thu, 2009-01-29 15:06

    The M-Labs "tools" aren't measurement tools; they are propaganda tools.

    We've run them against 3 Mbps, wide open Internet connections and they've reported a throughput of 39 Kbps. (We've tested the connections with reliable testers to verify that, indeed, we can get 3 Mbps out of them.) In many cases, they've even failed to run at all or become unresponsive in the middle of a test.

    Clearly, the M-Labs programs either are not capable of accurate measurements or are specifically intended to cast aspersions upon ISPs. We suspect that in fact the latter is the case, because it appears that they are being promoted by Google as part of its "network neutrality" war against ISPs. The worse they make ISPs look, the more likely it is that uninformed people will use the biased results as arguments for regulation of the Internet -- which is, of course, what Google wants. It is not surprising that Sascha and the New America Foundation, which is heavily funded by Google, would be promoting such misleading "tests." ISPs might do well to band together and sue M-Labs for mischaracterizing and slandering their services.

  2. sascha on Thu, 2009-01-29 21:25

    Brett -- I know you're convinced of a giant "network neutrality cabal" that's out to get you, but the notion that many of the top Internet researchers are out to "cast aspersions upon ISPs" is high comedy. I suspect one of two things:

    1. The measurement servers have been hit particularly hard (a couple million visitors in the past 24 hours) -- which can lead to degraded results (which is explained more fully in the documentation for the tools).

    and/or

    2. Human error in interpreting the test results. Check to see where the tools are actually reporting the bottleneck -- tools like NDT will tell you exactly where the congestion point is -- they may well be reporting to you that the problem lies outside your network.

    ***

    Either way, it's a cheap shot to be calling into question the independence of the people who have been working on the Measurement Lab project. These researchers are committed to independent and objective scientific inquiry -- they're certainly not biased against ISPs nor is there any sort of "war against ISPs." Finally, New America Foundation is not "heavily funded by Google" as you state -- as far as I know, Google's contributions to New America run in the tens of thousands of dollars range (out of a yearly budget of $12-15 million).


  3. Brett Glass (not verified) on Fri, 2009-01-30 12:34

    In that case, Sascha, let's see New America's 2008 Form 990, complete with the required list of funding sources.

    Oh, and who would the Chairman of your Board of Directors happen to be? No connection to Google at all, eh?

  4. sascha on Fri, 2009-01-30 13:41

    Brett,

    All of New America Foundation's 990s are freely available online. NAF's Employee Identification Number is 52-2096845 and you can use the 990 finder of your choice to take a look at our paperwork (I've used the Foundation Center's 990-finder a bunch and it's pretty good: http://foundationcenter.org/findfunders/990finder). The 2008 990 will be us there as well, just as soon as it's completed.

    As for Eric Schmidt, he joined New America's Board back in 2000 when he was CEO of Novell -- well before he was ever CEO at Google (see: http://www.markle.org/news/press_releases/2000/press_release_07102000.php). I know you really want to find a great big conspiracy; but the fact is, Eric Schmidt was involved with New America Foundation before he ever joined Google.


  5. Brett Glass (not verified) on Fri, 2009-01-30 14:38

    Sascha:

    New America's Form 990 is not present on that site. I've managed to find one on Guidestar.org, but -- guess what? -- the section which lists contributors and the amount they contributed has been redacted. I guess that New America doesn't want the public to know how much it's getting from Google, so that it can claim not to be a shill for it (even though it obviously is).

  6. sascha on Fri, 2009-01-30 17:09

    Let me ask you, Brett -- if I can find out how much Google has given to New America Foundation over the years, and if it's an insubstantial amount (say less than 20% of its yearly budget), would that change your mind? How about if it was less than 10% of NAF's yearly budget? 5%? How about 1%? I'm just curious at what point you would be willing to admit that your assessment was incorrect? And since percentage of yearly budget is a fairly objective metric, I figure it would be a good way to get an accounting for how you're making your determination.

    To be perfectly honest, I actually don't know what the percentage is -- it's never come up before. So there's always the possibility that I'll be eating crow on this. But I'm wondering what threshold you set in determining that New America Foundation is a shill for Google. For your part, let me know, and for my part, I'll see what information I can get from our finance department.

    Lets lay our cards on the table and see where we stand.


  7. Brett Glass (not verified) on Sat, 2009-01-31 18:37

    Yes, Sascha, let's lay our cards on the table. How much did Google contribute -- both in money and in kind -- to the New America Foundation in 2008? And were there indirect contributions? (Very often, corporations will "launder" some or all of their contributions to a group through a foundation.)

    Myself, I already have no cards to hide. I operate a small ISP which Google's war on ISPs threatens to extinguish. The New America Foundation has been lobbying for policies which would regulate the Internet and create a duopoly by driving small, rural, and competitive providers out of business -- and they are JUST EXACTLY the policies that Google advocates as part of its corporate agenda. Coincidence? I think not.

  8. sascha on Mon, 2009-02-02 16:39

    Google contributions to the New America Foundation last year were $9,000. The books aren't yet closed for 2008, but that's current as of this afternoon. Doing the math, that works out to 0.06% (six hundredths of one percent) of the $14,100,000 in revenues the New America Foundation received for 2008. There were no indirect contributions through intermediaries. In kind donations were equally negligible.

    Hopefully this puts to rest the concerns you've raised.


  9. Brett Glass (not verified) on Tue, 2009-02-03 11:49

    Actually, Sascha, it destroys your credibility. It's absurd to claim that Google, whose CEO is chairing your group (and whose water you're carrying in DC with expensive events), would give you only $9,000. This shows that not only is your group willing to lie about the facts and the technology -- it's also willing to lie about its support.

  10. sascha on Wed, 2009-02-04 14:57

    Whoa, you asked about Google contributions and I had our HR folks find out Google's contributions and made them publicly available as plain as day. Now, having found no smoking gun (because there isn't one) and unwilling to admit you were entirely wrong in your assessment, you're on to Eric Schmidt.

    I'm not certain how much Eric's donated personally -- since, as I've mentioned, he's been a part of New America for 8+ years now (from before his time at Google and since the founding of the New America Foundation). As the Chronicle of Philanthropy documents (see: http://philanthropy.com/news/government/index.php?id=3928), Eric made a $1 million dollar unrestricted donation when he stepped up as Board chair. There's no smoke and mirrors, Brett, this was publicly announced quite widely. As should be obvious, since the funding was unrestricted, there are not any strings attached. There was no quid pro quo or conspiracy -- just a guy who has believed in the mission of New America Foundation since its founding, since before he ever worked for Google.


  11. Brett Glass (not verified) on Mon, 2009-02-09 10:14

    Sascha, as you know (or should know), a contribution from Eric counts, legally, as a contribution from Google. And I find it hard to believe that there isn't other support, as well, that you're not talking about. The New America Foundation might well just rename itself the "Google Lobbying Foundation." Because that's what you are. You're a sellout. And you're harming America. It's sad. What's even sadder is that even if your astroturf group fails to do the harm that it intends to do on behalf of Google, you'll probably still have a job. Not so for those of us who have actual skin in the game and are doing USEFUL work that HELPS people. You ought to be thoroughly ashamed. How do you sleep at night, knowing that you're just a corporate tool?

  12. sascha on Tue, 2009-02-10 06:42

    Brett, you're hopeless. In the face of overwhelming evidence to the contrary, rather than shift your position to meet the facts, you stick to a position that is untenable given the facts.


  13. Dana Spiegel (not verified) on Thu, 2009-01-29 16:34

    Sascha,

    Congrats on the project announcement! I remember the conference that you organized with kc at UCSD, where everyone talked about how more data could be gotten about normal network usage. Its great to see that some of the ideas from that meeting are actually being implemented.

  14. sascha on Thu, 2009-01-29 21:42

    shhhhh -- that was in December 2006. Remember, the official story is that Google started this with their meeting in 2008. ;)

    Of course, my own efforts were only one of a zillion different attempts to get something started. In the end, I like to think of Vint as a catalyst -- things definitely did gel after the Google meeting. Now we just need to ensure that we take this momentum and build on it over the next several months. So we're actually just getting things started. But thanks for the kind words and thanks for being involved from before the (official) beginning.


    15. Post new comment

    The content of this field is kept private and will not be shown publicly.
    • Lines and paragraphs break automatically.

    More information about formatting options