Last night I was at the gym in the showers, washing off all the swimming pool chlorine, when a woman across from me asked, “Was the pool colder than usual today?” Well, being polite led us to having a whole conversation, breaking the very awkward rule of no talking to strangers while naked in the shower.
Turns out she went to UW for undergrad (about my parents’ time). And she went to Cal for grad school. And she’s an econ prof now - at tiny little Bay Area school you’ve never heard of. Now, for those of you who are not profs/didn’t spend a lot of time around them? Fact about profs: at least 90% of them love giving advice, and she was no exception.
Don’t become an academic.
The naked econ professor started off, from what it sounds like, a very strong grad. Interviewed at top schools, took a job at a big name state school in a terribly boring place, where, to quote her, the best restaurant in the town was Denny’s. Since then, she has slowly worked her way down the chain of university quality, through some combination of (a) tenure denial - she didn’t mention this, but it might have had something to do with it, (b) wanting to get the hell out of boring places, (c) more recently, being in a two-body problem.
In the end, she said she’d spent a lot of time unhappy. Living in places where she couldn’t meet people/date and couldn’t even eat out or have a good time. Moving a lot. Working a lot. She says that ultimately, she wished she’d gone in to industry. Now she lives in a good place, but what did she do all the work for, living in crappy places and, for most of it, being all alone?
This is a new perspective for me - in grad school, it’s mostly a win. Sure, you give up money, but I actually enjoy being in grad school (and not because I get out to SF all the time. I don’t). I get to pursue my own interests for a few years, make my own schedule, and I’m not at all lonely - I’m surrounded by tons of like-minded people. Even when people I know have said grad student didn’t go as they’d hoped, and left with a master’s degrees, for the most part they got something out of being in grad school and didn’t actively regret the decision.
The only grown-up academics I’ve met are the ones who’ve won the academic game: the tenured profs at places like UW and Berkeley - they (a) live in places where they have fun, have worked hard but (b) are all happily married, are (c) at strong schools, (d) have students they enjoy working with, and (e) have done work they’re proud of.
But what about the rest of the world? What makes it worth it?
For me, I think reaching my mid 40’s without (b), (d), and (e) would make me look back and regret it too - but (a) contributes to (b) and (c) contributes to (d) and (e), so it’s hard to disentangle it all. Is going on the academic track something you might look back on and regret, or is it like grad school, where when you look back - and even if you leave - you can point to things you learned or did that made it a good experience?
I listened to an interview on creativity on KQED this morning: http://www.kqed.org/a/forum/R201204041000. One interesting tidbit stuck with me - in a study of successful Broadway plays, they looked at the composition of casts and creative teams and how well their plays performed in terms of $$ and ratings:
Turns out (3) worked the best - (1) suffered from “same old same old”, (2) suffered from people who didn’t quite know how to work well together, and (3) was the sweet spot.Filing this away for choosing-your-next-research-group-to-work-with…
- Teams where the whole group had worked together before and knew each other well
- Teams where no one in the group knew each other well at all before starting the project
- Teams where some fo the group had worked together before and knew each other well, but there were a few newcomers as well.
I spent my senior year of college poking at, measuring, and experimenting with the IP Prespecified Timestamp option, a little-known, rarely used optional extension to the IP protocol that requests that routers forwarding a packet insert a timestamp into the packet before forwarding it. Conventional wisdom said that it was too rarely supported - and when supported, too buggy - to be useful, but with Reverse traceroute, our alias resolution paper, and my senior thesis, we found that some fraction of routers support it well enough, and in the wild world of confusing data that is active Internet measurement, this fraction was useful enough to do some very useful things.
It’s super exciting to me that my research baby (perhaps a better analogy is “first crush”) is now off attracting attention on its own in the wide world, first with CAIDA’s Motu aliasing tool and now, with a paper in PAM 2012 which revisits deployed support for IP timestamps. Their paper describes an active measurement study of support for IP timestamps that is in some ways broader and deeper than the data we were able to provide.
In “A Hands-On Look at the IP Prespecified Timestamp Option,” Walter de Donato, Pietro Marchetta, and Antonio Pescape quantify and analyze the support for IP timestamp for 1.7 million IP addresses, and extend their analysis not only to IP timestamp attached to ICMP probes (as we did), but also to UDP, TCP, and SKIP probes as well. Among their findings - about 40% of IP addresses respond to IP timestamp requests embedded in ICMP; other protocols (UDP/TCP/SKIP) provide some increase in the number of addresses you can get to respond to IP timestamp requests; these other protocols have only minimal marginal utility (All ~2% or less) on top of ICMP in the first place. That is, if you really really want to get timestamp replies from as many routers as possible, you should not just use ICMP probes but other protocols too - but you won’t huge gains by doing so. They, like CAIDA, also identify a number of “buggy” implementations I hadn’t seen in the past. When doing things at scale, things break in newer and more interesting ways.
In “Real Science” the gold star for a piece of work is “reproducibility”: did subsequent researchers come back and show data that repeats your findings, or does it refute them? I’m happy that the authors of this paper revisited the applicability of our proposed timestamp techniques, now with their dataset rather than ours. They found that our timestamp alias resolution technique works for about 12.9% of cases - a smaller figure than what we saw, but still very useful number of cases, given our conjecture that the “perfect” alias resolution technique will have to combine techniques to be effective - and that our Reverse traceroute timestamp step should work for 35% of IPs.
The full list of papers from PAM 2012 is here - measurement nerds and I (where I is a subset of measurement nerd, of course) are stoked to get to read through them this week.
A plethora of articles in the press recently have bashed Wikipedia for the organization’s lack of female editorship - a project that is supposedly “open to all” without female membership is clearly doin’ it rong. Por ejemplo:
- Where are the Women in Wikipedia?
- Study: Women and Wikipedia Don’t Mix
- Nine Reasons Women Don’t Edit Wikipedia
Last week, I wrote my first Wikipedia article (check it out!) as part of a class project for Brian Carver’s Cyberlaw course in the Berkeley I-School. After reading all of the above articles, I didn’t feel particularly keen to edit but for class, I did as I was asked. As I clicked “submit” to save my article for the first time, I braced myself for an onslaught of nitpicky editors, legal activists with an axe to grind, or… something.
TLDR? I did indeed interact with a bunch of Wikipedia editors, and they were all supportive and kind. At first blush (or prima facie, as I’m learning to say in my law class), I think they’re doing the right thing - or at least, it’s clear that there are parts of the community who are trying to change things.
- Editor Interaction #1: Nice guy! Welcomed me to Wikipedia, explained to me how Talk pages and Sandboxes worked as I got started writing my article.
- Editor Interaction #2: A form asked me how I felt after posting my first article - I mentioned that I had been nervous that I might break the Wikipedia formatting rules. Someone responded on my talk page, and told me to check out Wikipeda’s “Be Bold” policy.
- Editor Interaction #3: An editor posts to my talk page inviting me to the Wikipedia Teahouse. Editor Interaction #4 came today, similarly inviting me to the Teahouse.
If you click through to the Teahouse, it’s clearly aiming to broaden female participation - just look at the pastel background and references to tea. However, in substance, it does a lot of good not only to make Wikipedia more inviting to women, but to new editors in general. First, it provides profiles of Wikipedia editors, aiming to present well-rounded individuals with diverse backgrounds so that any new editor can feel like he or she “fits in”. Second, it provides a Q&A for new Wikipedia users to ask questions and chat with current editors. I didn’t feel patronized flipping through it, rather, I learned some useful information - and started to feel like I would like some of these other Wikipedia editors.
We’ll see how my development as a Wikipedian goes. I’m already eyeing some of the technical pages I think could use a fixup. But so far, so good, Wikipedia. I laud your efforts to welcome new users in to your community!
Okay, so the other night I was out to dinner with some friends, and one of them stole my phone and jokingly said he was going to Tweet that I was giving up Twitter for Lent. I stole the phone back, and told him that was never happening.
Then I stopped and thought “Why is that ridiculous? Shouldn’t I be able to give up Twitter for just a few weeks?”
To this thought, I say: Challenge Accepted.
I’m deleting the Twitter and FB apps from my phone and blocking both from my computer from Wednesday until Easter.
Some questions I’m curious about during this little self-experiment:
- Will I become less socially aware without technical media to keep me updated? Is it true that participating in social media is a necessary component for maintaining 21st century relationships?
- Will I become less current-events aware, or will I become slower to learn of current events? I keep track of the news via Google Reader and subscriptions to various news sites… and via whatever my friends post to Twitter and Facebook. I’m not sure what the balance of my content sources is.
- How will lack of access to social media impact my day-to-day behavior and social interactions? Will I require more facetime? As a student and a researcher, I spend a lot of time by myself, reading, working, and coding - not exactly the ideal set of activities for an extrovert. I find myself using Facebook/Twitter as a “hit” of social interaction when I find myself getting lonely. I suspect that without this artificial social hit, I’m more likely to hang out by the drinks in the lab, distract my officemates, and drag my housemates to playing games of Dominion. Some of this may be good (my housemates and I don’t hang out nearly enough) and some of it may be bad (distracting officemates may take more time than a quick Twitter checkin).
- Will I make it? I mean seriously… I’m pretty addicted to social networking. I might just cave mid-season.
What’s Lent? Lent is a season in Catholicism and other sects of Christianity leading up to Easter. It starts this Wednesday with Ash Wednesday (if you see people wandering around with black smudges on their foreheads, this is why) where we begin a season of fasting, self-reflection, and sacrifice. Catholics during this season don’t eat meat on Fridays (except fish) and they choose something to give up for the entire season - typically something they consider an indulgence or treat. Most religions have some period of fasting/sacrifice. The past few years I’ve given up meat for the whole season; one year I gave up alcohol… you get the picture.
What about the blog? My tumblr (this thingy) is a repository for most of my academic-networking-work related stuff, and isn’t particularly social, so I’ve decided it doesn’t count. I’ll keep posting here - and allow it to cross-post over to Twitter and Facebook - but I won’t see any responses to the Twitter’d/Facebook’d versions of my posts.
Tomorrow is my last day for online social networking - see you all on Easter!
Are you in high school, junior high, or do you work thereabouts? Does your institution use filtering software such as NetNanny or CyberPatrol? I’m curious what your opinions are on the effectiveness of said technology.
The backstory: in my Cyberlaw class today, we discussed the legal framework for protecting children from online obscenity and indecent content, such as pornography. Back in the day, Congress tried to pass a bunch of laws restricting people from posting such content online and making it accessible to minors; these bills were mostly struck down as First Amendment violations.
A key part of the argument they made was, “Why force the websites to filter themselves when you can have parents and schools install filters that will be just as effective, if not more?”
Here’s the thing: when I was in high school and junior high, the filters were TERRIBLE. They worked very poorly: kids would get blocked from all sorts of things they should have been able to access, and vice versa.
But, that was a ZILLION YEARS AGO. Like, seriously, a decade ago. Because I am old. And technology has progressed since then.
So, my question for you, sisters, sisters’ friends, teenagers, schoolteachers, and librarians, is how well does this stuff work today?
- Would you categorize filtering software as, for the most part “effective” at blocking things the adults want blocked, or “not effective” at blocking the stuff the adults want blocked? (Note that this question is not whether or not it’s fair that the adults block Facebook at school. This question is whether or not you’re usually blocked from things adults don’t want you to see, and whether or not you’re rarely accidentally blocked from things it is okay for you to see.)
- Have you ever been blocked from something you should have been allowed to see, such as information you were looking up for a report? What happened?
- Have you ever accidentally seen something you know should have been blocked? What happened?
- Do you think the filters the adults do want blocked are fair? Have you ever wanted to access something that you weren’t allowed to see at school? What was it, why did the adults want to block it, and why do you think you should be allowed to see it?
Update: Thanks! That was useful. Most people (via eyeballing) seemed to be in the 20ms-40ms range from their nearest EC2 site. We took the instances below down now.
Update update: no really, we took down the instances, it won’t work if you try to ping them now!
Do me a super quick favor? If you’re in the US, open up a console:
- In Mac, click the spotlight in the top right and type “Terminal”
- In Windows, click Start -> Run -> (Type this in the box: cmd)
- In Linux, I assume you know what you’re doing.
You should get a little black box asking you to enter a command. Type this:
Where <address> is one of the following:
You should get a response that looks something like this:
64 bytes from 220.127.116.11: icmp_req=3 ttl=48 time=27.5 ms
Copy/paste that response (or really, just the time value, in this case 27.5ms) and send it to me, and tell me where you ran the measurement from (home/work/school/a cafe in Seattle/San Francisco/Boston) and anything else you know about the network (I get my Internet from Comcast/Time Warner/AT&T/My tethered cell phone, etc).
Do that for each of the five addresses? It’ll be really quick, I promise.
You can quit ping by closing the terminal/console or by typing “Ctrl+C.”
What did I just do?
The list of numbers I gave you are a list of IP addresses: every computer on the Internet has an address called an IP address. These IP addresses belong to computers in Amazon’s EC2 Datacenters. in fact, it’s one IP address from every public “availability zone” in the US (so, one address is in Virginia, another one is in Oregon, etc).
“ping” is a command that sends a small message (a packet) to another computer on the Internet, and the other computer replies immediately (pong!). If you time how long it takes from when you send your “ping” to when you receive the “pong” back, you know roughly how close you are to the other computer. If the “round trip time” is say, 150ms, you’re probably talking to a computer on another continent, but if it’s say, 5ms, it’s someone super close to you. Packets travel at about the speed of light - it’s just electricity!
Why did you need that?
Each of those IP addresses belong to a server running some experiments I’m doing with Colin, Shaddi, Vyas, Arvind and Sylvia. Part of our project involves assuming people will be “close” to Amazon, and I’m curious how that pans out outside of the networks I’ve used in Berkeley (UC campus, my house, Shaddi’s house, and assorted Cafes).
This is a public post that I’m doing just for curiosity - the data is not part of any experiment and will not be part of any publication, besides my sticking it here on this blog!
Dear Media Industry,
Great job on providing reasonably-priced easily-accessible digital content. I’m really digging this whole Netflix/Amazon Prime/Spotify media-service model. I’m a happy paid subscriber to some services, and I even click on ads for others. I feel like we’re finally having a pretty good relationship here. There are still a few kinks to work out but on the whole I think we’re on the right track.
Here’s something you need to work out right now though: Globalization and the Internet. It’s a thing. Your customers are aware of it.
So, Spotify, when you tell me that my Metric album isn’t available for streaming in the US, I feel like you’re taunting me.
Also, Amazon Video? Downton Abbey Season 2? Don’t tell me to “look forward to it” when my UK friends have already seen it.
And, PS - confession time? I’m super tempted to access those things without paying you for them. Because I can get them now rather than later. But I’d rather pay you for them? If you’d let me?
So earlier this week I posted about the Stop Online Piracy Act, on why it won’t work to stop piracy (you can circumvent it in about 5 minutes), and why the technical requirements it’s demanding are bad for the security of the Internet.
A friend of mine has been enjoying the dinner-table-politicking that everyone tells you you’re supposed to avoid that you wind up doing anyway (Happy Chanukah!) and a family member of his made the statement during such an argument that filtering the Internet, in general, is a good idea. In talking about this with him, the wheels started turning in my head about what context this non-technical relative was probably missing before making that statement. I’m not going to argue with the statement that filtering the Internet is or isn’t a good idea, but I want to provide a little bit more context to what that means in a technical sense and provide a reason that this is a hard decision that most non-geeks probably haven’t considered.
So, for this thought experiment, let’s say that the US government passes SOPA2, a bill that allows the government to filter the Internet effectively (which, as I pointed out last week, SOPA doesn’t do). In this imaginary world, the US government has a full on government filtering policy, a la the Great Firewall of China or many of these countries. And, of course, our filter is only used to block things like pirated content or child pornography, but not to block things like search queries for “Tiananmen Square.”
As part of SOPA2, US citizens are banned from trying to circumvent the filtering mechanisms - it only makes sense, right? Not only do we have barbed wire fences to stop people crossing borders illegally, we also tell the citizens that it’s illegal to try to climb them. So the law is a combination of technical mechanism (a fence or filter) + public policy (that you’re not allowed to try to get around it). [*]
Here’s the wrench in that story that you might not have known about, though. For over the last ten years, the United States Navy has invested in anti-Internet filtering technology called Tor. Tor is commonly used by dissidents in some of the world’s most free-speech-oppressive countries to anonymously browse the Internet and circumvent national filters. Tor volunteers give workshops to journalists and political activists heading in to wiretapping, filtering countries (think China, Iran, and some of the US’s other best pals).
And here’s the tough choice I think everyone should think about: Tor (and other technologies) can be used to circumvent filters in other countries, including our imaginary SOPA2 United States which only filters pirated content and child porn, but doesn’t suppress free speech otherwise. The technology is agnostic to what you use it for. It gets around Internet filters, but whether you’re looking for documents on the UN Declaration of Human Rights (banned in your country) or pirated media (banned in your country), the technology can’t tell the difference. It will help you do either. All such technology will fundamentally suffer from this downfall.
So, you have two options on your hands:
- Continue to invest in and allow Americans to build anti-circumvention technology, knowing that to do so puts tools in to the hands of thieves, child pornographers, and the like to commit theft and abuse over the Internet.
- Cease to invest in anti-circumvention technology and ban American’s from contributing to it, knowing that to do so removes a valuable tool from the hands of the oppressed to exercise their free speech, leaving them once again vulnerable to imprisonment, torture, etc for speaking their minds.
I’m not going to argue what you should choose, but know that you are making this choice and that there’s no way around it.