Slashdot Mirror


Firefox Extension Makes Social-Network ID Spoofing Trivial

Orome1 writes "A simple-to-use Firefox plugin presented yesterday at Toorcon in San Diego has hit the security world with the realization that squabbles about Facebook's changing privacy settings and various privacy breaches simply miss the point. 'When it comes to user privacy, SSL is the elephant in the room,' said Eric Butler, the developer of the extension in question, dubbed Firesheep. By installing and running it, anyone can 'sniff out' the unencrypted HTTP sessions currently allowing users on that network segment to access social networks, online services and other website requiring a login, and simply hijack them and impersonate the user."

185 comments

  1. Illegal? by Anonymous Coward · · Score: 5, Informative

    I don't dispute author's work or goals (I've been using SSH tunneling on public WiFi for years to prevent just this) but he should have mentioned that clicking on information you gathered (and logging in as another user without their concent) is very likely against federal laws in US (and likely most other locations). Just gathering this information can likely be argued to be illegal as well (wiretapping?)

    So be careful where you click..

    1. Re:Illegal? by Anonymous Coward · · Score: 1, Informative
    2. Re:Illegal? by Romberg · · Score: 3, Funny

      yeah, like I'm gonna click on your link.....

  2. First haxx! by Anonymous Coward · · Score: 4, Funny

    Ha ha, anon is pwned :D

    1. Re:First haxx! by Anonymous Coward · · Score: 5, Funny

      WTF !, this guy is logged in as me !

    2. Re:First haxx! by Anonymous Coward · · Score: 1, Funny

      Dude, seriously, he probably isn't even using the plugin... Your password is one of the worst I've seen. Heck, even I cracked it (as you can see from this post)

    3. Re:First haxx! by Anonymous Coward · · Score: 2, Funny

      Remind me to change the combination to my luggage.

    4. Re:First haxx! by Anonymous Coward · · Score: 0

      Too late, already stole it! If you want your luggage back, post here.

    5. Re:First haxx! by Anonymous Coward · · Score: 0

      DISREGARD THAT

      I SUCK COCKS

    6. Re:First haxx! by Anonymous Coward · · Score: 0

      careful who you lend ur ipad to bro lol

    7. Re:First haxx! by iieugenio · · Score: 1

      GTFO samefag

  3. A better explaination by buchner.johannes · · Score: 5, Informative

    here: http://codebutler.com/firesheep

    They apparently call it "sidejacking", i.e. sniffing other users cookies from a wifi, and using it. Not new, but made userfriendly.

    --
    NB: The message above might reflect my opinion right now, but not necessarily tomorrow or next year.
    1. Re:A better explaination by thomst · · Score: 3, Informative

      here: http://codebutler.com/firesheep.

      Steve Manuel of TechCrunch claims that the Force-TLS 2.0 Firefox extension can defeat Firesheep. (You have to configure it manually for each site you want to protect, though, so it's somewhat of a PITA.)

      Another option is the HTTPS Everywhere Firefox extension from EFF and the Tor Project. Although HTTPS Everywhere has a predefined ruleset that includes some of the most popular Web sites, you'll still have to write your own ruleset for any site not on their default list.

      --
      Check out my novel.
    2. Re:A better explaination by Anonymous Coward · · Score: 0

      What keeps this from being deployed beyond social networking sites? Say, using someone ebay's account they haven't logged out of (not that logging out half the time actually deletes the cookie), and jacking in a bunch of bids on stupid or embarrassing items or just to financially mess or up screw with the feedback rating of a user?

      I've always wondered this. Apparently, others have too, because it looks like a lot of stuff aside from social networking isn't encrypted that relies on cookies for the more sensitive or monetary stuff.

      This is why I've always been suspicious of ebay. That, and one time I put in a high bid early on instead of sniping (I sniped because I could never figure out if this couldn't be done, then it appears to have been done) because I had an appointment, and there was 1 bid, $.41 under mine, that didn't get rebid (at that point, they could see my bid amount fully, and $1 more could have had the item). Then I read a couple years later on /. how ebay never protected people from server side includes, and sellers were embedding code in their auction descriptions to report back bid amounts, and ebay allegedly KNEW about the practice and didn't care (higher bids, after all, means higher profit, so no reason for them to protect the buyer if this wasn't a regular practice). Sellers would simply use another account, place a bid just under, and profit big time. (And no, I'm not pissed at the price I paid so much as I'm pissed at the supposed described and fair system being gamed.)

    3. Re:A better explaination by AK+Marc · · Score: 1

      I'll call it "freejacking" in honor of the best movie Emilio Estevez or Mick Jagger were ever in. Which was, coincidentally, the worst movie Anthony Hopkins was even in.

  4. and this is news ? by Torvac · · Score: 3, Insightful

    someone in the same network sniffing your unencrypted traffic is facebooks fault ? or the fact that someone made a UI to do it for dummies ?

    1. Re:and this is news ? by Anonymous Coward · · Score: 5, Insightful

      the fact that it's unencrypted is facebooks fault, it's not hard to push everything through HTTPS, there's no excuse these days

    2. Re:and this is news ? by Anonymous Coward · · Score: 0

      I couldn't get it to work, I suck at the internet :(

    3. Re:and this is news ? by arndawg · · Score: 1

      "not hard". Well maybe not for your blog with 2 users per week. But for facebooks loadsize it's not a matter of signing up with digicert and enabling SSL. But yeah. They probably should prioritize that instead of some fancy new web 2.0 feature.

    4. Re:and this is news ? by Anonymous Coward · · Score: 0

      Why isn't face book requiring SSL? That's one of the thing that seems so simple to me. If there is a user authenticated session, require SSL. IF the session is just people putting items in a basket, then I can see not requiring it yet (at that point all you risk is exposing their shopping cart; possibly undesireable but possibly acceptable as people can look in my shopping basket IRL).

    5. Re:and this is news ? by Ephemeriis · · Score: 5, Insightful

      someone in the same network sniffing your unencrypted traffic is facebooks fault ?
      or the fact that someone made a UI to do it for dummies ?

      The fact that it is unencrypted is, yes.

      --
      "Work is the curse of the drinking classes." -Oscar Wilde
    6. Re:and this is news ? by vlm · · Score: 1

      But for facebooks loadsize it's not a matter of signing up with digicert and enabling SSL.

      Problems can be solved with money. Their only income stream is selling private information. Therefore:

      Scenario one, your privacy is lost because they sell it to someone with money to pay for the dedicated SSL hardware cluster.

      Scenario two, your privacy is lost because semi-smart people skimmed it away.

      Since the end result is about the same, I'd rather reward the smart people than the greedy/rich people.

      --
      "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    7. Re:and this is news ? by biryokumaru · · Score: 1

      Why is it anyone's "fault?" Who cares? It's Facebook for science's sake! It's all just pictures of people's kids and crap, it doesn't matter at all if someone logs on as me and posts nonsense!

      [/perspective]

      --
      When you're afraid to download music illegally in your own home, then the terrorists have won!
    8. Re:and this is news ? by savvysteve · · Score: 1

      The fact that Facebook is getting all of this negative press is not surprising. It has been known for a long time that Facebook is the root of a lot of problems relating to fake antiviruses and spyware etc... If Facebook is pulling down billions a year then investing in a couple of million dollars in equipment or infrastructure to handle the load is minimal. At some point it will simply self destruct and users will begin a mass exodus once more and more articles about the dangers of Facebook are written and IT Professionals and techies begin informing everyone that using Facebook is dangerous especially on a Winblows PC.

    9. Re:and this is news ? by Anonymous Coward · · Score: 1, Informative

      "not hard". Well maybe not for your blog with 2 users per week. But for facebooks loadsize it's not a matter of signing up with digicert and enabling SSL.

      Facebook's issue isn't buying & installing a certificate, it's that they have so much web traffic that the CPU load of encrypting all that traffic (or buying dedicated encryption acceleration hardware) is significant.

    10. Re:and this is news ? by Anonymous Coward · · Score: 1, Interesting

      Post your user/pass if it doesn't matter. Put your action where your mouth/fingers is/are.

    11. Re:and this is news ? by The+Mighty+Buzzard · · Score: 2, Informative

      While I'm inclined to agree that any remotely commercial website should offer and default to encrypted transfers, it also serves you right if you use a service that doesn't encrypt everything. Using a service that doesn't at least offer you the option of encryption is akin to driving a car that you know has defective brakes (ha, car analogy!). If shit goes badly and you knew better, you've no one to blame but yourself. If you didn't know better, it's your own fault for not educating yourself about such basic things and I shall mock you.

      Unless you're a cookie baking grandmother willing to bribe me with baked goods. Principles be damned when there are fresh, warm cookies involved.

      --
      Violence is like duct tape. If it doesn't solve the problem, you didn't use enough.
    12. Re:and this is news ? by Anrego · · Score: 3, Insightful

      users will begin a mass exodus once more and more articles about the dangers of Facebook are written and IT Professionals and techies begin informing everyone that using Facebook is dangerous especially on a Winblows PC.

      Oh you can't seriously believe that!

      People have been screaming at the top of their lungs about how insecure facebook is and what they do with your information for years. Your average user just doesn't care as long as they can keep playing farmville!

    13. Re:and this is news ? by PopeRatzo · · Score: 4, Insightful

      Their only income stream is selling private information.

      Good point.

      I'm surprised so many people are upset about people stealing their private information, but have no problem with someone buying and selling their private information.

      --
      You are welcome on my lawn.
    14. Re:and this is news ? by Anonymous Coward · · Score: 0

      What I don't understand is why people are talking like Facebook isn't offered over HTTPS. There's no excuse not to be using https-everywhere. If you are using that plug-in, you're automatically using HTTPS on Facebook by default. I guess you could argue it's Facebook's fault for not forcing its users to use HTTPS, but anyone with a hint of a clue is already using it anyway.

    15. Re:and this is news ? by Afty0r · · Score: 1

      The fact that it is unencrypted is, yes.

      Wait, it's Facebooks' fault that you chose to browse their site unencrypted?

      You have the choice - if you visit https://facebook.com/ it will let you run your entire session on the site in https. They obviously support SSL for those who want it... I fail to see how it's their fault?

    16. Re:and this is news ? by Aqualung812 · · Score: 4, Informative

      You have the choice - if you visit https://facebook.com/ it will let you run your entire session on the site in https. They obviously support SSL for those who want it... I fail to see how it's their fault?

      Follow the link you attached. Log into Facebook. Click the Facebook icon on that page to return to your home page, or click on a link to a fan page you have, or click on a link to a friend's page. You just went from SSL to HTTP. They make it hard to STAY on SSL, even if you go through the work of going there manually.

      --
      Grammer Nazis - I mod you "troll" unless you actually add something on-topic. Yes, I know I have mispellings in my sig.
    17. Re:and this is news ? by Ephemeriis · · Score: 1

      Should be encrypted by default. Should not be an unencrypted option.

      --
      "Work is the curse of the drinking classes." -Oscar Wilde
    18. Re:and this is news ? by Culture20 · · Score: 1

      Why is it anyone's "fault?" Who cares? It's Facebook for science's sake! It's all just pictures of people's kids and crap*, it doesn't matter at all if someone logs on as me and posts nonsense!

      [/perspective]

      * that people have a general sense of being true. Great mischief can be done with data gathered, or accounts used/people impersonated.

    19. Re:and this is news ? by Stewie241 · · Score: 1

      "not hard". Well maybe not for your blog with 2 users per week. But for facebooks loadsize it's not a matter of signing up with digicert and enabling SSL.

      Facebook's issue isn't buying & installing a certificate, it's that they have so much web traffic that the CPU load of encrypting all that traffic (or buying dedicated encryption acceleration hardware) is significant.

      He needs to do more than get a certificate and add it to his server. He also has to buy more hardware to deal with the extra load that this will add.

    20. Re:and this is news ? by KBJorgensen · · Score: 2, Informative

      The Chrome extension KB SSL Enforcer automatically redirects you to SSL every time you visit Facebook (and other sites) and changes all links to point to SSL. Although I do agree that they should just use SSL by default on a site with so much personal info. Disclaimer: I made this extension.

    21. Re:and this is news ? by Anonymous Coward · · Score: 0

      Actually there is an excuse.

      SSL adds a nontrivial amount of calculations that needs to be done on your server on every request and with a few millions of users and an interactive service that does lots of small ajaxy stuff it quickly adds 6-digit figures to your infrastructure budget.

      It also increases latency noticeably from all the extra RTTs needed for the handshake.

      Another problem that once your page is served over HTTPS there is no trivial and secure way to do AJAX requests to HTTP resources due to SOP, even if those requests contain no private information.

    22. Re:and this is news ? by miserere+nobis · · Score: 1

      Okay, cool, just let me know how it goes the next time a prospective employer looks you up on the interwebs and finds all the Ku Klux Klan rally pictures I just posted to your facebook page.

    23. Re:and this is news ? by Anonymous Coward · · Score: 0

      That's because most people don't actually think of that stuff as private information, probably because it isn't. And because they realize that free web sites have to make money to exist, and they're paying for the services they use by providing their information to interested parties.

      But stealing is login information? That's different. It falls outside the social contract.

    24. Re:and this is news ? by Anonymous Coward · · Score: 0

      It should be noted that even encrypting everything is not enough. You are still vulnerable unless the relevant cookies are marked as only being allowed to be sent over encrypted connections.

    25. Re:and this is news ? by Wayne247 · · Score: 1

      I can see two semi-valid to valid excuses:
      1. The additional server load, and cost associated (electricity, maintenance, less clients per 'unit' served and so on)
      2. Impossible sniffing & tracing to debug live systems.

      Sniffing the data in-house troubleshoot is a great tool. Without it, you have to rely to server-side logging to capture your issues. If these are occasional, hard to reproduce, unclear problems, then you're in for some serious log digging, not to mention considerable work to request from all your services to write a lot of logging details. And then rotate those logs and so on.

      When all the data is unencrypted end-to-end, all you have to do is put a few network capture and analysis tools where you think the problem is, write some possible triggers for them to launch capture and synchronize together, and then voilà, you've got yourself "the bug" in the net, and now you can begin fixing.

    26. Re:and this is news ? by PopeRatzo · · Score: 1

      And because they realize that free web sites have to make money to exist

      We're not talking about sustenance web publishing, we're talking about fucking Facebook for god's sake. They aren't just "making money to exist".

      --
      You are welcome on my lawn.
    27. Re:and this is news ? by Anonymous Coward · · Score: 0

      If facebook is so unsecure, then why isn't no one poisoning their precious big juicy database already, eventually making it (and facebook as a company) lose some of its worth?

  5. Why no encryption? by AHuxley · · Score: 3, Interesting

    What is the cpu use and heat of the user base requesting and using ssl vs this bad news?
    "Double-click on someone, and you're instantly logged in as them."
    Whats the the extra use 15-20%? vs unencrypted HTTP.
    Would ssl been left off allow creative law enforcement uses?

    --
    Domestic spying is now "Benign Information Gathering"
    1. Re:Why no encryption? by betterunixthanunix · · Score: 4, Funny

      Facebook's servers are too busy violating your privacy to handle the extra load of encryption ;)

      --
      Palm trees and 8
    2. Re:Why no encryption? by sakdoctor · · Score: 1

      SSL can be delegated to a PCI-e crypto accelerator board.
      Perhap the same would work for privacy violation?

    3. Re:Why no encryption? by maxume · · Score: 5, Informative

      When Google switched Gmail over to HTTPS all the time for everything, they found it accounted for 1% of CPU load:

      http://unblog.pidster.com/imperialviolet-overclocking-ssl?c=1

      So Facebook probably wouldn't need to do much more than get their software set right.

      --
      Nerd rage is the funniest rage.
    4. Re:Why no encryption? by cerberusss · · Score: 4, Funny

      Facebook's servers are too busy violating your privacy to handle the extra load of encryption ;)

      Facebooks servers were hanging around in a dark alley one faithful night. My privacy just happened to think that particular night, let's take the shorter route home. It's as if Facebooks servers sniffed she was coming, despite her high privacy settings. They libpcaptured her, then stripped all of her headers and checksums, right to her to the bare profile while taunting her loudly. Some traffic just passed by without doing anything. My privacy was violated again, and again and Facebooks servers just kept going and going. Then they left my privacy "face"-down in a shallow ditch, some shreds of unique ROWIDs covering her bloodsoaked profile.

      --
      8 of 13 people found this answer helpful. Did you?
    5. Re:Why no encryption? by Anonymous Coward · · Score: 0

      one faithful night

      Chapter one: The night was faithful. Wait, no: the night... the night was fateful. The night was fateful! No wait, faithful, faithful. The night was faithful! The night was faithful and loyal, loyal and faithful. The night was loyal and faithful, faithful and loyal, loyal and faithful; that's faithful. The night was faithful.

  6. https everywhere by NuShrike · · Score: 1

    Plugin-rebuttal.

    1. Re:https everywhere by Anonymous Coward · · Score: 0

      Re: Buttal plug in.

    2. Re:https everywhere by skywatcher2501 · · Score: 3, Informative
    3. Re:https everywhere by anti-pop-frustration · · Score: 4, Interesting

      https everywhere is indeed a great extension, and everybody should be using it.

      But some of the services that Firesheep target don't offer an https option *at all*. This is no rebuttal, it only proves Firesheep developer's point : these services have an unappropriate level of security.

      The worst offender is probably Yahoo! Mail. They don't even offer https to their paying customers! For one of the leading webmail service this is utterly unacceptable. https for login is a fig leaf, the only thing this does is give users a false sense of security.

  7. What permissions do you need ? by Haedrian · · Score: 1

    What permissions do you need for this? Do you have to be the owner of the network in order to sniff things out in this manner? Or is it possible for me to steal accounts off a public network?

    If its the former, then there's nothing too special - sniffers can do that already.

    If its the latter, then its time to put on the tinfoil hats.

    1. Re:What permissions do you need ? by pinkeen · · Score: 2, Informative

      It is wifi sniffing. The data is in the air. All you need is to be in the range of client's radio transmissions. If the network is encrypted then you need WEP/WPA(2) key.

    2. Re:What permissions do you need ? by mbone · · Score: 3, Informative

      What permissions do you need for this? Do you have to be the owner of the network in order to sniff things out in this manner? Or is it possible for me to steal accounts off a public network?

      None, no, and most emphatically yes.

    3. Re:What permissions do you need ? by Stray7Xi · · Score: 3, Informative

      What permissions do you need for this? Do you have to be the owner of the network in order to sniff things out in this manner? Or is it possible for me to steal accounts off a public network?

      You need to be administrator to place your network card into promiscious mode or rfmon for wireless.

      So in a public wifi network you're screwed. In a public ethernet network it depends if it's a switched or hubbed network. But even in a switched network you could be vulnerable to this via ARP poisoning.

      The takeaway is what we've known for decades, if you want private communications use encryption.

    4. Re:What permissions do you need ? by rHBa · · Score: 1

      I always thought you had to put your wireless NIC in monitor mode to do this, a la airodump/wireshark etc...

      If the addon can do this for any wireless enabled computer capable of running firefox then I'm impressed (especially on Windows), it usually requires specific wireless cards and aftermarket drivers for it to work.

      If that isn't the case how does this plugin capture data not aimed at your local IP address?

  8. Another point is not "missing the point" by Chriscypher · · Score: 5, Insightful

    squabbles about Facebook's changing privacy settings and various privacy breaches simply miss the point.

    Another point does not "miss the point".

    Transport security != corporate marketing of private data

    --
    "You have liberated me from thought."
    1. Re:Another point is not "missing the point" by pinkeen · · Score: 1

      Aircrack sells anything? I thought it was a toolset...

    2. Re:Another point is not "missing the point" by Anonymous Coward · · Score: 1, Informative

      It also misses the point that Facebook is about *SHARING* data. The idea is you are sharing things with people. If you want to keep things private ... Facebook is not the place to do it.

      If I do not want people to know something. It is simple I do not put it on the internet.

      Facebook has created this idea of privacy itself that it will never match. They did it accidentally by putting a login on the front page in the first place. People have an expectation of privacy when you have to log in. Facebook was just using it so you can post as yourself. Not as a privacy feature.

      Also people have a gut reaction of 'lets just ssl everything'. That is not practical. As ssl breaks caching of many things. By default ssl is not cached both on the browser and at many proxy servers (this is a good thing in many cases). But in this case that picture of you wolfing down hotdogs somewhere probably will never change. Yet it will not be cached if you pipe it over ssl. Not everyone has DSL or higher. There are people who are stuck on dialup and it will not get any better any time soon for them.

    3. Re:Another point is not "missing the point" by epine · · Score: 1

      It also misses the point that Facebook is about *SHARING* data. The idea is you are sharing things with people. If you want to keep things private ... Facebook is not the place to do it.

      Duh! But I did enjoy watching the metal boomerang slice your hand off. I'm sure you won't mind if I share (having assumed your Facebook identity on the sly) that you're dumping your main squeeze because she slept with your boss, but you don't care because you're out now.

      It's unfortunate that authentication and privacy get so badly conflated. The need for SSL certificates derives from the authentication function, but you can't establish a private connection without one, for no technical reason at all, but it's a nice racket. Sometimes I want privacy (e.g. plaintext passwords being exchanged) and sometimes I want authenticity (that only I post under my own identity).

      Hopefully you fingers came off clean and can be surgically repaired. Next time, type smarter.

      Concerning Facebook, it's a no-fly zone for me. I preferred sharing back in the day when sharing was only weakly transitive unless especially ghastly, malicious or inbred.

  9. Promiscuous mode on any adapter? by SpinningCone · · Score: 5, Interesting

    I used to do sniffing and stuff like this a couple years ago and the biggest hurdle was finding a wireless adapter which would allow promiscuous mode. aircrack sells one that comes with 1st party drivers to allow sniffing. I used a linksys usb adapter since there were 3rd party drivers that allowed it.

    unless something has changed I thought most wireless driver didn't support promiscuous mode for sniffing.

    1. Re:Promiscuous mode on any adapter? by maxume · · Score: 1

      On Windows, sure, on Linux, not so much:

      http://backtrack.offensive-security.com/index.php/HCL:Wireless

      --
      Nerd rage is the funniest rage.
    2. Re:Promiscuous mode on any adapter? by SpinningCone · · Score: 1

      yeah it was easier for linux but the plugin doesn't even have linux support yet that's why i'm wondering how it works. even more curious that a browser plugin has that level of access to the system.

    3. Re:Promiscuous mode on any adapter? by Anonymous Coward · · Score: 1, Informative

      Its pretty well documented which cards support promiscuous mode these days. You are correct that most drivers dont support it however. I have an ALFA

      http://www.aircrack-ng.org/doku.php?id=faq&DokuWiki=8a7d493ef4d8a6ce9779a7c8b2f0259a#what_is_the_best_wireless_card_to_buy

    4. Re:Promiscuous mode on any adapter? by Anonymous Coward · · Score: 0

      And Vista improves the situation by introducing a new driver model for wireless drivers that allow for 802.11 monitor mode, but WinPcap does not support that either:
      http://www.winpcap.org/pipermail/winpcap-users/2009-March/003112.html
      MS's NetMon does support it, however.

  10. How does it work? by pinkeen · · Score: 3, Interesting

    The article is extremely light on details. The plugin's page doesn't tell much either. I'm curious how does it capture the WIFI packets. Is it possible to capture them when not in monitor mode?

    1. Re:How does it work? by will_die · · Score: 3, Informative

      You first need to installWinPcap this is the program that does the actual work. You then log on to the wifi, using password if required, and the program starts looking for know cookies. If it finds them it captures the info and gives you a nice userfriendly way of using them.
      It can capture the wifi since anyone can capture them if you are within range of the transmissions. You if you are not monitoring when the signals go out you cannot capture them.

    2. Re:How does it work? by pinkeen · · Score: 3, Interesting

      That wasn't my question. When in monitor (promiscous) mode, adapter can capture but cannot associate and give you internet connection. So, when you capture packets you need another wlan adapter or ethernet nic for your internet conncetion to actually use this stolen cookies. There's no mention of it on the site. So I wondered that maybe the plugin does some magic and captures packets while the same adapter is associated with an ap.

  11. Other People in the Room by SudoGhost · · Score: 2, Insightful

    the realization that squabbles about Facebook's changing privacy settings and various privacy breaches simply miss the point.

    I'm much more concerned about that then someone on my network stealing my password. If they're on my network, they could steal my password? This is not new, nor is it news. The number of people on the internet out to get your personal information is much, much higher than the number of people on your network out to do the same.

    This is just a high-tech version of this:

    'When it comes to user privacy, other people are the elephant in the room,' said SudoGhost, random douchebag author of the post in question, dubbed 'Other People in the Room'. By being in the room and watching the screen/keyboard, anyone can 'sniff out' not only the unencrypted HTTP sessions, but virtually any keystroke, allowing your mom to access social networks, online services and other website requiring a login, and simply hijack them and find out where you really were Saturday night."

    1. Re:Other People in the Room by statusbar · · Score: 3, Insightful

      How many people use wireless at a conference, or a coffee shop, or a hotel?

      --
      ipv6 is my vpn
    2. Re:Other People in the Room by Anonymous Coward · · Score: 0

      I certainly use WiFi in all of those places. The step after making the connection to the WiFi access point is to make a VPN connection to work. After that, my traffic is tunneled securely (at least securely from the point of view of someone on that WiFi connection trying to snoop out unencrypted packets). I wouldn't want to use one of these access points without VPN for anything but checking the weather or something - nothing that requires login or private information that's for sure.

      However, the obvious problem is the "average user" who either doesn't have a clue about security or who doesn't have a VPN connection they can make to up the bar in trying to hack them.

    3. Re:Other People in the Room by gmurray · · Score: 1

      I don't use other people's wifi for the same reason I wouldn't share a junkie's needle.

    4. Re:Other People in the Room by statusbar · · Score: 1

      Of course not, unless you are accessing the net via VPN, or SSH tunnel to a proxy, or of course when you are running firesheep...

      --
      ipv6 is my vpn
  12. Am I the only one who finds it amusing... by Viol8 · · Score: 2, Interesting

    ... that the bleating masses who so readily rushed to put their entire lives and details on social networking sites despite all the warnings are now running around shouting at all the chickens that are coming home to roost?

    For the rest of us with some common sense this is just hilarious.

    1. Re:Am I the only one who finds it amusing... by betterunixthanunix · · Score: 3, Informative

      To be fair, most of those people do not actually care. There is a small minority of users who do care, but for some reason continue to use those websites; the rest just want to follow the crowd without stopping to question anything.

      --
      Palm trees and 8
    2. Re:Am I the only one who finds it amusing... by beh · · Score: 1

      For the rest of us with some common sense this is just hilarious.

      You're making a bad judgment here - there is a lack of common sense in both IT geeks like 'us' und normal users (anyone outside IT).

      The issue with facebook and security has nothing to do with common sense per se, but with IT training. You and I may know a few things about security, which may lead us to accept some things, but reject others.
      People outside IT do not have this type of training, nor would it be easy to bestow it on them. It IS the kind of people (the 90+% of the planet) which can not easily follow what's going on.

      You can partially try and explain it to them, saying that Facebook's login is akin to a 'secret knock' at the kids secret hideout (you know - when we were kids and actually played outside, meeting each other). The password is like the secret knock letting you in. This secret knock has the problem that anyone within earshot will hear it and will likely be able to reproduce it. And - as you may have learnt as a kid, you don't knock on the secret club house door if you see anyone within earshot (who you know doesn't belong).

      It's easy to explain, and a workable analogy. The problem for people now is, that their facebook cookie is a secret knock that needs to be transported halfway around the world in order to 'let you in'/'let you have the data you want'. And this is where the problem sets in - as a kid you may have learnt that anyone at a greater distance can't hear you knock.

      On the Internet, potentially anyone along the way can - and from your home to facebooks servers can be quite a distance.

      The most likely point where your data may be intercepted is still at your own home - and again, you may feel safe, as you don't see anyone you don't know sitting in your own living room with his laptop running (and potentially listening in). The fact that your neighbour can, is already difficult enough to grasp for most people outside IT...

    3. Re:Am I the only one who finds it amusing... by kevinNCSU · · Score: 1

      I hate to break it to you but the intersection of the set of people whom you consider to be the "bleating masses who so readily rushed to put their entire lives and details on social networking sites" and the set of people who read about the opensource project known as firesheep AND are really concerned about someone packet sniffing on their own network and then doing something malicious with it (just logging in is likely completely illegal) is probably incredibly small so no one is running around shouting now that the chickens have come home to roost. This is just a self satisfying notion that people want to believe from their nerd cave, that those dang sheep finally got their comeuppance for frolicking in the sun instead of hunkering down underground and now are all whaling why or why didn't we listen to the nerds, but this simply isn't the case. Not yet anyways.

      Lastly, if you're going to mix animal metaphors at least make them make sense. Sheep don't give a shit about chickens roosting. Now wolves in sheep's clothing on the other hand, that's got some promise in this situation.

    4. Re:Am I the only one who finds it amusing... by mbone · · Score: 1

      Look, I know plenty of people who use Facebook and the like basically as a means to post blogs (or, as "twitter with 420 character posts"). They don't put up anything personally sensitive, but they would still be pissed off if someone stole their info and started putting up posts in support of neo-Nazi child pornography or whatever.

    5. Re:Am I the only one who finds it amusing... by Klinky · · Score: 1

      This is session cookie hijacking, it could be used to spoof your Slashdot credentials just as much as someone's Facebook account. Someone just put "Social Network" in the headline to make it seem more hip. Cookie spoofing has been known since the invention of Cookies.

    6. Re:Am I the only one who finds it amusing... by Viol8 · · Score: 1

      Well for a start mixing metaphors doesn't mean just using 2 in the same sentence and secondly if you think living your life on a social networking site is "frollicking in the sun" then I'd suggest you get out more my friend.

    7. Re:Am I the only one who finds it amusing... by mcgrew · · Score: 1

      those dang sheep finally got their comeuppance for frolicking in the sun instead of hunkering down underground and now are all whaling

      Whaling is against international treaty these days, haven't you heard? And what does hunting whales have to do with facebook, anyway?

    8. Re:Am I the only one who finds it amusing... by kevinNCSU · · Score: 1

      The Council of Wool, being the governing body of Sheep, never signed onto that treaty due to the long history of conflict in sheep-whale relations.

      God catch ;)

    9. Re:Am I the only one who finds it amusing... by Anonymous Coward · · Score: 0

      With this in the news, they just might start caring.

      But what will the NSA do if everyone starts using SSL?

    10. Re:Am I the only one who finds it amusing... by Anonymous Coward · · Score: 0

      Yep, it is off-topic, and if I had mod points, I would mod it as such.

  13. No HTTPS encryption by DrYak · · Score: 4, Insightful

    Kudos to FaceBook and most other networks for NOT using encryption for anything but the log in, making such hacks possible !
    I know that HTTPS would put some stress on the servers, specially with something as big as Facebook.
    But, come-on. Social networks have become so important for some people, that the risks of vandalism/identiy spoof/deffamation, etc. are significant and would benefit from some more protection.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
    1. Re:No HTTPS encryption by cindyann · · Score: 1

      Since when have they used SSL/HTTPS as the default for anything, let along signing on?

      I still have to manually change http to https in the URL every time they decide to sign me off.

    2. Re:No HTTPS encryption by muckracer · · Score: 4, Informative

      > Kudos to FaceBook and most other networks for NOT using encryption for anything but the log in [--DrYak]

      > I still have to manually change http to https in the URL every time they decide to sign me off. [--cindyann]

      Install the HTTPS-Everywhere FF Plugin. It will SSL-encrypt Facebook and a host of other domains. Only draw-back: Chat doesn't work via SSL atm.

      https://www.eff.org/https-everywhere

      And while you're at it, also install the BetterPrivacy Add-on:

      https://addons.mozilla.org/en-US/firefox/addon/6623/

      which will get rid of the LSO cookie Facebook sets each time you use it. Best used in conjunction with AskforSanitize.

    3. Re:No HTTPS encryption by FrostDust · · Score: 2, Interesting

      Do they have any guarantee that all of their users have a browser that supports HTTPS?

      To Facebook, it's better to allow access to as many users as possible, than lock some out in the name of security.

    4. Re:No HTTPS encryption by lavagolemking · · Score: 4, Informative

      Facebook does submit your information over HTTPS; they just load the page over HTTP by default. Passive sniffing won't work on it. Here, take a look at the following code from http://www.facebook.com/:

      <form method="POST" action="https://login.facebook.com/login.php?login_attempt=1" id="login_form">

      The problem with this approach is, while it saves server resources, an attacker could trivially perform a man-in-the-middle attack on an average person connecting to http://www.facebook.com/ rewriting the above code to HTTP or running a squid proxy or something, and they would never notice because their browser says "http" like always.

      That said, if you're worried about it you could always install HTTPS Everywhere and it will make Facebook always load using SSL.

    5. Re:No HTTPS encryption by Confusador · · Score: 3, Insightful

      There's a world of difference between having a fallback for those who can't use the secure site (with a warning that it is not secure, even) and not having an option for those who can.

    6. Re:No HTTPS encryption by IAmGarethAdams · · Score: 1

      Well, the problem being exposed isn't related to the login form and password sniffing, but merely that after login a stolen session cookie gives you full access to someone's logged-in session (as it is in fact designed to) The problem is potentially less of an issue than having the password submission over HTTP - once you have a user's Facebook password, chances are you have the password to some of their other online accounts, whereas this attack only gives you access to the current logged in session - but it's definitely an issue that users needs to be aware of

    7. Re:No HTTPS encryption by kriebz · · Score: 1

      Are there such things? I mean, really... find me a browser that has enough Javascript support to have any hope of rendering Facebook, that doesn't at least support 128-bit RSA SSL. It's not 1998.

    8. Re:No HTTPS encryption by sznupi · · Score: 2, Informative

      http://m.facebook.com/ ...not saying the mobile browsers can't have the security, just that "hope" isn't required to render Facebook without js.
      And apparently such access is quite popular - there were some news from FB itself about explosive growth; also according to stats of Opera Mini (the #1 mobile web browser worldwide by site hits, despite many of its users being evidently rather frugal with numbers of sites visited / data transferred), Facebook is quite often near the top of popularity.

      --
      One that hath name thou can not otter
    9. Re:No HTTPS encryption by Matt+Perry · · Score: 1

      You don't even need HTTPS. HTTP already supports authentication mechanisms. If we'd use digest authentication for logins then we wouldn't have to bother with cookies at all. Unfortunately, there's no way to make a pretty login page for digest (or plain) authentication. The browser pops up a username/password dialog instead. Therefore, web sites avoid it and opt instead for the mess of cookies and all their security issues.

      --
      Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
    10. Re:No HTTPS encryption by miserere+nobis · · Score: 1

      Yes, but the same thing works for other sites as well, like Yahoo, which is a mail provider, and if you can hijack someone's email account, you can then visit other sites' "forgot my password" links and get passwords or password reset links sent to the mail account, and thereby gain access to everything else, too. So it is a bigger breach than just the existing session on the existing site.

    11. Re:No HTTPS encryption by Anonymous Coward · · Score: 0

      Unless there are any HTTP authentication mechanisms that prevent session hijacking from sniffers, you'll still need HTTPS.

    12. Re:No HTTPS encryption by Anonymous Coward · · Score: 0

      They could use an authentication scheme that changes the persistence token on every request... sure you can still have your session hijacked... but at least you'd know because you got logged out. It would probably be lighter than HTTPS and would narrow the window of opportunity. Of course, you are right that HTTPS would be a better solution.

  14. Cookie theft by Securityemo · · Score: 5, Insightful

    It's "just" WiFi cookie theft. You can do that easily with wireshark and copy/paste, this just makes it a bit faster. The problem lies in session cookies, and this is a problem known for what, almost a decade now?

    --
    Emotions! In your brain!
  15. It's news in that people STILL don't get it by Toe,+The · · Score: 1

    The news is that still hardly anyone understands SSL or what it is for.

    People like to see that little lock sign (or whatever obscure message their browser displays) when they log into their bank. But I sincerely doubt that the great majority of people have any idea that things like e-mail transactions can be routed over SSL or why that might be a good (i.e., critically important) idea.

    Just scan your local neighborhood and look at (for an analogous example) how many people are still using WEP and thinking that somehow they are protecting themselves.

    1. Re:It's news in that people STILL don't get it by icebraining · · Score: 1

      Email (IMAPS/SMTPS to your server) over SSL is nice but ultimately irrelevant, as you don't know if the rest of the path is encrypted. Only OpenPGP is safe.

      WEP is similar; it's not a real protection, but stops the random kid trying to use your 'net to download stuff.

    2. Re:It's news in that people STILL don't get it by Toe,+The · · Score: 1

      My point is about revealing passwords. If you're communicating with an e-mail server over insecure protocols, what's protecting your password from being snooped?

      It blows my mind to think of people using insecure mail protocols over open or readily hackable wifi networks, using the same password for everything.

      It is like everyone is walking around with their wallet dragging ten feet behind them by a bit of weak thread. They're practically insisting on being robbed by any nearby thief.

    3. Re:It's news in that people STILL don't get it by Anonymous Coward · · Score: 0

      I use WEP just so the only people who get on my network are the ones who really want it. If you're going to steal my internets I expect you to have to do one or two more clicks at least!

  16. Er, It's the lack of SSL by mbone · · Score: 1

    It is the lack of SSL that is the problem here, and it is the non-use of SSL that 'is the elephant in the room,'

    This has been going on for a long time now - attend a NANOG meeting and use unencrypted logins, and you may well see your password on the screen by the end of the meeting - the white hat guys routinely sniff the wireless for passwords.

  17. The next step by Fusione · · Score: 1

    Public wifi isn't secure.. in other slashdot news, water is wet and fire burns. Really though, the next step in this equation is for someone to run this at a really busy hotspot in NYC, and then to anonymously publish the results online. Bang! media coverage. Bang! reputation loss and user defection for compromised services. Bang! solution for problem gains financial incentive, and gets fixed.

    1. Re:The next step by Anonymous Coward · · Score: 0

      Publish what? The session cookies?

  18. WPA2 will work better against this hack by whitesea · · Score: 1

    What is the problem? Protect your WiFi connection with WPA2 and this hack does not work. All around me almost any network is protected and these are regular folks, not some security gurus. Yes, their information may be stolen further down the wire, but this is not new. While I am all for SSL protection, this particular hack can be fought off by individual users. Even more, while HTTPS has to protect each individual site you go to, WPA2 creates a secure wireless tunnel that protects all your communications. Move along, nothing to see here :-).

    1. Re:WPA2 will work better against this hack by Instant_Karmma · · Score: 2, Interesting

      This works on any network segment, including wired. How many people do you know that use Facebook, Amazon, etc. from their desks? Sure, your traffic could always be monitored by the PFY's in the data center, but now your pointy-haired boss has a tool that allows him to see what you've been buying. No thanks.

    2. Re:WPA2 will work better against this hack by Anonymous Coward · · Score: 0

      This works on any network segment, including wired.

      How many people do you know that use Facebook, Amazon, etc. from their desks? Sure, your traffic could always be monitored by the PFY's in the data center, but now your pointy-haired boss has a tool that allows him to see what you've been buying.

      No thanks.

      Ah no. In most wired networks, the data is switched, and only travels down the wires it needs to. Your packets are not broadcasted to your boss's network jack. Unless he asks IT to set that up in the switch anyway.

    3. Re:WPA2 will work better against this hack by crunzh · · Score: 1

      Only if its not a switched network. WPA2 will work against this as the users can't see each others traffic.

      --
      Visit http://www.crunzh.com/ for free software. Mac/Lin/Win
    4. Re:WPA2 will work better against this hack by Mashiara · · Score: 1

      Actually on wired network it depends on the switching hardware whether you're getting packets meant for others on your port or not (discounting active mac/arp spoofing but with properly configured high-end HW you will find yourself in an isolated network segment really quickly if you try that)

    5. Re:WPA2 will work better against this hack by Archangel+Michael · · Score: 1

      Your scenario would require more than just plugging into a network, it would require a switch in promiscuous mode or a hub. A normal switch the traffic is not sent to all ports on the network; only broadcast traffic goes to all ports.

      And if you're running switches or a hacker has access enough to your switches to turn on Promiscuous mode, you have other problems, and securing web traffic is not one of the big ones.

      The fact that you got modded "Interesting" on /. makes me sad.

      --
      Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    6. Re:WPA2 will work better against this hack by imemyself · · Score: 1

      Not sure what all of the other people posting on this topic were thinking. Have any of you heard of ARP poisoning? Having a switched network offers little to no protection from MITM attacks unless you and your attacker are on separate L3 networks.

      --
      Every time you post an article on Slashdot, I kill a server. Think of the servers!
    7. Re:WPA2 will work better against this hack by yuhong · · Score: 1

      And don't forget Hole196 too, allowing ARP poisoning over WPA(2).

  19. But will it... by koterica · · Score: 1, Interesting

    run (on) linux? Apparently not. I guess I wont be using it.

  20. Mobile Apps by Don_dumb · · Score: 1

    It seems that this is most concerning for those loggining in while using public networks (such as accessing with a cafe's WiFi).

    So this leads me to ask if I am safer when using the Facebook/Amazon/eBay app rather than the mobile browser. Is the security of the iPhone or android apps better than the web security for Facebook?
    Or can I make my access of these sites more secure myself somehow?

    --
    If this were really happening, what would you think?
    1. Re:Mobile Apps by DJ+Rubbie · · Score: 1

      One way to find out is to do a promiscuous tcpdump of your local network traffic while using that app - if you can read personally identifiable items in plain text, you are simply not safe.

      --
      Please direct all bug reports to /dev/null
  21. Use md5 (or something) over the wire by Compaqt · · Score: 3, Informative

    Leaving aside md5 cracks (use another algo if you want):

    md5 the password with Javascript on the client end before sending it. Then un-md5 it with PHP on the server.

    Plenty of security-conscious CMS's have been doing this before Mark Z even thought of an electronic facebook.

    --
    I'm not a lawyer, but I play one on the Internet. Blog
    1. Re:Use md5 (or something) over the wire by gmurray · · Score: 2, Insightful

      md5 is a hash algorithm. How would that help? If someone can snoop your md5 hash they can replay it to gain access to the server, and then change your password (provided the server doesn't provide a challenge to perform this action). All md5 does is protect your actual password, which is small protection if your account can be illicitly accessed anyway. None of these services send a password in plaintext (hopefully). That isn't the issue. The issue is that they use replayable tokens and don't use encryption to send them on the wire.

    2. Re:Use md5 (or something) over the wire by gmurray · · Score: 5, Insightful

      furthermore the entire usefulness of md5 is that you can't un-md5 it ;-)

    3. Re:Use md5 (or something) over the wire by ogapo · · Score: 2, Informative

      I think you may not understand how a cryptographic hash works. In the scheme you are describing, the password is typically hashed on the client side (along with some value specified by the server which changes every time). When the server gets the hash, it hashes the password (as stored in the DB and possibly also hashed) along with the same value and compares the result. Regardless, what this plugin does is not steal passwords, but simply looks for authenticated credentials (usually cookies). See, once you authenticate, the server gives you a cookie (your session identifier) that you pass back with every request to prove you are who you say you are. Since the traffic is not encrypted, this can be intercepted by anyone on a network between you and the facebook servers. If you live on a college campus or work for an ISP, this could very well be many people. Even if Facebook is smart enough to tie this session to your IP, it's likely that someone in a correct network position to sniff your packets can also viably spoof your IP (both sending and receiving). This is effectively the same as them hijacking your account except the ability goes away when your session expires.

    4. Re:Use md5 (or something) over the wire by Anonymous Coward · · Score: 2, Insightful

      This won't work as the extension sniffs out cookies, not passwords.

      Even then, this won't help as the extension could be changed to sniff the hashed password (it's just send as plain text over HTTP), and send that hash itself.

    5. Re:Use md5 (or something) over the wire by Culture20 · · Score: 4, Funny

      md5 is a hash algorithm. How would that help? If someone can snoop your md5 hash they can replay it to gain access to the server, and then change your password (provided the server doesn't provide a challenge to perform this action). All md5 does is protect your actual password, which is small protection if your account can be illicitly accessed anyway. None of these services send a password in plaintext (hopefully). That isn't the issue. The issue is that they use replayable tokens and don't use encryption to send them on the wire.

      Well, then md5 the hash. It's just like using triple-DES or double rot-13 (one of the two, or maybe a happy middle-ground). ;)

    6. Re:Use md5 (or something) over the wire by jwietelmann · · Score: 4, Informative

      Hash = 1-way crypto

      The only way to "un-md5" anything is to crack it. Also, I'm not sure you actually put any real thought into this.

      Since it's best practice to store only password hashes (and not the passwords themselves) in your database (or whatever), your process is apparently:

      1. Client md5's the password, sends it to server
      2. Server "un-md5"s the password (let's say for argument's sake that this makes perfect sense)
      3. Server md5's the un-md5'd password
      4. Server checks hash against user's hash in the database
    7. Re:Use md5 (or something) over the wire by Mashiara · · Score: 2, Informative

      You are missing the point.

      The problem is not reading the password as plaintext from the cookie (now that would be monumentally stupid design) but that since the cookie equals valid session authentication copying the cookie equals session hijacking (or sidejacking since the original cookie is still there on the original users machine).

    8. Re:Use md5 (or something) over the wire by Mashiara · · Score: 1

      The login forms/submissions AFAIUnderstand do go over SSL so doing encrypting the password is kinda pointless there.

      Also there is no such thing as "un-md5", now the password might be encrypted (des/aes/whatever) but hashes are by definition one-way.

    9. Re:Use md5 (or something) over the wire by gmurray · · Score: 1

      How exactly would that help? You could md5 hash a password and a timestamp, and this would at least limit the amount of time that a hashed password could be replayed, but it would not prevent the replay of the password. The nature of a hash is that it isn't something that you decode. It obscures something from view, so that a party on the other end, if it knows the same secret, can verify that you know the secret, without divulging the secret publicly.

      But if someone can snoop your hash, they can replay it and pretend they know the secret, without actually knowing it.

      This is why a hash protects the secret, but doesn't protect the service from replay attacks, you need encryption also.

      A hash is a good idea to be used in concert with encryption because then, even if the encryption is broken, the secret is not exposed. But a hash in itself is not a secure way to assert identity.

    10. Re:Use md5 (or something) over the wire by Anonymous Coward · · Score: 0

      1) This about cookies, not passwords
      2) Your idea fails to protect passwords, sure the hacker won't know the actual password but he can just send the md5 like the javascript would.

    11. Re:Use md5 (or something) over the wire by Culture20 · · Score: 1

      Woosh. The double-rot13 wasn't a clue?

    12. Re:Use md5 (or something) over the wire by gmurray · · Score: 1

      my sarcasm detector needs more coffee, still, my comments may offer more elucidation to the original replier, so I'm not sorry I ducked under your humor.

    13. Re:Use md5 (or something) over the wire by PatPending · · Score: 2, Funny

      md5 the password with Javascript on the client end before sending it. Then un-md5 it with PHP on the server.

      Or use quad-ROT13 instead.

      --
      What one fool can do, another can. (Ancient Simian Proverb)
    14. Re:Use md5 (or something) over the wire by Thinboy00 · · Score: 1

      Leaving aside md5 cracks, WTF do you mean by "un-md5 it"? You can't do that!

      --
      $ make available
    15. Re:Use md5 (or something) over the wire by ghjm · · Score: 1

      Or, depending on who you are, the usefulness might be that no-one else can.

    16. Re:Use md5 (or something) over the wire by Compaqt · · Score: 1

      Did I really say un-md5 ?!! Sorry, I meant "compare the hash sent by the client to that saved in the DB".

      Even so, this technique uses cookies, and not the password or hash. (Note to self: Read the articles!)

      Typo3 is one CMS that you can set to check the incoming IP and make sure it's the same as the IP that originally authenticated.

      Drupal 6 is abysmal in that it doesn't even use salt; probably half the passwords in table users are likely to be in an md5 database somewhere.

      --
      I'm not a lawyer, but I play one on the Internet. Blog
    17. Re:Use md5 (or something) over the wire by Compaqt · · Score: 1

      Sorry, man, you caught me. Lesson: Don't post while drowsy.

      un-md5?

      Is that slated to be the next Slashdot meme?

      --
      I'm not a lawyer, but I play one on the Internet. Blog
    18. Re:Use md5 (or something) over the wire by Compaqt · · Score: 1

      Please somebody mod my original post as Funny and not Informative to avoid future PHP-Nukes.

      What you actually need to do at the very least is:

      1. md5 (or another algo) with Javascript on the client and compare that hash to the one saved in the DB. If the password is stored in cleartext (which it shouldn't be, but sometimes external systems are out of your control), md5 it with PHP.

      2. Some people use SSL on the login page.

      3. But this attack shows crackers just intercepting an replaying the creds. Discouraging that might involve IP or other checks. Defeating it might involve total encryption.

      --
      I'm not a lawyer, but I play one on the Internet. Blog
    19. Re:Use md5 (or something) over the wire by Compaqt · · Score: 1

      By now you may have seen my follow-up that I mistyped when I said "un-md5" (meant compare hashes on the server).

      But I disagree that all logins (even for large sites) are encrypted.

      For example, I use Slashdot just fine without JavaScript. I haven't checked the source, but the standard HTML FORM element doesn't encrypt anything when sending form submissions over the network. So the password must obviously be sent (at least the first time) in the clear.

      That's why I was encouraging people to md5 their passwords on the client before sending it over. That won't stop this attack, but it'll stop others (security in-depth).

      --
      I'm not a lawyer, but I play one on the Internet. Blog
    20. Re:Use md5 (or something) over the wire by Tepic++ · · Score: 1

      If you're trying to implement login like this and HTTP Digest authentication isn't an option, then I'd suggest reading the HTTP Digest authentication spec. and implementing that using URL parameters. I think that should actually then get you a fairly secure login.

      This isn't really related to the issues in the article though.

    21. Re:Use md5 (or something) over the wire by gmurray · · Score: 1

      But this attack shows crackers just intercepting an replaying the creds. Discouraging that might involve IP or other checks. Defeating it might involve total encryption.

      Anything that you send in the clear to assert your identity can be replayed. IPs etc are easy to spoof so are not an adequate counter. You can include a timestamp in the hashed information so that the hashed info cannot be reused more than a certain amount of time after it is generated, but you have to allow for transmission delay and the server's clocks being out of sync, so if an attacker is quick enough they can replay your tokens even if you have made them time sensitive.

      Using a request counter + timestamp or a one time password in the token would be much more preferable, but is more expensive to assert with each request, and is still suseptible to interception and spoof attacks, if not replay attacks.

      There's really very little that is a valid substitution for encrypting all traffic here, IMO. The rampant use of unencrypted transmission of tokens on these sites today rely on the fact that it is harder to stage a man in the middle attack once requests leave your local network. But as tools for use on public wi-fi networks become easier to use and more prevalent this is only going to become a larger and larger problem.

    22. Re:Use md5 (or something) over the wire by nomorecwrd · · Score: 2, Funny

      Bettter yet 1024-ROT13... it's a little time consuming, but totally worth it.

    23. Re:Use md5 (or something) over the wire by t1oracle · · Score: 1

      You can prevent replays by generating a random one time use key and using that in the MD5 hash on both ends to confirm the secrete and the random key. That way, once it has been used once it is no longer valid thus defeating the replay. Unfortunately, the login process is only one issue. You still have to protect the session id or they can hijack that and gain access.

    24. Re:Use md5 (or something) over the wire by shmlco · · Score: 1

      "Typo3 is one CMS that you can set to check the incoming IP and make sure it's the same as the IP that originally authenticated."

      If you snatched someone's cookies over free WiFi at a coffeehouse, you probably HAVE the same IP address as they do, since all the server sees is the coffeehouse's gateway IP address.

      IOW, that won't help either.

      --
      Any sect, cult, or religion will legislate its creed into law if it acquires the political power to do so.
    25. Re:Use md5 (or something) over the wire by t1oracle · · Score: 1

      You don't un-MD5, you MD5 on both ends. MD5 the password on the client, MD5 on the server, compare the results. The hash will have collisions so this does introduce the possibility of false positives, but that should be small and hard to predict without knowing the password for a good hash algorithm.

    26. Re:Use md5 (or something) over the wire by gmurray · · Score: 1

      As I noted in my post, your suggestion still leaves you open to interception attacks. Which are, admittedly harder to perform than replay attacks, but are nonetheless problematic. My statement that all tokens sent in the clear were replayable was a bit inaccurate though, yes, as I contradicted with suggestions later in the same post! :)

    27. Re:Use md5 (or something) over the wire by gmurray · · Score: 1

      uh... I'm referring to a post further down :)

    28. Re:Use md5 (or something) over the wire by Anachragnome · · Score: 1

      "If you snatched someone's cookies over free WiFi at a coffeehouse, you probably HAVE the same IP address as they do, since all the server sees is the coffeehouse's gateway IP address."

      "Even then, this won't help as the extension could be changed to sniff the hashed password (it's just send as plain text over HTTP), and send that hash itself."

      All this talk of sniffing cookies and hash has got my stomach rumbling.

      Enough already.

    29. Re:Use md5 (or something) over the wire by hairyfeet · · Score: 1

      Plus I'd add that tying ANYTHING, from authentication to bans, to an IP address given that we are running out of IPV4 and most likely will be seeing more and more bigger and bigger NATs is a bad idea to beat ALL bad ideas.

      While I'm not into social junk and therefor don't have a clue how they authenticate (but really not surprised if its easily hacked, considering the amount of spam I've gotten from friends that use FB) tying ANYTHING into IP before we've had the switch to IPV6 is just a universally BAD idea.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    30. Re:Use md5 (or something) over the wire by Mashiara · · Score: 1

      By now you may have seen my follow-up that I mistyped when I said "un-md5" (meant compare hashes on the server).

      Yup, that was in fact a completely secondary point to me as I first thought that is what you *must* have meant , since hashes are not reversible, and only seconds later decided that maybe pointing the fact out might be a good idea.

      But I disagree that all logins (even for large sites) are encrypted.

      For example, I use Slashdot just fine without JavaScript. I haven't checked the source, but the standard HTML FORM element doesn't encrypt anything when sending form submissions over the network. So the password must obviously be sent (at least the first time) in the clear.

      That's why I was encouraging people to md5 their passwords on the client before sending it over. That won't stop this attack, but it'll stop others (security in-depth).

      A fair point, though that would require either plaintext passwords (*very* bad) or unsalted passwords (slightly bad) in db (or first validating the username to get the specific users salt to be passed on to the client but that is again rather bad).

      IMO someone attacking the server gaining access to wholesale set of plaintext or unsalted (rainbow tables here we come) usernames&passwords is in fact worse than someone sniffing plaintext passwords in POSTs in you network segment (or between you and server but that's less likely).

    31. Re:Use md5 (or something) over the wire by uninformedLuddite · · Score: 1

      That wasn't a duck! That was a squirrel

      --
      The new right fascists are bilingual. They speak English and Bullshit.
    32. Re:Use md5 (or something) over the wire by uninformedLuddite · · Score: 1

      mmmmmmm, hash

      --
      The new right fascists are bilingual. They speak English and Bullshit.
    33. Re:Use md5 (or something) over the wire by Terrasque · · Score: 1

      In addition to the things other posters have pointed out, how will this actually change anything?

      Scenario A:
      You send clear pw to server, server verifies it, sets cookie.
      Dr. Evil steal password, sends the password to server, and verifies as them (or just gobble cookie, but lets do it the hard way)

      Scenario B:
      You send MD5 pw to to server, server verifies it, sets cookie.
      Dr. Evil steal the MD5'ed password, sends the MD5 to server, and verifies as them.

      You haven't really changed anything. It's still the same replay attack. The password itself is obscured, that's true. But now you don't need the password anymore, so no problem.

      You'd need a challenge / response system, where the response is calculated (with challenge + pw) via JS in the browser, and then the response is verified by server. More complicated, but doable. Still MITM vulnerable. Cookie is still insecure.

      At the end of the day, instead of chasing your own tail, you either give up or go SSL. Or you cook up some shit that at least look pretty complex.
      Which leads me further to :

      Plenty of security-conscious CMS's have been doing this before Mark Z even thought of an electronic facebook.

      Fixed : Plenty of clueless CMS's have been doing hare-brained stuff before Mark Z even thought of an electronic facebook.

      --
      It's The Golden Rule: "He who has the gold makes the rules."
    34. Re:Use md5 (or something) over the wire by Anonymous Coward · · Score: 0

      I'd be interested in knowing how do you un-md5...

    35. Re:Use md5 (or something) over the wire by riegel · · Score: 1

      lost humor aside...

      You make a good point. So my comments are to the OP.

      If I use Client side Javascipt to generate the MD5 hash then the "secret" key has to be sent to the client. Otherwise how would it know the secret key to use to create the hash.

      --
      http://p8ste.com - Web based Clipboard
    36. Re:Use md5 (or something) over the wire by LordLimecat · · Score: 1

      Im not sure why this was modded informative, since you cant "un-md5" anything-- as a hash algorithm, it is a 1-way thing. Even brute-forcing wont necessarily reveal the original data-- unlike encryption which is a 1-to-1 mapping (plaintext X always maps to crypted Y, and vice versa), hashing is a many-to-1 mapping-- like lossy encrpyion, you lose some of the original data in the hashing.

      The solution is to use SSL, which people already do (in theory) when connecting to signin pages. But that doesnt eliminate MITM attacks if users ignore the SSL warnings. At the end of the day, if the user is going to connect to unencrypted WiFi and ignore SSL warnings, there is very little that can be done.

    37. Re:Use md5 (or something) over the wire by Compaqt · · Score: 1

      Well, it (my comment) was Funny just yesterday. Somebody please re-mod it to Funny.

      --
      I'm not a lawyer, but I play one on the Internet. Blog
    38. Re:Use md5 (or something) over the wire by ecotax · · Score: 1

      And what are step 2 and 3 good for? The folllowing (just step 1 and 4) should be fine:

      1. Client md5's the password, sends it to server
      2. Server checks hash against user's hash in the database
      --
      "Money is a sign of poverty." - Iain Banks
    39. Re:Use md5 (or something) over the wire by Anonymous Coward · · Score: 0

      Now you just need to intercept the MD5 and use that instead of the actual password.

    40. Re:Use md5 (or something) over the wire by Chrisq · · Score: 1

      When you realise what GP was saying it will be a "Doh" moment.

    41. Re:Use md5 (or something) over the wire by Alex+Belits · · Score: 1

      gb2/g/

      (ths is just like gb2/b/ but without any implication of potential epicness).

      --
      Contrary to the popular belief, there indeed is no God.
  22. My comments by formfeed · · Score: 2, Funny
    I'd like to declare that all comments under my user name that are controversial or could get me in trouble were made by someone else.

    Someone, who obviously must have sniffed out my wireless cookies. -Shame on them.

  23. Re:Er, It's the lack of SSL -- not needed by Anonymous Coward · · Score: 0

    I know you wouldn't be arrogant enough to try to invent your own encryption algorithm, so don't be dumb enough to try to invent your own authentication protocol.

    Follow Microsoft and steal it from the best: http://web.mit.edu/Kerberos/#what_is

    Invented by the professors at MIT to secure their own logins from the students at MIT. A.k.a. "forged in the fires of Mount Doom" and so safe from most threats.

  24. I must be really really old... by X.25 · · Score: 1

    I really miss the old good days, where talks on security conferences would blow you away, and where people would actually talk about new security related things, rather than showing 76th way of automating a process/procedure that has been known for 10 years (always involving grabbing [flavor of the month service]'s password).

    Oh well, guess people were in security world for different reasons 10 years ago...

    1. Re:I must be really really old... by Anonymous Coward · · Score: 0

      Maybe they're just making the point that not only is shit not getting fixed, the exploit works through a plugin.

    2. Re:I must be really really old... by Hartree · · Score: 1

      I must be really really old...

      Yeah, you are. And me too. I bet a lot of the young'uns here have never heard of the protocol in your username.

  25. Where are the Google apologists now? by Anonymous Coward · · Score: 0

    Where are the guys who keeping saying sending out unencrypted packets are the users' fault?

    Hey, you should know your connection to Facebook is not encrypted, so anyone sniffing your packets is your own fault.

    Oh, this rule only applies when otherwise Google would be blamed? My bad.

  26. Spread this by Amorymeltzer · · Score: 1

    This needs to be heard by everyone. NOW. Sure, your New York Times access is largely trivial, but Facebook and gmail access? That's someone's life. Amazon, and soon Netflix, PayPal, and eBay? That's someone's money. Maybe once people start losing money and their jobs websites will realize the severity of security, as that's usually when it hits home. But until then, very neat.

    Protect yourself: https://addons.mozilla.org/en-US/firefox/addon/12714/

    --
    I live in constant fear of the Coming of the Red Spiders.
  27. KB SSL Enforcer by brunes69 · · Score: 2, Interesting

    This is why I use this Chrome extension - https://chrome.google.com/extensions/detail/flcpelgcagfhfoegekianiofphddckof

    Basically for any site you go to it AUTOMATICALLY redirects you to the SSL version of that site if it exists. Including ssl.facebook.com.

    Yes ssl.facebook.com should be the default, as should most sites, but until they are this extension is invaluable IMO.

    1. Re:KB SSL Enforcer by jspraul · · Score: 1

      FYI, that extension can still leak cookies on the first request because of limitations in Chrome's extension API. See the page you linked for more details.

    2. Re:KB SSL Enforcer by Anonymous Coward · · Score: 0

      Unlike the Firefox extension (HTTPS Everywhere), this apparently does hit the non-SSL version of the page if that is what you type in. So someone sniffing has a chance to grab your cookies then, no?

      Still it's better than nothing, and I'm going to start using it (switched to Chrome recently) in case one of my coworkers is sniffing.

  28. eBay Gaming by Anonymous Coward · · Score: 1, Interesting

    I get so irritated with the gaming of eBay's system. It wasn't until I put up a nice guitar that I learned if you want to get a good deal on something, bid as a brand-new user. What was happening was some brand-new user was bidding on my auction in the smallest increments allowed. This made it look like the auction was being bid up by me, so probably a lot of eBay regulars didn't want to bid because of this douche. I didn't really notice or understand what was happening until the douche won the auction at a price that was hundreds below where the guitar should have sold. And I'm too effing moral to do the other eBay game I hate which is when a seller cancels an auction simply because the price is too low. I've had that happen on a number of occasions where it was very clear I was going to get a sweet deal on an auction and seller suddenly cancels. Of course, you place a bid and there's no fucking way to get out of it without getting permission from the seller, but if you're a seller you can get out of the auction any damn time you please.

  29. I don't like sending data over an open channel... by Antony+T+Curtis · · Score: 1

    When I am using public WiFi, I tend to SSH-tunnel to my proxy at home for web browsing,
    It usually makes for a better browsing experience too because DNS on public WiFi usually sucks and the compression over SSH means that most web pages loads quicker.

    --
    No sig. Move along - nothing to see here.
  30. Woohoo! Easy to use tools ... by Anonymous Coward · · Score: 0

    Many of my "personalities" love this: The Professor is eager to learn who is twittering about his cheap hairpiece rather than taking notes on the lecture. Gods-gift-to-women wants to learn the name of the cutie updating facebook in the coffee shop. We're going to have great fun! (Of course, this also means no more nefarious activities on the wifi in my therapist's waiting room. She's concluded some of us are paranoid.)

  31. oh lawdy by Anonymous Coward · · Score: 0

    I can't wait for 4chan to get their shit-disturbing hands on this.

  32. ALL network communication should be encrypted by Danathar · · Score: 1

    Although I'm not holding my breath for IPv6 to be widely adopted any time soon....the fact that encryption is mandated in the protocol as an option is something that is long overdue. Clear text non-encrypted network traffic is something everybody should avoid if possible. (which is REALLY hard without a lot of work).

    Maybe if encryption was mandated in packets sniffing this sort of stuff would not be a issue? (yes)

  33. Nice hack, James by Anonymous Coward · · Score: 0

    Thank you for putting in the hands of Joe User and his ten-times-a-day script kiddie son, a nice jacking extension for Firefox. Sure, unencrypted HTTP is crap security, and it has been for ages. But the web won't go all-SSL over night as a result of your ingenious little idea. And in the meantime you'll have caused of a lot of misery with your shitty ass cracking toy.

  34. Not that hard to block by filebottle · · Score: 1

    This should be easy to block on the server side without using SSL. Just save the IP address in the session when you first assign the session id, and then check the ip each time the session id is used. I haven't had a chance to test it yet but I put the code up here: http://filebottle.com/FireSheepBlock.html

    1. Re:Not that hard to block by Terrasque · · Score: 1

      A small flaw in your brilliant plan..

      This works over local network segment.. So, if you sit on the same network as the victim, theres one thing you usually have in common.. No, its not the haircut. You use the same gateway and public ip as the victim.

      Thank you for playing, come again.

      --
      It's The Golden Rule: "He who has the gold makes the rules."
  35. Speaking for myself (who I am)... by seebs · · Score: 1

    Speaking as seebs, who I actually am, I think this addon is a brilliant example of the importance of making a threat concrete and specific in order for people to understand it. I, for one, welcome our new us overlords.

    Consider:

    http://www.csd.uwo.ca/staff/magi/personal/humour/Computer_Folklore/Robin%20Hood%20And%20Friar%20Tuck.html

    This is not a new technique. This is not a bad thing, particularly. And compared to the severity of the problem, I think it's pretty tastefully understated.

    And again, this is actually seebs. Really!

    --
    My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
  36. The real threat is in sniffing mobile 3G data by kintek · · Score: 1

    I work for a web design brisbane company called Kintek. Interesting, however it still uses wincap which means its actually sniffing the LAN packets or ARP poisoning to see other users insecure data correct? Most websites dont re-direct you to https before asking you to login which basically means your login details are sent insecurely over the LAN before they get to the internet.

    Theres a cool tool which lets you see all this already called Cain and Abel: http://www.oxid.it/cain.html

    Using ARP poisioning you can actually man in the middle anyone on a LAN unless they use https or anti ARP poisoning tools.

    The real security threat will probably come down to sniffing mobile phone 3G data. And I wonder whats possible in that realm. If you can send it you can receive it, and I doubt its encrypted well enough to prevent reading, especially when the Police want the ability to pick messages out of the airALA The Wire. :)

  37. Indeed by DrYak · · Score: 1

    If we'd use digest authentication for logins then we wouldn't have to bother with cookies at all.

    Unless there are any HTTP authentication mechanisms that prevent session hijacking from sniffers, you'll still need HTTPS.

    Highlighted for you. Digest-authentication is replay-resistant and MITM resistant (but not spoof-resistant), because the exchanged data change at each request :
    - at each request, the web server sends a different salt, and test if the hash of the salt + hashed-password corresponds to what it predicted.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
  38. "Bleating masses"? by PCM2 · · Score: 1

    You should try Facebook sometime; I bet you're loads of fun at a party.

    --
    Breakfast served all day!
  39. Re:Details... by KingSkippus · · Score: 1

    The OP worded it badly

    What they meant is that most CMSes store the password in the database as an MD5 (or some other algorithm) hash. The technique he's describing is to hash the password client-side and send the MD5 over the wire to the server, which can then compare it to what's stored in the database to grant access.

    Still, while better than nothing, there are a few problems with this, including:

    • Using MD5 or any other algorithm is susceptible to so-called "rainbow table" attacks. For example, if I see 5f4dcc3b5aa765d61d8327deb882cf99 floating across the wire, I know the user's password without any fancy decryption techniques. There are rainbow tables for just about every algorithm out there. This can be thwarted by adding a password "salt."
    • It depends on the user having javascript enabled. While most browsers do, some people (like me) use things like FlashBlock to avoid annoyances and sundry evil.
    • As mentioned above, it does little to counter replay attacks, where someone who has sniffed the MD5 hash off the wire simply sends it to the server and impersonates the user. To mitigate that, you could in theory use some time-based salt, like the UTC epoch timestamp or a one-off seed sent from the server.

    While there are ways around it, really, the simplest thing to do is simply use SSL for secure connections. Truth is, the smartest thing to do is use combinations of all of the above, because even SSL isn't guaranteed to be secure if you don't have absolute control over the hardware you are browsing on.

  40. I use NX nomachine as my personal SSL cloud server by ElliotWilcox · · Score: 1

    I submitted an article on using Ubuntu and NX nomachine at home to prevent exactly this type of attack. No one commented http://slashdot.org/submission/1365444/A-personal-private-cloud-server

  41. sidejacking by Maave · · Score: 1

    This technique is called "sidejacking". It works by piggy-backing a user's cookies. I've seen programs and setups to do it before but I wouldn't have imagined that a Firefox addon would be made to do it. This addon is very thorough, a sweet tool for wardrivers.