« first day (1211 days earlier)      last day (3678 days later) » 

12:57 AM
hehehe some are fantastic ... not booties, that's just the image but it's NSFW tho
 
 
1 hour later…
2:10 AM
@RоryMcCune And already 6 downvotes !
@tylerl I got today an email from the local sysadmin people: they were taking down a number of systems in emergency because they had to patch for the new scary OpenSSL bug.
It has gone completely viral.
 
Night folks, I think I've gone anemic from all this heartbleed already :|
 
 
2 hours later…
4:31 AM
Seeking devop with 2+ years in Heart Bleed
 
- What should clients do about heartbleed?
- Are clients affected by heartbleed?
- How affected are clients by heartbleed?
- Are clients as severely affected by heartbleed as servers?
- Is SSH affected by heartbleed?
- Is Apache affected by heartbleed?
- Is Tomcat affected by heartbleed?
- Which clients are affected by heartbleed?
- Is Facebook affected by heartbleed?
- Is OpenVPN affected by heartbleed?
- Which services are affected by heartbleed?
- What should the average user do about heartbleed?
- Is heartbleed hertbleed heartbleed?
- How heartbleed is heartbleed actually is?
 
5:34 AM
@ThomasPornin meh. the scarce resource is brain cycles. And if I need to spend even one extra cycle to "parse" the separation of blocks, it's not worth it. I put opening braces on new lines so the block pops out.
 
@TildalWave Oh, thank you for tagging that as NSFW. I couldn't have figured it out from the thumbnail.
 
@TildalWave what the hell is a "vine"?
 
@AviD Instagram for videos.
 
@TerryChia isnt that youtube?
or vimeo?
 
@AviD Ok, Twitter for videos.
 
5:43 AM
and if its a vine, why is it on youtube?
@TerryChia so visually useless.
 
@AviD Short, few second videos basically. It's on YT because it's a compilation.
 
@TerryChia Love, you explained this to me before but I forgot, so I'll have to ask again. My tweets are private (my tweeting wall or whatever it's called is set to private). If I mention somebody, (like asking something from somebody I'm following). Example: "@TerryChia Are you going to xCon this year?"
Will they see it or not?
 
@Adnan I won't see it if I'm not following you.
 
@TerryChia Would that change if I set my tweeting wall to public?
 
@Adnan Yes (I think).
And it's called a twitter feed. :P Not "tweeting wall". That sounds funny.
 
5:46 AM
@TerryChia okay 90% of that is worser thant useless crap. what the hell?
 
@AviD Teenagers.
 
@TerryChia Is your twitter tweeter twit @TerryChia?
 
@TerryChia uhh, yeah okay, that actually explains it.
 
I guess it's called "Twitter username"
Okay, @AviD, I'll try it on you.
 
try what?
heh. I just got tweeted.
twerted?
 
5:49 AM
@AviD Twitted?
 
@Adnan Nah it's not. And I'll prefer to keep it private sorry. :)
@AviD Twerked?
 
@TerryChia Hey, no problem at all.
@AviD So, you saw it?
 
Just a thought, since we determined that heartbleed clients are vulnerable... anyone bothered to recompile Python yet? :P
 
@Adnan yeah. and responded. so is my response private too?
 
BEST TWITTER ACCOUNT EVER!
@SwiftOnSecurity, The Internet
Tayor Swift's thoughts on #infosec. By @explanoit
6 tweets, 84 followers, following 34 users
 
5:52 AM
@AviD Perfect. Now let's remove that abomination. I removed mine
@AviD How did you make your tweeter wide and big like that?
@RоryMcCune's and mine are still small and narrow
2
 
@Adnan I just maximized my window. But how did you see that??
 
@AviD Go to my tweeter and look at my page. It's narrow and weird
 
@Adnan TWSS?
 
@TerryChia But hey, really, was there some opt-in Twitter upgrade?
 
@Adnan Hahahahaha!
 
5:56 AM
@Adnan and mine isnt?
 
@Adnan No clue.
 
@AviD Why does mine look like that? Maaaan, every time I look at it and compare it to yours I feel sad :(
2
 
perhaps because you changed the template etc.
@Adnan starwhoe ;-)
 
@AviD Last one, yup.
@AviD Oooh, how can I change that template again?
 
@Adnan clikc "Me", then click "Edit Profile"
 
6:00 AM
@AviD Yeah? It shows me a place to change my picture
No mention of templates
 
@Adnan sorry, "theme"
dunno, maybe its the header pic.
 
@Adnan Maybe @AviD is just a beautiful and unique snowflake.
 
if I was, I would be water by now. Summer is here.
 
@AviD Don't worry. Winter is coming.
Caught the new GoT episode yet?
@ThomasPornin, @Code, anyone else with good crypto knowledge --- can I get some thoughts on this? github.com/pyca/cryptography/issues/846
 
 
1 hour later…
7:09 AM
ya know what's super odd about twitter
 
@RоryMcCune Do tell
 
I put a picture in the chat transcript of that heartbleed honeypot
and it got tweeted by some guy from the transcript
and now that tweet has been retweeted all over the place
like the most I've seen a tweet I did get tweeted is one/two retweets
this one is more like 50
and all it does is link to that picture in the chat transacript
v.odd
oh yeah and I get an e-mail every time someone retweets it..
 
@RоryMcCune Turn it off. :P
 
@TerryChia yeah it was more that I've never had cause to need to before (used to be interesting to see retweets) and I've no idea why this tweet is getting RT'd so much
 
Twitter gives you pretty fine grained control over email notifications.
@RоryMcCune Funny stuff gets retweeted a lot.
 
7:14 AM
@TerryChia could be that but I bet if I'd RT'd the tweet I got the image from it wouldn't have gone round as much..
I guess that's down to the number of followers the original tweeter has..
 
@RоryMcCune Why do some questions become rep trains? Same theory.
It's completely random.
 
@ThomasPornin ridiculous!
@TerryChia true it's pretty random but I think yesterday mentioning heartbleed was a +500% powerup
I looked at the "winning" answer for that "should I change all my passwords" question, bad idea, now I'm angry about how wrong it is and it's not 8:30 yet
 
@RоryMcCune Link?
 
13
A: Should I change all my passwords due to heartbleed

Thomas W.Short answer: Yes, all passwords. Long answer: At first sight, you only need to change the secret key of the certificate. But due to several reasons, all passwords are affected. Here's why: Reason 1: Chained attack Someone captured the secret key of the certificate. From that time on, he coul...

luckily I'm saved the time of putting together a rebuttal 'cause it just got closed
 
Ugh, TL;DR. Why is it wrong?
 
7:24 AM
@TerryChia hey says to change all passwords 'cause they might have been compromised either directly or 'cause the admin might have logged in to the site and had his password compromised
so every time a site admin might have done something risky you reset all your passwords?
 
Heh.
 
every time an apache bug comes out which is RCE you reset all your passwords
 
Time to start drinking I think?
 
assuming of course that changing all your passwords is a risk free activity
I'll bet that if people to a huge password reset sweep they'll set it to be the same.
on multiple sites
that or they'll forget them quickly
far better advice would be to get 2FA on critical account and start using a password manager
that advice I could get behind
 
Almost 0.4% of ALL questions on Sec.SE are now about heartbleed!! About ONE attack.
Wow.
 
7:31 AM
I'm getting an urge to merge!
All of them
Into one
And then ban heartbleed questions forever more
....
 
100% @Terry would masturbate to that.
 
@Adnan Wut?
 
@TerryChia You know, because you've requested a ban on a few times.
 
@RoryAlsop it'll run and run this one, I'm started to get e-mails from services I use about it
and that dumb story in the BBC about changing all your passwords <sigh>
 
@Adnan I only asked the mods to nuke the tag. :P
 
7:34 AM
@TerryChia and that would be fantastic
 
rassin' frassin' university professors
all websites should tell you they've updated their SSL
should they tell you every MS patch tuesday too
and about the fact they use PHPMyAdmin
 
@RоryMcCune Are you sure @TildalWave did not take over your account?
 
Honestly, this is the most obnoxious security incident I've witnessed in my adult life.
 
I woke up grumpy
people were banging around and shouting as I woke up (my in-laws are here and getting ready to head off)
 
@RоryMcCune You need to step away from the Internet then. :)
 
7:40 AM
but really it's cause this is really annoying, first security thing in the news for a while, and the advice is all wrong...
 
@RоryMcCune Really? Where were you when the Snowden stuff happened?
 
@RоryMcCune In-laws asking about hearbleed?
 
instead of taking a chance to talk about defence in depth or perhaps why corporates rely on open source software but don't contribute to it's security, it's things like "change all your passwords"
@Adnan they asked about changing all their passwords after that BBC news story
 
@RоryMcCune I was thinking my father was the only ******* one
 
@TerryChia that's not really IT Sec though (apart from the bit where the most secure organisation on the planet let a contractor take unlimited data from them and couldn't even say what he took)
 
7:42 AM
Got a call from him this morning telling me that all the Internet has been hacked by the Russians and CIA
 
@Adnan hell no after the news coverage in the UK, I'll bet a v.large number of IT/ITSec people will get those questions today
 
@RоryMcCune Dual EC DRBG? RSA?
 
hey @RoryAlsop how many people have you had ask so far
@TerryChia sure you'd think that would be bigger, malicious planting of weaknesses in crypto.... did it even get any mainstream press
noooo
@TerryChia you're just trying to make me grumpy now aren't you..
 
@RоryMcCune Hehe.
 
7:55 AM
I'm writing an article about heartbleed, maybe I should post it on Sec.Se?
as it seems that we have mass panic here where people start claiming it's a virus
and that all security is compromised on the intertubes
 
Only if you manage to write it without mentioning the H* word. I developed an allergy to that.
 
@CodesInChaos H?
 
@CodesInChaos you're allergic to hemorrhoids ?
 
@CodesInChaos I'm trying to asses the actual impact rather than fud
@CodesInChaos heartbleed? I've used it once
 
8:02 AM
the actual impact is hard to assess, since we're all uncertain as to how long this has been exploited in the wild for.
 
@FalconMomot that's the point
 
Is it correct the dodgy code has existed for 2 years?
 
@FalconMomot but you are only impacted if you actually used openSSL and while a lot of services use it, that doesn't even cover half the internet
@FalconMomot @rorym @CodesInChaos if you want I can let you proof read
 
@deed02392 Yup, ever since the implementation of heartbeat into OpenSSL
 
@LucasKauffman there is also this neat thing called protected mode where you can't access the memory of random processes from one process, so the compromise is limited to data the process running openssl can work with I think.
 
8:05 AM
@FalconMomot you mean private SSL keys :p
 
@Adnan Jesus
 
@LucasKauffman I'm just about ready to go to bed, but please do link :)
 
Also, I hate to be that guy but, NSA?
 
@LucasKauffman well yeah, among other things
 
8:05 AM
@deed02392 I'd imagine they might be more subtle.
 
@FalconMomot I'm still working on it, if not tonight have a look at it tomorrow
 
@FalconMomot Two years to find it
 
@deed02392 but it's hardly subtle at all.
 
@FalconMomot How do you define subtle if not that it took so long to discover? How much more subtle could you be for such an effective attack?
 
@deed02392 you'd try to target, a little bit at least, so that the whole world doesn't explode when it's discovered. ;)
 
8:08 AM
I would agree with you that it's not subtle at all but that's easy to say now we know the exploit. I am amazed it has been in the wild for 2 years
 
@deed02392 that it was speaks mostly to the insane complexity of TLS (among other things) and the fact that detailed knowledge about it is perhaps more limited than it should be.
 
nice @Adnan
 
RoryM - I'm delighted in on holiday. Might write up a canned response tonight before I go back in to the office tomorrow
 
See also: BEAST
 
8:10 AM
We have fifty fucking h****b**** questions. Seriously.
 
@FalconMomot Which no doubt they'd be aware of and so yes it would be considered subtle enough, especially with the lack of spotlight they enjoyed 2 years ago
 
Actually, does anyone want to write the definitive blog post on this?
 
43 mins ago, by Adnan
Almost 0.4% of ALL questions on Sec.SE are now about heartbleed!! About ONE attack.
 
s/attack/vuln/
 
I will be at home later this afternoon and will aim to cleanup the bleeding heart crap. Useful flags will help me
 
8:11 AM
@RoryAlsop Oh good lord! Thank you! Finally
 
@RoryAlsop I'm doing one atm, want less fud, more facts
 
Hey @CodesInChaos, mind weighing in on this? github.com/pyca/cryptography/issues/846
 
@TerryChia There was a related discussion in the context of NSS: bugzilla.mozilla.org/show_bug.cgi?id=868948
 
@CodesInChaos Ah thanks, I'll read through it.
 
Constant time GCM implementations without AES-NI costs quite a bit of performance compared to variable time GCM.
 
8:15 AM
@LucasKauffman sure
@LucasKauffman if you're looking for conclusions, can I suggest you add a bit about potential useful outcomes of this, something about "should perhaps companies look at the open source software they rely on for their billion dollar enterprises and consider contributing to security reviews of them"
6
 
Ha @Adnan did you know AMD released an APU with an architecture codename of 'Kaveri'
 
@RоryMcCune excellent idea
 
@RоryMcCune starred
 
@deed02392 Now way! Does it mean something else is some other language?
 
@Adnan Haha I don't know, but kaveri suits it if it's supposed to act as some kind of co-processor
There's not even a wiki article on it, though
 
8:22 AM
Well, I took a look at the list
Desna
Ontario
Zacate
Llano
Hondo
Trinity
Weatherford
Richland
Kaveri
Kabini
Temash
IGP
Wrestler
WinterPark
BeaverCreek
All of them, including Kaveri, seem to be river names
 
Ahh I should have looked at them in context
I only saw it because it appeared in a list of AMD processors that support AES-NI
 
@deed02392 Would have been more interesting if they actually meant the Finnish word.
 
And I'm pretty sure Bulldozer, Piledriver and Jaguar aren't rivers :P
Advanced Encryption Standard Instruction Set (or the Intel Advanced Encryption Standard New Instructions; AES-NI) is an extension to the x86 instruction set architecture for microprocessors from Intel and AMD proposed by Intel in March 2008. The purpose of the instruction set is to improve the speed of applications performing encryption and decryption using the Advanced Encryption Standard (AES). New instructions {| class="wikitable" |- ! Instruction ! Description |- | AESENC | Perform one round of an AES encryption flow |- | AESENCLAST | Perform the last round of an AES encryption flow |-...
 
:D
 
How did your experimenting go, did you find any private keys in the end?
 
8:35 AM
@RоryMcCune it's kind of sad that this isn't standard procedure, especially for something so obviously security-critical. Without an audit you're paying for, the code is effectively completely untrusted.
 
@FalconMomot It's very sad and really quite shameful of companies making extensive use of it
openssl.org could do with a makeover and some way of promoting sponsors or something
perhaps they need to look more like a business and not a collective of geeks that companies think can't be wrong and can be exploited for free functionality
 
hmm
on the other hand, piracy in enterprise environments is hardly uncommon
and when you have sub-week release cycles it gets difficult to do detailed audits of things
#devopsconsideredharmful
hey maybe for my master's thesis instead of doing it on entropy analysis I should do it on why "devops" is a monstrosity and needs a massive paradigm shift to not be a liability.
 
@FalconMomot it's extremely sad, especially when you look at the roster of companies using it
 
@RоryMcCune also governments
 
@FalconMomot yep, and you can bet that after this bug, a load of offensive security types will be poring over the codebase looking for more problems...
 
8:42 AM
the Canada Revenue Agency got hit with it, all of 21 days before the filing deadline
@RоryMcCune I'm surprised it's taken them this long. Those vulns are at this point as good as (or better than) vulns in a TCP/IP stack.
undoubtedly more abundant too.
 
@FalconMomot well I'd bet some of the top-end types will have had this + others in the bag for a while, but the range will widen out a lot now I'm sure
 
I feel almost certain this has been being exploited in the wild for a long while before we discovered it
 
it does have a pretty strong version dependency though
 
Yeah, ironically the later versions which the larger more sensitive corporations are most likely to have 'upgraded' to
I bet the assholes still with exploitable NTP servers aren't susceptible
 
@TerryChia no no, winter is come and gone.
and - yeah. GoT is great.
@RоryMcCune the fact that it's so popular?
 
8:50 AM
I lol'd
I was really tired when I watched the new GoT, I feel like I need to watch it and maybe the last half of season 3 again
 
@AviD yeah, an image from the chat transcript gets RT'd like 50 times, no information useful or otherwise just a picture
 
Is there a standard in development for an Internet of Things protocol? E.g. a wifi kettle and a Wifi enabled power socket using a compatible protocol?
 
@deed02392 there was some standard for home automation that likely covers it: X10.
though X10 with 802.11 would be silly
and I guess it's not really much of an internet of things protocol at all
 
Mmm that's a lower level protocol than I'm thinking of
 
I think you missed the point of that xkcd... namely the dripping sarcasm. Your answer is also a bit blown out of proportion, but that at least is in tune with most of the mass media's treatment of this. — AviD ♦ 20 secs ago
 
8:58 AM
@AviD heh
 
@RоryMcCune also - what web site allows an admin to log on and directly access the database??
 
@AviD badly designed ones :)
 
no, I dont even think that.
 
@AviD you've never seen phpmyadmin exposed to the Internet?
 
I have actually seen a couple of business apps like that, but those are features. not in common web apps...
 
8:59 AM
if so you've not seen enough low-end web app reviews..
 
@RоryMcCune and that gives direct access to the db?
 
not allowing admins access to the database seems a little like view-based access control to me?
 
@AviD yeah phpmyadmin is direct database manipulation
and it's had loads of security bugs
 
oh hmm. then again, you're using PHP, so none of these other issues is your biggest problem.
 
I've seen it quite a bit in the past, just sitting there somewhere predictable like /phpmyadmin
@AviD you dissing facebook again.
 
9:00 AM
hehe
 
@JeffFerland won't be happy
 
@FalconMomot no, it's indirection and encapsulation. nothing to do with views.
 
it just doesn't feel like the right place to enforce access constraints.
though making a web form that forwards totally arbitrary stuff to the database does seem to serve no purpose besides making attackers have an easy go of it.
and of course I didn't think of the case right away where the application "admin" might not have access to the backing database at all, but hey.
the application has creds on the database, at some point, and probably also has them in memory.
 
@Xander You should answer this security.stackexchange.com/questions/55351/…
 
@FalconMomot no, its not about "enforcing access constraints", its about not even creating such a channel.
 
9:13 AM
@AviD there has to be at least some channel. it's just a matter of defining what can go over the control channel, but that has to be at least everything the application can do.
 
an indirect one.
there is a world of difference in allowing the application to send specific commands, as opposed to allowing an admin to access the database directly.
it's like the difference between an API and an open file share.
 
@deed02392 Oh, @Kisuminttu wanted me to tell you that there are companies like Roccat use a lot of Finnish words for their product names: Apuri, Syva, Isku, Kova, Savu, etc.
 
ROCCAT Kone XTD
haha
Is Roccat a Finnish company? Seems odd considering it has two 'c's in the name
 
@deed02392 Yup, that's @Kisuminttu's mouse.
@deed02392 German
 
isn't that odd
I know of quite a lot of Finnish companies, I wonder how many average people realise how significant Finland really is
Huhtamaki, Nokia, Rovio (Angry Birds), Kone (escalators/lifts all over the world)
 
9:19 AM
@Adnan Actually, I have the original version of that mouse... IMHO the "upgrades" to that original model haven't contributed any good features : )
 
@Kisunminttu Where did you find a computer to login?
 
I just started looking at the names and apparently they've expanded to some other language as well...
@Adnan In the room where I work. : )
Most of my work is done on the computer
 
@Kisunminttu I feel like I don't know you.
@Kisunminttu One of ROCCAT's mouse pads is called Siru.. hehehehe
 
@Adnan What do you mean?
Yeah I noticed that. Höhöhö : P
 
@deed02392 Whoa!! You can list Finnish companies that aren't Nokia or Rovio!
You're a fine one, young man.
 
9:24 AM
\o/ You just start to notice it very quickly when you're familiar with the language
 
Hmmm... I used to have a keyboard from Roccat and it got awful reviews from people. I hated that thing and destroyed it in the end. I was just trying to find out what the model was but they've apparently stopped making it. Hehehe.
 
@Kisunminttu I've got a microsoft keyboard that is God-awful, the keys are extremely stiff
I took it apart and washed it and it's still crap. I think the holes for the key pegs are too small
I need to find a suitable lubricant
 
@deed02392 Heeeey! What's with you going around asking my girlfriend about "suitable lubricants"?
5
:D
 
@deed02392 I bet it's not half as horrid as the Roccat one was. Every time you used caps lock -> instant freeze, need to reboot. Trying to get the multimedia keys to work? Wishful thinking. Trying to write anything on it was impossible, it skipped almost 30% of the key presses. The only thing that worked on that thing was changing profiles... As a keyboard, it sucked.
I should have "lubricated" it with gasoline and "polished" it with a match.
2
 
I still need to get a new keyboard.
 
9:28 AM
Oh right, I forgot you two are in a relationship
@Kisunminttu lol
It's so frustrating because my old one wasn't half bad
 
@TerryChia Anything mechanical is awesome, but hey, I'm biased.
2
 
@Kisunminttu Yeah, I want a good one though.
 
@Adnan :*
 
It isn't fair. I might see if MS will send me a new one
 
Gotta run, break is over.
 
9:29 AM
@TerryChia She can help you with that. She's like the expert on these matters. At least compared to me.
@Kisunminttu Love, when you're back, recommend some models for @Terry
 
@Adnan Hehehe. I'm the hardware know-how part of our relationship.
And yep, I sure will!
 
@Kisunminttu I don't like the sound of that.
:p
 
@Kisunminttu ahem
 
@Adnan Oh. that.
 
@Kisunminttu Fly now.. fly! Back to work. :*
 
9:31 AM
@Adnan Enjoy your day, love : )
 
n'awww söpö
 
@AviD Non-applet Java programs have full permissions anyway (just like native programs – the JVM is a native program) so I don't know what a security flaw there would look like. — ntoskrnl 20 hours ago
can somebody please explain to me what he meant?
 
9:50 AM
Ahhh.. now I have to read the question, that answer, your comments, aaand his comments.
 
@AviD I think he's just an idiot.
 
@Adnan best value for link.
@TerryChia I dont think so, at least not usually.
 
Applet sandbox escape vulns don't apply to Android etc but other JVM vulns sure do.
 
@TerryChia right, exactly - but noone mentioned applets or sandbox until he did.
 
@AviD Hence "he's just an idiot".
 
9:52 AM
@TerryChia in this case, its possible.
 
A code execution vuln in the JVM can be pretty nasty.
 
@AviD Ooooh... now I know what he means.
 
@Adnan because he said its correct?
 
Ok, I can see his point WRT to Android.
Because Android does not use the JVM runtime.
But I still don't understand why he is even bringing Applets into the conversation.
 
@TerryChia no?
 
9:54 AM
@AviD Android uses the Dalvik runtime.
 
He's stating that native (non-applet) apps run and (almost) interact directly with the machine, then there's no difference between malicious code and vulnerable JVM.
 
ah fine, fair enough.
 
Not the regular JVM you run on your desktops/servers etc.
 
@TerryChia its still a jvm though.
well, technically not.
because it doesnt run java directly....
 
@AviD It's not.
 
9:55 AM
@AviD The error in his logic is very clear when you have a Java server running on a JVM susceptible to some RCE vuln.
 
@Adnan still not getting what that means.
 
It doesn't support a bunch of languages that compiles to the JVM for example.
 
fine, but all that is irrelevant - he was making claims about the JVM.
 
@AviD Yeah, hence "he's just an idiot".
That's like saying anything that doesn't run in a sandbox will not have security flaws.
 
@AviD He thinks that because an application is native to the machine, then malicious code (in the native application) is equally or more sever than a vulnerable JVM.
He's wrong, of course.
 
9:59 AM
@Adnan and he's also right.
but thats not what he was saying.
 
@AviD In the comment you linked, that's exactly what he's saying.
 

« first day (1211 days earlier)      last day (3678 days later) »