Please Do Not Change Your Password

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
A new study says that many security measures are a waste of time. Why? I don't know, I guess you’ll have to read the original article (no password or log-in required).

Just when you have them all sorted out, along comes another “urgent” directive from the bank or IT department — time to reset those codes, for safety’s sake. And the latest lineup of log-ins you’ve concocted won’t last for long, either. Some might temporarily stay in your head, others are jotted on scraps of paper and stuffed in a wallet. A few might be taped to your computer monitor in plain view (or are those are from last year’s batch? Who can remember?).
 
Password policies are a double-edged sword. Too bad many IT depts. don't recognize the other edge as much as they should. :(
 
Yeah, but this is well known already. Didn't Bruce Schneier already comment extensively on this thing?
 
i never understood why i needed to change my 14 character password every 3 months. if wasn't hacked in those 3 month jeez i guess its working.

a password that you remeber and never have to write down is more secure then changing it every so often.
 
The government is getting STUPID about this stuff, at least in the dept. I'm in. They just up'd the requirement to 14+ characters, 2 numbers, 2 lowercase, 2 uppercase, and at least 2 special characters. We have to update out passwords every 6 months and it can't be the same as any of our 10 previous passwords. I'm seeing more and more people in my dept. write their PWs down and try to hide them somewhere (and even occasionally on the monitor like the article states). Since I'm the IT/security guy, I have to keep calling them out on it too. It may not be perfectly safe too, but since I have to know literally 30+ passwords for various sites, servers, accounts, etc., I put all mine in a password (hope I don't forget that one! lol) protected Excel 2007 file on my encrypted user profile. I have to add/edit to it prolly a couple of times a week too
arghh.gif
.
 
You guys should use something a little more security-conscious than Excel for that...

“Most security advice simply offers a poor cost-benefit trade-off to users,”
Well, duh. Security always gets in the way. You need to find a balance.
 
eye, finger print and voice scanners would help speed everything up. you would not need all these passwords...but few departments can afford it I am sure.
 
eye, finger print and voice scanners would help speed everything up. you would not need all these passwords...but few departments can afford it I am sure.

Relying on biometrics alone is equally foolish. You still need passwords.

Personally I prefer the password + token approach to dual-factor authentication. It doesn't require special hardware on the machine you're using, and IMO is probably more secure.
 
i never understood why i needed to change my 14 character password every 3 months. if wasn't hacked in those 3 month jeez i guess its working.

a password that you remeber and never have to write down is more secure then changing it every so often.

And how do you know it wasn't hacked in those 3 months?

Hint, any decent IT department isn't too worried about sabatoge. That's what backups are for. Espionage on the other hand... that's where the real damage is. And unlike with sabatoge, there's no glaring neon sign yelling that you've been hacked.
 
And how do you know it wasn't hacked in those 3 months?

Hint, any decent IT department isn't too worried about sabatoge. That's what backups are for. Espionage on the other hand... that's where the real damage is. And unlike with sabatoge, there's no glaring neon sign yelling that you've been hacked.


theirs always what if's
 
The government is getting STUPID about this stuff, at least in the dept. I'm in. They just up'd the requirement to 14+ characters, 2 numbers, 2 lowercase, 2 uppercase, and at least 2 special characters. We have to update out passwords every 6 months and it can't be the same as any of our 10 previous passwords. I'm seeing more and more people in my dept. write their PWs down and try to hide them somewhere (and even occasionally on the monitor like the article states). Since I'm the IT/security guy, I have to keep calling them out on it too. It may not be perfectly safe too, but since I have to know literally 30+ passwords for various sites, servers, accounts, etc., I put all mine in a password (hope I don't forget that one! lol) protected Excel 2007 file on my encrypted user profile. I have to add/edit to it prolly a couple of times a week too
arghh.gif
.

Lol consider yourself lucky... I'm in the military and we've had the whole 14+ characters, 2 numbers, 2 lowercase, 2 upper case, 2 special case, and you can't use any previous words, or passwords used before. No sentances or phrases, no family names and such. Plus a CAC card with a pin identification thats 9 characters long.

Now we have to change them every 3 weeks, 20 characters with 3 of upper case, lower case, special, and numbers. As well as the CAC card with the 9 character pin that has to be changed just as often with differnet combinations and no simple patterns.

So... I am quickly running out of keys to use... :(

This is freaking rediculous, because among the 15 different passwords I have to remember that are like this, I have all my personal passwords to remember as well.

And they wonder why we need 3+ attempts to get into are accounts. Hell it takes me up to 5 trys some times to figure out what password I used for what account and what varitaion of it, it was.

Oh and when you get locked out you have to go to the military police, get a statement from them, get a writen statement from your commander and S-2, and then call a number to set up an appointment, to go verrify it is your account just to unlock it.

And people wonder why IT folks, or people that work on computers go crazy or want to kill things...
 
24 character password with caps, lowercase, numbers and symbols, gl hacking me with a brute forcer or something
 
theirs always what if's

And there's always stupid ways to address them.

However, having a policy to change passwords on a reasonable schedule is hardly a stupid way to mitigate some of the damage from a compromised password. Note I said reasonable. If it was every 3 days, then yeah, I'd be right there with you. But every 3 months? Seems quite reasonable to me.
 
I think it is still good practice to change passwords every so often. The article even states that towards the end. Even though you have to remember that they tried to quantify the results of not practicing good safe security habits, and although they concluded that most folks don't bother because of IT's approach to their employees in training them still don't get them to follow the rules, they still concluded that prevention and best practices are safer than not, even though their data show that even if you are a slouch and don't practice safe habits, the chances are still slim, although they admit that their data is somewhat flawed in stating "slim" - not enough data to support the too many variables with how people actually practice these safe habits.

A balance between functionality/convenience and security is always a tough challenge but should be adhered to in my opinion. Someone mentioned Bruce Schneier and yes, he did talk about that at length. Someone also stated the difference between sabotage and espionage and they are also correct. Schneier once said, amateurs go after machines, professionals go after individuals. To extrapolate on that, I think most attackers are going after your money, individuals money, more than before when it was mostly mass attacks against corporations. But there is also a grave concern from within the corporation, podslurping and data theft.

I like to use http://lastpass.com, as a password manager, and a Yubikey is also a great product. There are also some free open source programs like keepass and locknote that are quite lean and robust.

Philip
 
We change passwords when someone leaves. They don't need to know our passwords and those left can use a password manager and be happy to still have a job. :eek:
 
unfortunately our university forces us to change our passwords every few months, and it can't even be similar to the previous passwords either, I hate it.
 
unfortunately our university forces us to change our passwords every few months, and it can't even be similar to the previous passwords either, I hate it.

This is nothing new. ESPECIALLY for colleges / universities / companies that work with/receive funding from the government.


As a person who works in IT (like 2/3 of the board) I'm personally a fan of passwords changing every 90 days at a minimum. Not allowing to use your last 10, 24, 45 passwords is a bit excessive. Not allowed to use your last 4-5, fine.

What most people don't realize its pretty darn easy to create a secure password. Take for favorite activity, vehicle, person, month, whatever... (Drinking) and swap out some letters for numbers, add a year at the end, then another letter.. Drinking becomes Dr1nk1ng2010a

Next 90 days... change the last letter. As long as no one finds out your particular key word/phrase, the password is pretty safe against bruteforce.
 
Next 90 days... change the last letter. As long as no one finds out your particular key word/phrase, the password is pretty safe against bruteforce.

Doing this pretty much negates any value in changing it every 90 days. If you're going to have such a trivial modification you may as well just leave the password alone.
 
Will biometric authentication every fully replace passwords?

Doubtful. As a single-factor I think passwords are about as good as it gets. If the users use them properly (ie. not divulge them and select good ones), they are secret and essentially impossible to reproduce.

Biometric factors are all basically 'public' information; a coke can contains your prints, a hidden IR camera can record your iris patterns etc. That means it's relatively easy to construct a model that is sufficiently good to fool the biometric sensor. So even in the ideal situation, they aren't secret, and while they might be difficult to reproduce, it would be possible to spend time doing so away from the target machine, which isn't possible with passwords.
 
I'm personally a fan of passwords changing every 90 days at a minimum. Not allowing to use your last 10, 24, 45 passwords is a bit excessive. Not allowed to use your last 4-5, fine.

What most people don't realize its pretty darn easy to create a secure password.

At my job, I was frustrated just about every-other password change as I couldn't find something I was sure I'd remember while being different enough to not have the system trip on it as being to common or like the previous passwords...

Then someone suggested their "method"... And then each month when I have to change everything I never again tripped on their system...

The bizarre thing was that the passwords truly were extremely all like one another, and each new one was also very similar to the one being changed...

I found it odd that this "method" worked while IMO greatly lowering security...

FWIW: When a new guy was having issues getting tripped-up on new passwords being too alike the old one, I told him vaguely about the existence of this "method"... He wanted me to explain so he wouldn't have any issues anymore, but I told him explaining it would just make his and my passwords now even less secure...

And three years later, the guy that originally explained it to me, I saw him typing in his password... And realized he was still using it... When I mentioned it to him, he seemed annoyed that I watched him type his password... I had to explain to him that finger-pecking when entering your password didn't help... I always at least touch-typed mine in...

Simply put: Finger-pecking makes it clearer where your password began and ended... With this "method", if you knew the first and last letter, you basically knew the entire password... If you knew just the first letter, you might guess all of the rest before the account would lock you out from too many failed attempts... As I said, strange the system never caught how alike pattern-wise they all were... and the "method" always worked...

I won't be shocked if at another job I find this "method" doesn't work anymore, though...


As far as personal passwords: I have one that I've used for over 17 years... I actually wasn't even initially allowed to use it (it would also be a verbal password) 'til I explained how I planned to remember it...
 
At my job, I was frustrated just about every-other password change as I couldn't find something I was sure I'd remember while being different enough to not have the system trip on it as being to common or like the previous passwords...

Then someone suggested their "method"... And then each month when I have to change everything I never again tripped on their system...

The bizarre thing was that the passwords truly were extremely all like one another, and each new one was also very similar to the one being changed...

I found it odd that this "method" worked while IMO greatly lowering security...

FWIW: When a new guy was having issues getting tripped-up on new passwords being too alike the old one, I told him vaguely about the existence of this "method"... He wanted me to explain so he wouldn't have any issues anymore, but I told him explaining it would just make his and my passwords now even less secure...

And three years later, the guy that originally explained it to me, I saw him typing in his password... And realized he was still using it... When I mentioned it to him, he seemed annoyed that I watched him type his password... I had to explain to him that finger-pecking when entering your password didn't help... I always at least touch-typed mine in...

Simply put: Finger-pecking makes it clearer where your password began and ended... With this "method", if you knew the first and last letter, you basically knew the entire password... If you knew just the first letter, you might guess all of the rest before the account would lock you out from too many failed attempts... As I said, strange the system never caught how alike pattern-wise they all were... and the "method" always worked...

I won't be shocked if at another job I find this "method" doesn't work anymore, though...


As far as personal passwords: I have one that I've used for over 17 years... I actually wasn't even initially allowed to use it (it would also be a verbal password) 'til I explained how I planned to remember it...

I'm not the brightest code breaker in the world, but I would have to say your method is probably based on pattern. For example, you could decide on the first letter of your password and use a pattern to continue the password. lets say that I wanted the first letter of my password to be the letter "e". I could then make a numerical pattern using the alphabet and determine the next letters. I could say that I am going to increase by even numbers from "e", like +2, +4, +6, +8, etc. So each consecutive letter in this password for this example would be e, g, k, q, y. This is a simple example, but it creates a seemingly random password that you can remember by a pattern method. You could create a more complex pattern that makes the password look even more random. To change your password, you just change the first letter and follow the same pattern method. All you need to know is the first letter and the pre-determined pattern.
 
Doubtful. As a single-factor I think passwords are about as good as it gets. If the users use them properly (ie. not divulge them and select good ones), they are secret and essentially impossible to reproduce.

Biometric factors are all basically 'public' information; a coke can contains your prints, a hidden IR camera can record your iris patterns etc. That means it's relatively easy to construct a model that is sufficiently good to fool the biometric sensor. So even in the ideal situation, they aren't secret, and while they might be difficult to reproduce, it would be possible to spend time doing so away from the target machine, which isn't possible with passwords.
It's just that passwords are so crude and can eventually be defeated w/ a fast enough computer (which partially explains why the min. password length is increased every few years - most authentication systems now require at least 8 characters, some 10). I remember in the 1990s and early 2000s, most only required at least 6 characters.

The main knock of biometric authentication is if someone somehow gets your biometrics (like you mentioned w/ the soda can, etc.), then they can get in as you and you can't change it. Whereas if someone gets your password and you find out about it, you can have it changed.
 
It's just that passwords are so crude and can eventually be defeated w/ a fast enough computer (which partially explains why the min. password length is increased every few years - most authentication systems now require at least 8 characters, some 10). I remember in the 1990s and early 2000s, most only required at least 6 characters.
This only really applies to passwords for e.g. encrypted data files, where an attacker can steal the file and have his way with it. As far as access to network and services (like this article is discussing), running a bruteforce attack against the auth server, or even an individual machine will be quickly noticed and blocked, the account will be locked out etc. It's not really feasible to attack passwords this way.
 
Back
Top