Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Apr 20, 2014, at 20:20, Joe User mailingli...@rootservice.org wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA512 On 20.04.2014 18:40, Matti Nykyri wrote: On Apr 20, 2014, at 15:38, Mick michaelkintz...@gmail.com wrote: On Sunday 20 Apr 2014 10:10:42 Dale wrote: Just a 1/3 of all websites offer TLSv1.2 at the moment and hardly any public sites offer it as an exclusive encryption protocol, because they would lock out most of their visitors. This is because most browsers do not yet support it. MSWindows 8.1 MSIE 11 now offers TLSv1.2 by default and has dropped the RC4 cipher (since November last year). I understand they are planning to drop SHA-1 next Christmas and have already dropped MD5 because of the Flame malware. This should push many websites to sort out their encryption and SSL certificates and move away from using RC4 and SHA1 or MD5. As I said RC4 has been reverted to by many sites as an immediate if interim defence against the infamous BEAST and Lucky Thirteen attacks. This is a problem all Microsoft's customers are facing. Take a look on Linux Distros from 2000 when WinXP has been developed, and you'll see, that the Linux Distros weren't better in this. Same for the time when WinVista was developed, and the same for Win7 and Win8. So don't blame Microsoft for things that they did as good as everybody else did, that would be unfair. Ok, that's a good point. Sorry, didn't really think about it that way. It's mostly a user issue for not updating their software. But still the point is correct that the ones that are suffering of this are their customers, although its not Microsoft's fault. But the number of people using a Linux Distro from the year 2000 is neglible... And of course there are many reasons for that. But what is something to blame Microsoft for is the order of preference that MSIE selects it's cipher. I don't know if user can change this order but i think it would be better to order them by security and not by some other factor ei speed. But thats just my oppinion and I usually try to stay away from windows :) Anyways I just wonder who trusts software whose source code isn't open and and reviewed by a large community that don't have a financial interest on you. Ouch, wrong argument, realy! Nobody in the large opensource community had ever reviewed the heartbeat code in more than two years. This was not a harmless bug in a mostly unused library, it was a realy big issue in one of the most used library in the world and *nobody* saw it. Has openssl ever been carefully audited? I don't think so and i bet that there are more heartbleed like bugs in openssl. Yes heartbleed was solely a bug in openssl and yes it was truely severe and that should never ever be allowed to happen. On the other hand schannel (the Windows cryptolib) is regularly audited. Sorry, but the large opensource community is blind on both eyes, whereas the closed source community is only blind on one eye. But I still disagree... Everybody has some goals why they are doing something.. Some of these goals might be private and some are public. The public and private goal need not to correlate. For any PLC their true goal is to make money for their stock holders. People are by nature greedy and put their own interests above everybody-else's. I think there are less of these greedy people within the open-source community than in general. How can you say that nobody is auditing the security of open-source software? We audit all the software and hardware we use! And every company should. Open-source is just easier coz you have the source to look at. Hardware is the trickiest one to audit of-course. Big agencies have capital to put their people to work in the closed source companies and try inject their goals to the code. It is even harder if you inject the vulnerability to hardware as claimed by Snowden. If you look at Linux kernel I think that is a quite good example on how software should be developed. The update cycle is fast and the few bugs that are found get fixed rapidly. And better the program is written the easier it is to debug and avoid security disasters. Just be reviewing a file you can see how well it is organized and that tells you about the quality of the program. All these things are mostly opinions and speculation because all the data has not been disclosed. Snowden revealed it to some extent but with that content you can analyze the hole extent of operations. What would you do if there were no limits? -- -Matti
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Saturday 19 Apr 2014 19:41:02 Dale wrote: Mick wrote: and look for this info: New, TLSv1/SSLv3, Cipher is RC4-SHA Server public key is 2048 bit Secure Renegotiation IS NOT supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-SHA I have this little padlock looking thing too. I dug around and found this info: CN = VeriSign Class 3 Extended Validation SSL SGC CA OU = Terms of use at https://www.verisign.com/rpa (c)06 OU = VeriSign Trust Network O = VeriSign, Inc. C = US PKCS #1 RSA Encryption There is another place with info but it doesn't allow me to highlight it so that I can copy and paste. Hmm. Anyway, is that reasonable for a bank to use? In case you haven't noticed, I'm not a wealth of info on encryption, just rich in questions. I just know that it is supposed to make things unreadable without a password, pass key or whatever. This is currently my bank. http://cadencebank.com/ Since they changed to a card that a lot of stores don't take, that could be changing real soon. You need to go to the URL that they provide for secure banking, not the home page of their main website. They seem to offer a lot of services under different URLs. Not all of them have the same level of protection. Picking two URLs at random: The Fluent account login page takes me to: https://portal.cadencebank.com/consumer/ and openssl s_client tells me: == New, TLSv1/SSLv3, Cipher is AES128-SHA Server public key is 2048 bit Secure Renegotiation IS supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: AES128-SHA == So, they use TLSv1, as opposed to the latest TLSv1.2 and their digital signature is with the AES symmetric cipher with 128bit keys. This is considered safe enough for today. They also use the SHA1 hash which is less secure (if you are paranoid that someone may change the packets payload in flight). Since 2004 it was found that practical collision attacks could be launched on MD5, SHA-1, and other hash algorithms and NIST has launched a competition for the next secure hash SHA3. However, MD5 and SHA1 are used so widely today it could take a lng time for them to disappear. However, picking up another banking service of theirs I see that they are using RC4 with MD5: == New, TLSv1/SSLv3, Cipher is RC4-MD5 Server public key is 2048 bit Secure Renegotiation IS supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-MD5 == RC4 is considered completely broken today, even for Microsoft! :-) http://en.wikipedia.org/wiki/RC4 The good news are that your bank's servers do not leak any secrets at this moment and it seems they never did (they use SUN servers). -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Sunday 20 Apr 2014 01:18:43 Peter Humphrey wrote: On Saturday 19 Apr 2014 18:43:50 Matti Nykyri wrote: Well you can use ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhide Res ults=on Well, that's an eye-opener and no mistake. I see my bank is rated B overall. Could be worse I suppose. Maybe I should forward the results to them. Many banks, businesses and public institutions have to cater for the lowest common denominator, or their help lines would be inundated with irate customers being asked to first reboot their MSWindows PC. Until the beginning of April 2014 this would have been a WinXP user with MSIE 8.0. In Europe up to 25% of all PCs are still on WinXP. This counts out anything exotic in encryption capabilities, like ECDHE and ECDSA, because it is only the latest versions of Firefox and Chrome that can use these. This is the reason that banks also employ some other means of authentication, in addition to your user ID; e.g. they typically ask you to enter a few characters out of your password (different each time), or additional secret data like the name of your favourite teacher, mother's maiden name and the like. Unless someone was recording each and every login of yours with the bank and kept a record of each and every password character you ever typed they may still not be able to login, without locking up the account and triggering an offline replacement of your password. So I suspect they assume that the Internet connection to their servers should be treated as aheam! less than private and have deployed additional means of at least stopping unauthorised transactions online. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: On Saturday 19 Apr 2014 19:41:02 Dale wrote: Mick wrote: and look for this info: New, TLSv1/SSLv3, Cipher is RC4-SHA Server public key is 2048 bit Secure Renegotiation IS NOT supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-SHA I have this little padlock looking thing too. I dug around and found this info: CN = VeriSign Class 3 Extended Validation SSL SGC CA OU = Terms of use at https://www.verisign.com/rpa (c)06 OU = VeriSign Trust Network O = VeriSign, Inc. C = US PKCS #1 RSA Encryption There is another place with info but it doesn't allow me to highlight it so that I can copy and paste. Hmm. Anyway, is that reasonable for a bank to use? In case you haven't noticed, I'm not a wealth of info on encryption, just rich in questions. I just know that it is supposed to make things unreadable without a password, pass key or whatever. This is currently my bank. http://cadencebank.com/ Since they changed to a card that a lot of stores don't take, that could be changing real soon. You need to go to the URL that they provide for secure banking, not the home page of their main website. They seem to offer a lot of services under different URLs. Not all of them have the same level of protection. Picking two URLs at random: The Fluent account login page takes me to: https://portal.cadencebank.com/consumer/ and openssl s_client tells me: == New, TLSv1/SSLv3, Cipher is AES128-SHA Server public key is 2048 bit Secure Renegotiation IS supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: AES128-SHA == So, they use TLSv1, as opposed to the latest TLSv1.2 and their digital signature is with the AES symmetric cipher with 128bit keys. This is considered safe enough for today. They also use the SHA1 hash which is less secure (if you are paranoid that someone may change the packets payload in flight). Since 2004 it was found that practical collision attacks could be launched on MD5, SHA-1, and other hash algorithms and NIST has launched a competition for the next secure hash SHA3. However, MD5 and SHA1 are used so widely today it could take a lng time for them to disappear. However, picking up another banking service of theirs I see that they are using RC4 with MD5: == New, TLSv1/SSLv3, Cipher is RC4-MD5 Server public key is 2048 bit Secure Renegotiation IS supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-MD5 == RC4 is considered completely broken today, even for Microsoft! :-) http://en.wikipedia.org/wiki/RC4 The good news are that your bank's servers do not leak any secrets at this moment and it seems they never did (they use SUN servers). Yet. I would rather not be the next customer to have his ID stolen like Target, I think the chain Micheal's was stolen in the past couple days but not positive on that yet. That bank is not a small bank and I pay fees each month for them to be able to keep their stuff updated. If they can't be bothered to keep it updated and then turn around and give me a card that sucks, well, oh well. picture a thumbs up here Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Apr 20, 2014, at 11:49, Mick michaelkintz...@gmail.com wrote: On Sunday 20 Apr 2014 01:18:43 Peter Humphrey wrote: On Saturday 19 Apr 2014 18:43:50 Matti Nykyri wrote: Well you can use ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhide Res ults=on Well, that's an eye-opener and no mistake. I see my bank is rated B overall. Could be worse I suppose. Maybe I should forward the results to them. Many banks, businesses and public institutions have to cater for the lowest common denominator, or their help lines would be inundated with irate customers being asked to first reboot their MSWindows PC. Until the beginning of April 2014 this would have been a WinXP user with MSIE 8.0. In Europe up to 25% of all PCs are still on WinXP. This counts out anything exotic in encryption capabilities, like ECDHE and ECDSA, because it is only the latest versions of Firefox and Chrome that can use these. Yes, this is true. Even gentoo doesn't have a stable firefox that supports TLSv1.2 highest security ciphers C030 and C02C (ECDHE-RSA/ECDSA-AES256-GMC-SHA384). But wht banks should do they should support the most secure ciphers and sort their ciphers lists so that the most secure are at the top. Because what I understood is that browsers will by default use the first cipher in the order the server sent them it supports and not go through the entire list. A security aware user can ofcourse disable all the bad ciphers he foesn't want to use in his own browser. Now if he tries to connect to a poorly secured site the connection will fail until a common cipher is found. But what is important you will know when you try to make an insecure connection. This is the reason that banks also employ some other means of authentication, in addition to your user ID; e.g. they typically ask you to enter a few characters out of your password (different each time), or additional secret data like the name of your favourite teacher, mother's maiden name and the like. Unless someone was recording each and every login of yours with the bank and kept a record of each and every password character you ever typed they may still not be able to login, without locking up the account and triggering an offline replacement of your password. NSA has this capability. Also i think most of the largest ISPs are capable to do it. All this requires is enough HD space, private key of any CA enabled x509 certificate and access to any router between you and the bank or DNS poisoning of your computer. So I suspect they assume that the Internet connection to their servers should be treated as aheam! less than private and have deployed additional means of at least stopping unauthorised transactions online. -- -Matti
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Sunday 20 Apr 2014 10:21:08 Matti Nykyri wrote: On Apr 20, 2014, at 11:49, Mick michaelkintz...@gmail.com wrote: On Sunday 20 Apr 2014 01:18:43 Peter Humphrey wrote: On Saturday 19 Apr 2014 18:43:50 Matti Nykyri wrote: Well you can use ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhi de Res ults=on Well, that's an eye-opener and no mistake. I see my bank is rated B overall. Could be worse I suppose. Maybe I should forward the results to them. Many banks, businesses and public institutions have to cater for the lowest common denominator, or their help lines would be inundated with irate customers being asked to first reboot their MSWindows PC. Until the beginning of April 2014 this would have been a WinXP user with MSIE 8.0. In Europe up to 25% of all PCs are still on WinXP. This counts out anything exotic in encryption capabilities, like ECDHE and ECDSA, because it is only the latest versions of Firefox and Chrome that can use these. Yes, this is true. Even gentoo doesn't have a stable firefox that supports TLSv1.2 highest security ciphers C030 and C02C (ECDHE-RSA/ECDSA-AES256-GMC-SHA384). But wht banks should do they should support the most secure ciphers and sort their ciphers lists so that the most secure are at the top. Because what I understood is that browsers will by default use the first cipher in the order the server sent them it supports and not go through the entire list. I think the browsers go through the list, but agree to support the first server preferred cipher that is also supported by the client, even if it is lower in the list of preferred ciphers on the client: http://news.netcraft.com/archives/2013/06/25/ssl-intercepted-today-decrypted-tomorrow.html A security aware user can ofcourse disable all the bad ciphers he foesn't want to use in his own browser. Now if he tries to connect to a poorly secured site the connection will fail until a common cipher is found. But what is important you will know when you try to make an insecure connection. This is the reason that banks also employ some other means of authentication, in addition to your user ID; e.g. they typically ask you to enter a few characters out of your password (different each time), or additional secret data like the name of your favourite teacher, mother's maiden name and the like. Unless someone was recording each and every login of yours with the bank and kept a record of each and every password character you ever typed they may still not be able to login, without locking up the account and triggering an offline replacement of your password. NSA has this capability. Also i think most of the largest ISPs are capable to do it. All this requires is enough HD space, private key of any CA enabled x509 certificate and access to any router between you and the bank or DNS poisoning of your computer. In Europe I think that the situation for ISPs capturing data is not settled yet. I seem to recall that Germany and Belgium disputed in court a European Directive (Data Retention Directive 2006) to capture and store users data. I think that they eventually were forced to implement part of the directive - who needs GDR's STASI these days! :p In the UK data is kept for 1-2 years, but that is only what they let us know. A few days ago the EU Court of Justice declared the directive invalid/unlawful, but that has been kept quiet in the media. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Sunday 20 Apr 2014 10:10:42 Dale wrote: Mick wrote: SSL-Session: Protocol : TLSv1 Cipher: RC4-MD5 == RC4 is considered completely broken today, even for Microsoft! :-) http://en.wikipedia.org/wiki/RC4 The good news are that your bank's servers do not leak any secrets at this moment and it seems they never did (they use SUN servers). Yet. I would rather not be the next customer to have his ID stolen like Target, I think the chain Micheal's was stolen in the past couple days but not positive on that yet. That bank is not a small bank and I pay fees each month for them to be able to keep their stuff updated. If they can't be bothered to keep it updated and then turn around and give me a card that sucks, well, oh well. picture a thumbs up here Just a 1/3 of all websites offer TLSv1.2 at the moment and hardly any public sites offer it as an exclusive encryption protocol, because they would lock out most of their visitors. This is because most browsers do not yet support it. MSWindows 8.1 MSIE 11 now offers TLSv1.2 by default and has dropped the RC4 cipher (since November last year). I understand they are planning to drop SHA-1 next Christmas and have already dropped MD5 because of the Flame malware. This should push many websites to sort out their encryption and SSL certificates and move away from using RC4 and SHA1 or MD5. As I said RC4 has been reverted to by many sites as an immediate if interim defence against the infamous BEAST and Lucky Thirteen attacks. According to the Netcraft SSL Survey (May 2013) only a third of all web servers out there offer Perfect Forward Secrecy to ensure that even if the encryption keys were to be compromised, previous communications cannot be retrospectively decrypted. Elliptic Curve algorithms are not yet included in many browsers and in any case the security of these in a post-Snowden world should be questionable (well, at least the arbitrarily specified NIST-NSA sponsored curves, which OpenSSL is heavily impregnated with). What I'm saying is that there may be no perfect banking website out there, because Internet security is screwed up at the moment, but it is always worth looking for a better bet. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Apr 20, 2014, at 15:38, Mick michaelkintz...@gmail.com wrote: On Sunday 20 Apr 2014 10:10:42 Dale wrote: Just a 1/3 of all websites offer TLSv1.2 at the moment and hardly any public sites offer it as an exclusive encryption protocol, because they would lock out most of their visitors. This is because most browsers do not yet support it. MSWindows 8.1 MSIE 11 now offers TLSv1.2 by default and has dropped the RC4 cipher (since November last year). I understand they are planning to drop SHA-1 next Christmas and have already dropped MD5 because of the Flame malware. This should push many websites to sort out their encryption and SSL certificates and move away from using RC4 and SHA1 or MD5. As I said RC4 has been reverted to by many sites as an immediate if interim defence against the infamous BEAST and Lucky Thirteen attacks. This is a problem all Microsoft's customers are facing. I wonder why they don't demand more. I hope this publicity that snowden and heartbleed has brought to an average user will change their interests to demand better privacy. Anyways I just wonder who trusts software whose source code isn't open and and reviewed by a large community that don't have a financial interest on you. According to the Netcraft SSL Survey (May 2013) only a third of all web servers out there offer Perfect Forward Secrecy to ensure that even if the encryption keys were to be compromised, previous communications cannot be retrospectively decrypted. Elliptic Curve algorithms are not yet included in many browsers and in any case the security of these in a post-Snowden world should be questionable (well, at least the arbitrarily specified NIST-NSA sponsored curves, which OpenSSL is heavily impregnated with). What I'm saying is that there may be no perfect banking website out there, because Internet security is screwed up at the moment, but it is always worth looking for a better bet. It is really hard to fight for privacy, because we have large companies and agencies that actively are lobbing politicians and standards for their own personal interests. In order for the security to get better an average user need to gain an interest to it. This seems unlikely because now a days everybody is uploading all their secrets to a cloud computing service etc. But I hope this publicity will change it even slowly. Another thing is that system administrators need to gain more knowledge on securing their services. For that I think this conversation is quite helpful. A lot of people read this list and it can be found by google. Openssl and gnupg are not very easy to use for someone who doesn't have any knowledge on cryptography. For example openssl will try to use md5 by default even in gentoo if you just try to create x509 cert. And many manual pages are way behind... Newest algorithms are almost never listed there. So you have to truly dig in or ask somebody to find safe and up-to date answers. -- -Matti
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
-BEGIN PGP SIGNED MESSAGE- Hash: SHA512 On 20.04.2014 18:40, Matti Nykyri wrote: On Apr 20, 2014, at 15:38, Mick michaelkintz...@gmail.com wrote: On Sunday 20 Apr 2014 10:10:42 Dale wrote: Just a 1/3 of all websites offer TLSv1.2 at the moment and hardly any public sites offer it as an exclusive encryption protocol, because they would lock out most of their visitors. This is because most browsers do not yet support it. MSWindows 8.1 MSIE 11 now offers TLSv1.2 by default and has dropped the RC4 cipher (since November last year). I understand they are planning to drop SHA-1 next Christmas and have already dropped MD5 because of the Flame malware. This should push many websites to sort out their encryption and SSL certificates and move away from using RC4 and SHA1 or MD5. As I said RC4 has been reverted to by many sites as an immediate if interim defence against the infamous BEAST and Lucky Thirteen attacks. This is a problem all Microsoft's customers are facing. Take a look on Linux Distros from 2000 when WinXP has been developed, and you'll see, that the Linux Distros weren't better in this. Same for the time when WinVista was developed, and the same for Win7 and Win8. So don't blame Microsoft for things that they did as good as everybody else did, that would be unfair. Anyways I just wonder who trusts software whose source code isn't open and and reviewed by a large community that don't have a financial interest on you. Ouch, wrong argument, realy! Nobody in the large opensource community had ever reviewed the heartbeat code in more than two years. This was not a harmless bug in a mostly unused library, it was a realy big issue in one of the most used library in the world and *nobody* saw it. Has openssl ever been carefully audited? I don't think so and i bet that there are more heartbleed like bugs in openssl. On the other hand schannel (the Windows cryptolib) is regularly audited. Sorry, but the large opensource community is blind on both eyes, whereas the closed source community is only blind on one eye. - -- Kind Regards, Mit freundlichen Grüssen, Markus Kohlmeyer Markus Kohlmeyer PGP: 0xEBDF5E55 / 2A22 1F71 AA70 1AD1 231B 0178 759F 407C EBDF 5E55 -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.22 (MingW32) iQIcBAEBCgAGBQJTVAHEAAoJEHWfQHzr315VBREP/11tSjfwXJEOmJ4kC/PG+8lj LN1vb/vMLWDny7HaTYRqCX8k49rhLCXTXeCB5sb10qeUX3rtqrudrYc/sGwI/H3E euhiU4oGK4YLDEQkXQDtFaSjpV0hayNQZruiz7x1mFO5FB05IqBmbSmxZLgKndE/ ydRevZPjvX475yVQ8YBsZRlLP8JQnOfec+4pxh8X+pq8/7THWOpdMqmaXdHwZVkM p0aZX3NBKQyFCtV/JTriQ/ByRC3OfDQbPY75qup13kr74B1222EGFpStU7TGedgR A+N36/9VrH/DCObQ78c8hsDfjLgRE5eJXBuYlL9dznKM68DrXGVP3hRkvuRozdQx KlNeQalkVd2p561dY0m92sCqitQIlTJ4uYEMeo4uIURYfBecQz/21Hk4cAATd+Yo IFamC3N8lO/iww4JsAzPgaywi2mkBzxH5o0jmSjmUW/aq/Mype7FADE9yFBNYr9P +E/d8AItXM7ouVfE2n55iz7i8wKIeyxZ2Ob1JMP0Bidxv9haihYMA6VYlmkryPmk TtR+oLxsGywRz21rBghOvzuyy7cyfBX2uHCo55VmxYIVGMfnYKF4ww/5iomgVtwp wTRo8JYPu8ixW9GAwtfoydEr4N4WMWyqHveNX+YqQMiQqlz2U1MieL/ILyHplRh8 PU4uCElXb5sYR6/w0KEV =5FIt -END PGP SIGNATURE-
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: On Sunday 20 Apr 2014 10:10:42 Dale wrote: Mick wrote: SSL-Session: Protocol : TLSv1 Cipher: RC4-MD5 == RC4 is considered completely broken today, even for Microsoft! :-) http://en.wikipedia.org/wiki/RC4 The good news are that your bank's servers do not leak any secrets at this moment and it seems they never did (they use SUN servers). Yet. I would rather not be the next customer to have his ID stolen like Target, I think the chain Micheal's was stolen in the past couple days but not positive on that yet. That bank is not a small bank and I pay fees each month for them to be able to keep their stuff updated. If they can't be bothered to keep it updated and then turn around and give me a card that sucks, well, oh well. picture a thumbs up here Just a 1/3 of all websites offer TLSv1.2 at the moment and hardly any public sites offer it as an exclusive encryption protocol, because they would lock out most of their visitors. This is because most browsers do not yet support it. MSWindows 8.1 MSIE 11 now offers TLSv1.2 by default and has dropped the RC4 cipher (since November last year). I understand they are planning to drop SHA-1 next Christmas and have already dropped MD5 because of the Flame malware. This should push many websites to sort out their encryption and SSL certificates and move away from using RC4 and SHA1 or MD5. As I said RC4 has been reverted to by many sites as an immediate if interim defence against the infamous BEAST and Lucky Thirteen attacks. According to the Netcraft SSL Survey (May 2013) only a third of all web servers out there offer Perfect Forward Secrecy to ensure that even if the encryption keys were to be compromised, previous communications cannot be retrospectively decrypted. Elliptic Curve algorithms are not yet included in many browsers and in any case the security of these in a post-Snowden world should be questionable (well, at least the arbitrarily specified NIST-NSA sponsored curves, which OpenSSL is heavily impregnated with). What I'm saying is that there may be no perfect banking website out there, because Internet security is screwed up at the moment, but it is always worth looking for a better bet. Well, my bank only got a C for it's grade. For what it costs every month, it should get a A+. I don't have one of those free checking accounts. I pay fees each month for mine. Plus I have already been planning to switch ever since they switched my debit card from Visa to Discover. I'm tired of finding something online or going into a business to buy something and then find out they don't take Discover. It's just a matter of speed of switching that has changed. Basically, just one more nail in the coffin. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Apr 19, 2014, at 18:29, Dale rdalek1...@gmail.com wrote: Mick wrote: Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Well you can use ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhideResults=on -Matti
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Saturday 19 Apr 2014 16:29:34 Dale wrote: How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Some banks have reverted to RC4 to protect against TLS v1.0 attacks from the BEAST. I don't think that FF shows the algos used for key exchange and encryption in enough detail. You can see them if you use Chromium and click on the green padlock. I use openssl s_client, e.g.: openssl s_client -connect www.wellsfargo.com:443 and look for this info: New, TLSv1/SSLv3, Cipher is RC4-SHA Server public key is 2048 bit Secure Renegotiation IS NOT supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-SHA -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: On Saturday 19 Apr 2014 16:29:34 Dale wrote: How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Some banks have reverted to RC4 to protect against TLS v1.0 attacks from the BEAST. I don't think that FF shows the algos used for key exchange and encryption in enough detail. You can see them if you use Chromium and click on the green padlock. I use openssl s_client, e.g.: openssl s_client -connect www.wellsfargo.com:443 and look for this info: New, TLSv1/SSLv3, Cipher is RC4-SHA Server public key is 2048 bit Secure Renegotiation IS NOT supported Compression: NONE Expansion: NONE SSL-Session: Protocol : TLSv1 Cipher: RC4-SHA I have this little padlock looking thing too. I dug around and found this info: CN = VeriSign Class 3 Extended Validation SSL SGC CA OU = Terms of use at https://www.verisign.com/rpa (c)06 OU = VeriSign Trust Network O = VeriSign, Inc. C = US PKCS #1 RSA Encryption There is another place with info but it doesn't allow me to highlight it so that I can copy and paste. Hmm. Anyway, is that reasonable for a bank to use? In case you haven't noticed, I'm not a wealth of info on encryption, just rich in questions. I just know that it is supposed to make things unreadable without a password, pass key or whatever. This is currently my bank. http://cadencebank.com/ Since they changed to a card that a lot of stores don't take, that could be changing real soon. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Matti Nykyri wrote: On Apr 19, 2014, at 18:29, Dale rdalek1...@gmail.com mailto:rdalek1...@gmail.com wrote: Mick wrote: Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Well you can use ssllabs.com http://ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhideResults=on -Matti I get this. https://www.ssllabs.com/ssltest/analyze.html?d=cadencebank.comhideResults=on I don't know a lot about this encryption stuff but mine don't look to good. :/ You got your test graded and mine seems to be bad enough to not even deserve a grading. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
-BEGIN PGP SIGNED MESSAGE- Hash: SHA512 On 19.04.2014 21:33, Dale wrote: Matti Nykyri wrote: On Apr 19, 2014, at 18:29, Dale rdalek1...@gmail.com mailto:rdalek1...@gmail.com wrote: Mick wrote: Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Well you can use ssllabs.com http://ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhideResults=on - -Matti I get this. https://www.ssllabs.com/ssltest/analyze.html?d=cadencebank.comhideResults=on I don't know a lot about this encryption stuff but mine don't look to good. :/ You got your test graded and mine seems to be bad enough to not even deserve a grading. Dale :-) :-) You have to use the https-URLs like this one: https://www.ssllabs.com/ssltest/analyze.html?d=secure1.cadencebanking.comhideResults=on Very secure your banks customer-login ;) Time to move to a safer bank... - -- Kind Regards, Mit freundlichen Grüssen, Markus Kohlmeyer Markus Kohlmeyer PGP: 0xEBDF5E55 / 2A22 1F71 AA70 1AD1 231B 0178 759F 407C EBDF 5E55 -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.22 (MingW32) iQIcBAEBCgAGBQJTUtHkAAoJEHWfQHzr315VmSsQALxIc+McPiO66/eHX2SURyiP ruo56b2NnB/8h2PNM6u9yeqfqwQeMjuj3CmNM6NnM8R4S1CUBGNP14iubAvgKdWc JSjkWZZMO/zkdpJQPatWsFriKNRAWncybrc/cQ/h6b9SXsHu0PYtk3grFlKm7fSp k8bLWTJCTy3pKQW86g3rJ5vGjXP8n1BjF0GgQ5WJjBKJK8iUn8oGzG0M4He2Zi33 t1HONE1lpkpgUG1r7PycKCobQMypMbd/h5zXLrd6RwM5qP6tK/ZzNsFkGXoT3bWt icZCdlH2vmWUlFMB0fdGORyu0u0qhQ5VVL4PS+rd//aq8WhAobGoTomRU+woSBQY xuzqmm1ssxrqiPeHCbXvTOnl6EPAWIcCj9YloonPLB9SpNizXufqD5exVJWme/2H BZYMmfRK2kkIhvEnxb2js7COQtyVcajQ8paQfaYMmKjk/HtdZwfvYy4OV9TLXI7g zfPKo9vnXPnmg2DvVFnhRZC+KTlqyMC8fZ3TI6uPOWhaQIspx6gN0FpA8gEzD1mt 6Z38C7IdJrv72B3hDoXrZnIA6kmCEVQgky3G5Um1h+qNT19Xb2UuJPnYJoG9Q7KP idvxPaZBlXUvUePB2CZu2o1plS4m6Nyoglv6HCCdRB9tbonx5NOas6LefiZCpHkk ceNA84TmDiuVdt32nRZY =mc4j -END PGP SIGNATURE-
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Joe User wrote: On 19.04.2014 21:33, Dale wrote: Matti Nykyri wrote: On Apr 19, 2014, at 18:29, Dale rdalek1...@gmail.com mailto:rdalek1...@gmail.com wrote: Mick wrote: Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ How does one find out what their bank uses? I'd like to check on what mine uses. I have Seamonkey and Firefox installed here IF it matters. Well you can use ssllabs.com http://ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhideResults=on -Matti I get this. https://www.ssllabs.com/ssltest/analyze.html?d=cadencebank.comhideResults=on I don't know a lot about this encryption stuff but mine don't look to good. :/ You got your test graded and mine seems to be bad enough to not even deserve a grading. Dale :-) :-) You have to use the https-URLs like this one: https://www.ssllabs.com/ssltest/analyze.html?d=secure1.cadencebanking.comhideResults=on Very secure your banks customer-login ;) Time to move to a safer bank... Well, I have had doubts about their security for a while now since I think they run windoze anyway. This sort of confirms it. They changed from Visa for their debit card to Discover about a year ago. I'm get pretty fed up with going places and them NOT take my card and me being stuck in awkward situations. Then finding out that their security is just barely half what it should be, yep, time to find a new bank. I been putting this off for a while now. As some know, my brother had cancer and I been busy dealing with that. We lost him about a month ago so I'm trying to play catch up. He beat the cancer but we think he took to much meds by mistake and it was to late by the time he realized it. Changing banks is on my todo list and may have just took a higher priority. It just went from not worth much to not worth spit. ;-) At least now I know how to check any potential new banks that I am interested in too. Thanks for sharing that howto info. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Saturday 19 Apr 2014 18:43:50 Matti Nykyri wrote: Well you can use ssllabs.com. I use it for debuging. Here is what Bank of America uses: https://www.ssllabs.com/ssltest/analyze.html?d=www.bankofamerica.comhideRes ults=on Well, that's an eye-opener and no mistake. I see my bank is rated B overall. Could be worse I suppose. Maybe I should forward the results to them. -- Regards Peter
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Matti Nykyri wrote: On Apr 17, 2014, at 23:17, walt w41...@gmail.com wrote: On 04/17/2014 11:43 AM, Matti Nykyri wrote: I don't know much about the secp521r1 curve or about its security. You can list all available curves by: openssl ecparam -list_curves I don't either, but I hope this guy does :) http://www.math.columbia.edu/~woit/wordpress/?p=6243 Good article :) The overall picture I had about EC is more or less the same as described in the article. But you always have to make a threat analysis and it depends on the private data you are protecting. By definition any private data will be disclosed given enough time and resources. So if your adversary is NSA... Well protecting the communication of regular internet user and your production server with SSL and x509 certificates will just not secure the content. I'm 100% certain that NSA has access to at least one CA root certificates private keys. With those they can do a man-in-the-middle attack that the regular user will most likely never spot. I my own security model I'm protected from NSA by the fact that it will disappear in the flow of all other traffic because NSA is not stealing credit card numbers :) ECDSA with ECDHE is fast and secure according to public sources. The problem is totally different if you are protecting the secrets of your company that are within the interest of NSA. I'm lucky I don't have to try that. On this topic about NSA, I read a article that claimed the NSA was able to view httpS traffic live or close to live since they had some backdoor access keys. I don't recall where the article was but since this is a knowledgeable bunch, is this true? If for example I go to my bank or credit card website, can they easily view that traffic? One reason this jumped out at me was that in the article, it was claimed that a group of people was going to rewrite the code/software/whatever for httpS and other encryption tools. If someone has links to such info for me to read and pass on to others, that would be great too. Thanks. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Friday 18 Apr 2014 15:27:12 Dale wrote: Matti Nykyri wrote: On Apr 17, 2014, at 23:17, walt w41...@gmail.com wrote: On 04/17/2014 11:43 AM, Matti Nykyri wrote: I don't know much about the secp521r1 curve or about its security. You can list all available curves by: openssl ecparam -list_curves I don't either, but I hope this guy does :) http://www.math.columbia.edu/~woit/wordpress/?p=6243 Good article :) The overall picture I had about EC is more or less the same as described in the article. But you always have to make a threat analysis and it depends on the private data you are protecting. By definition any private data will be disclosed given enough time and resources. So if your adversary is NSA... Well protecting the communication of regular internet user and your production server with SSL and x509 certificates will just not secure the content. I'm 100% certain that NSA has access to at least one CA root certificates private keys. With those they can do a man-in-the-middle attack that the regular user will most likely never spot. I my own security model I'm protected from NSA by the fact that it will disappear in the flow of all other traffic because NSA is not stealing credit card numbers :) ECDSA with ECDHE is fast and secure according to public sources. The problem is totally different if you are protecting the secrets of your company that are within the interest of NSA. I'm lucky I don't have to try that. On this topic about NSA, I read a article that claimed the NSA was able to view httpS traffic live or close to live since they had some backdoor access keys. I don't recall where the article was but since this is a knowledgeable bunch, is this true? If for example I go to my bank or credit card website, can they easily view that traffic? If your bank was using certain versions of openssl over the last two years, then *any* attacker who knew of the heartbleed bug would have been able to steal the private key of the server and decrypt all communications with it. It is rumoured (but could be FUD) NSA are likely to have known of this vulnerability for at least since November 2013. One reason this jumped out at me was that in the article, it was claimed that a group of people was going to rewrite the code/software/whatever for httpS and other encryption tools. If someone has links to such info for me to read and pass on to others, that would be great too. HTTPS on its own does not mean much, if it is using insecure (less secure) algorithms. RC4 and DES are no longer considered secure, but there are websites and browsers that still use them in preference to more secure cryptos. Some elliptic curves based algorithms peddled by NIST at the behest of NSA are now considered suspicious, if not downright compromised by design. http://safecurves.cr.yp.to/ -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: On Friday 18 Apr 2014 15:27:12 Dale wrote: On this topic about NSA, I read a article that claimed the NSA was able to view httpS traffic live or close to live since they had some backdoor access keys. I don't recall where the article was but since this is a knowledgeable bunch, is this true? If for example I go to my bank or credit card website, can they easily view that traffic? If your bank was using certain versions of openssl over the last two years, then *any* attacker who knew of the heartbleed bug would have been able to steal the private key of the server and decrypt all communications with it. It is rumoured (but could be FUD) NSA are likely to have known of this vulnerability for at least since November 2013. I'm a little vague on some things but it seems the claim was that NSA had some sort of backdoor that was built in from the beginning of the project for encryption which sounded like it would include httpS and others. Again, the details are fuzzy. I would say that I need to bookmark this sort of thing but I already have so many bookmarks that it is very hard to dig through them as it is. Adding more may be counterproductive, yet again. One reason this jumped out at me was that in the article, it was claimed that a group of people was going to rewrite the code/software/whatever for httpS and other encryption tools. If someone has links to such info for me to read and pass on to others, that would be great too. HTTPS on its own does not mean much, if it is using insecure (less secure) algorithms. RC4 and DES are no longer considered secure, but there are websites and browsers that still use them in preference to more secure cryptos. Some elliptic curves based algorithms peddled by NIST at the behest of NSA are now considered suspicious, if not downright compromised by design. http://safecurves.cr.yp.to/ Neat link. Lots of red stuff, which I assume is bad. ;-) Will dive into that more later on. Thanks. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Friday 18 Apr 2014 19:08:21 Dale wrote: I'm a little vague on some things but it seems the claim was that NSA had some sort of backdoor that was built in from the beginning of the project for encryption which sounded like it would include httpS and others. Again, the details are fuzzy. I would say that I need to bookmark this sort of thing but I already have so many bookmarks that it is very hard to dig through them as it is. Adding more may be counterproductive, yet again. I think that you are referring to their Dual_EC_DRBG (Dual Elliptic Curve Deterministic Random Bit Generator) which is/was used by RSA Security (not RSA the algorithm developed by Ron Rivest, Adi Shamir and Leonard Adleman). http://www.computing.co.uk/ctg/news/2295881/rsa-warns-customers-against-nsa-compromised-security-product# I don't know if Schneier said, stay away from elliptic curve algos and use symmetric keys instead, because of this. Others have tried to crack elliptic curves and have not been successful - so one has to tread carefully. Given the NSA/NIST and big corporates are all in it up to their neck, I would guess that distrusting *everything* they have or could be behind is a healthy attitude to take at the moment. ;-) -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
Mick wrote: On Friday 18 Apr 2014 19:08:21 Dale wrote: I'm a little vague on some things but it seems the claim was that NSA had some sort of backdoor that was built in from the beginning of the project for encryption which sounded like it would include httpS and others. Again, the details are fuzzy. I would say that I need to bookmark this sort of thing but I already have so many bookmarks that it is very hard to dig through them as it is. Adding more may be counterproductive, yet again. I think that you are referring to their Dual_EC_DRBG (Dual Elliptic Curve Deterministic Random Bit Generator) which is/was used by RSA Security (not RSA the algorithm developed by Ron Rivest, Adi Shamir and Leonard Adleman). http://www.computing.co.uk/ctg/news/2295881/rsa-warns-customers-against-nsa-compromised-security-product# I don't know if Schneier said, stay away from elliptic curve algos and use symmetric keys instead, because of this. Others have tried to crack elliptic curves and have not been successful - so one has to tread carefully. Given the NSA/NIST and big corporates are all in it up to their neck, I would guess that distrusting *everything* they have or could be behind is a healthy attitude to take at the moment. ;-) Well, I just wondered if it was true or not. If the NSA has some sort of back hack then encryption to them is meaningless. Thing is, I don't know if it is true or not. I wouldn't be surprised if it is for sure. I try to keep things as secure as I can and protect myself from the bad guys but this sort of things makes me wonder if it really does much if any good. If companies/governments have backdoor ways to get passed it, then there is no way to know who else can use that too. All it takes is for one employee/contractor with the knowledge to decide to sell out and then the whole thing is compromised. Imagine if it were to come out that there is a backdoor key to all the encryption that is currently in use. That would really throw a wrench into the whole internet community. I just read that yet another store has been hacked into and customer info stolen here in the USA. Waiting to see it from a reputable source before getting to deep into it. Of recent, I have seriously thought of encrypting my /home partition. I'm not a crook but like a guy said once in a TV interview, if a person looks long enough and hard enough, they will find something then build a career off building the rest. There are to many laws for anyone to really be able to safely say they have never broken the law before. I thought I read that article on Linux Journal but I can't find it there so it must have been somewhere else. shrugs Thanks. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Friday 18 Apr 2014 21:27:28 Dale wrote: Mick wrote: On Friday 18 Apr 2014 19:08:21 Dale wrote: I'm a little vague on some things but it seems the claim was that NSA had some sort of backdoor that was built in from the beginning of the project for encryption which sounded like it would include httpS and others. Again, the details are fuzzy. I would say that I need to bookmark this sort of thing but I already have so many bookmarks that it is very hard to dig through them as it is. Adding more may be counterproductive, yet again. I think that you are referring to their Dual_EC_DRBG (Dual Elliptic Curve Deterministic Random Bit Generator) which is/was used by RSA Security (not RSA the algorithm developed by Ron Rivest, Adi Shamir and Leonard Adleman). http://www.computing.co.uk/ctg/news/2295881/rsa-warns-customers-against-nsa -compromised-security-product# I don't know if Schneier said, stay away from elliptic curve algos and use symmetric keys instead, because of this. Others have tried to crack elliptic curves and have not been successful - so one has to tread carefully. Given the NSA/NIST and big corporates are all in it up to their neck, I would guess that distrusting *everything* they have or could be behind is a healthy attitude to take at the moment. ;-) Well, I just wondered if it was true or not. If the NSA has some sort of back hack then encryption to them is meaningless. Thing is, I don't know if it is true or not. I wouldn't be surprised if it is for sure. I try to keep things as secure as I can and protect myself from the bad guys but this sort of things makes me wonder if it really does much if any good. If companies/governments have backdoor ways to get passed it, then there is no way to know who else can use that too. All it takes is for one employee/contractor with the knowledge to decide to sell out and then the whole thing is compromised. Imagine if it were to come out that there is a backdoor key to all the encryption that is currently in use. That would really throw a wrench into the whole internet community. I just read that yet another store has been hacked into and customer info stolen here in the USA. Waiting to see it from a reputable source before getting to deep into it. Of recent, I have seriously thought of encrypting my /home partition. I'm not a crook but like a guy said once in a TV interview, if a person looks long enough and hard enough, they will find something then build a career off building the rest. There are to many laws for anyone to really be able to safely say they have never broken the law before. I thought I read that article on Linux Journal but I can't find it there so it must have been somewhere else. shrugs Encryption still works, at least for some attackers. The fact that burglars can pick locks doesn't mean that you should leave your door unlocked. FWIW I just checked my bank's website encryption ... they *still* use RC4!!! O_O I guess they are keen to make sure all these customers with WinXP and MSIE 7.0 can still login? For crying out loud! It seems that RSA's days may be numbered and elliptic curve cryptography would be the way forward, not because of resource constrained mobile devices, but also because of recent advances in crypto-analytics which may make RSA obsolete: http://www.technologyreview.com/news/517781/math-advances-raise-the-prospect-of-an-internet-security-crisis/ -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On 04/17/2014 11:43 AM, Matti Nykyri wrote: I don't know much about the secp521r1 curve or about its security. You can list all available curves by: openssl ecparam -list_curves I don't either, but I hope this guy does :) http://www.math.columbia.edu/~woit/wordpress/?p=6243
Re: [gentoo-user] Re: Heartbleed fix - question re: replacing self-signed certs with real ones
On Apr 17, 2014, at 23:17, walt w41...@gmail.com wrote: On 04/17/2014 11:43 AM, Matti Nykyri wrote: I don't know much about the secp521r1 curve or about its security. You can list all available curves by: openssl ecparam -list_curves I don't either, but I hope this guy does :) http://www.math.columbia.edu/~woit/wordpress/?p=6243 Good article :) The overall picture I had about EC is more or less the same as described in the article. But you always have to make a threat analysis and it depends on the private data you are protecting. By definition any private data will be disclosed given enough time and resources. So if your adversary is NSA... Well protecting the communication of regular internet user and your production server with SSL and x509 certificates will just not secure the content. I'm 100% certain that NSA has access to at least one CA root certificates private keys. With those they can do a man-in-the-middle attack that the regular user will most likely never spot. I my own security model I'm protected from NSA by the fact that it will disappear in the flow of all other traffic because NSA is not stealing credit card numbers :) ECDSA with ECDHE is fast and secure according to public sources. The problem is totally different if you are protecting the secrets of your company that are within the interest of NSA. I'm lucky I don't have to try that. -- -Matti
