134 replies to this topic

[blackpirate]

@willie
i see that exist some cracks for!
RaidenFTPD.v2.4.770.Cracked-AGAiN

edit:
yeAhh...2005 rls by SnD
[ CrackeR .............................. Teddy Rogers ]        
[ Rlz type ................................... Loader ]        
[ Prog name ......................... RaidenFTPD 2.xx


Cheers

Edited by blackpirate, 30 March 2010 - 11:55 AM.

[Erelim]

(Disclaimer: My knowledge is VERY limited, some of what I say may not even be possible xD)

My initial response to this question was 'Of course, anything that a computer can do has to be orderly and thus anything can be reversed or cracked'. However thinking about this a little I am not so sure. It really depends a lot on what your definition of cracking is and how long is reasonable.

As mentioned in this thread it is possible to use a proper encryption scheme to encode a program. Then, baring the five dollar wrench path, it would become impossible to crack a program based on the time it would take to break the encryption. However using this same scenario it still depends on what you define cracking as and what your goal is in the end. If your goal is to create a perfect keygen that anyone can use or to crack the encryption so that any entry will decode the app; You may be SOL. However, if your goal is simply to remove the program from the encryption a number of methods could work (a keylogger, a memory dump if possible, etc) but this then goes to the definition of cracking vs hacking vs just getting the job done...

[chickenbutt]

I think with the right engineering and manufacturing resources I could design something with no crypto-engine but a variant of DLPAR that people in the RE scene couldn't defeat. I'm talking about serious RE people too, not Geohot grade.

Features:
-ROM loader with e-fuse
-Dedicated hardware thread for memory checksums
-write protect based on isolated vectors
-DLPAR also with isolated vectors

This is based on a IBM design except with a flaw fixed. It'd take some exausting die-mesh manipulation to do anything. Nobody can do that right now.

If vendors wanted crypto offer a DMA RSA library.

With software I'd use inline ring3 VM with threads and thread checksums. Use heap a lot. Offer RSA based server auth for special cases. Hope one day hardware isolation that doesn't have the bad memory design of current x86 families comes along so people will quit cracking it.

Edited by chickenbutt, 26 June 2010 - 05:16 AM.

[Dooms_day]

of course everything is crackable, it was coded by humans, it can be cracked by humans

also note that when you argue that no one has came out with a crack to a crackme, it doesnt mean it cannot be cracked, it just means that no one has taken the time and thought to crack it

[E33]

If the program was coded by the spirit of god, then we can say it's un-crackable otherwise every program is possible to be cracked, only depend on the talent of the cracker.

[Chinga]

This is turning into some kind of debate about 'celestial programs'. I think there are some that are uncrackable, no matter the skill of the reverser.

[Thomas Long]

I'm sorry do disappoint you, but I know from a friend, that not everything is crackable when time
is limited to few years and when it is not allow to replace some function by ones self functions!

Edited by BytePlayeR, 26 March 2011 - 06:01 AM.
Changed annoying size.

[chickenbutt]

Not till resin guarded chip logic and hardware isolation without software-based or secure 'mail-boxing' comes along.

[quosego]

we can only hope that day will come soon..

[The Dark]

As far as I know given enough time on a program it can be cracked the only difficulties would probably be some Microsoft Products that being Office 2010.  However rarely there are programs which are uncrackable if the person gives enough time in order to Reverse Engineer it.  Operating Systems are a great example in this case though.

[chickenbutt]

The Dark, on 02 November 2011 - 12:29 AM, said:

As far as I know given enough time on a program it can be cracked the only difficulties would probably be some Microsoft Products that being Office 2010.  However rarely there are programs which are uncrackable if the person gives enough time in order to Reverse Engineer it.  Operating Systems are a great example in this case though.

Office 2010(latest) has been both cracked and license-spoofed. It's only a server check and a lot of multi-bin license checks. It's not even remotely hard, just time consuming.

Things that really aren't crackable:
-Dongle and ISO protectors with no inline key data and strong layered crypto(if you discount bruting for years)
-Hardware isolated decryption and signing where there are no accessible buffers or memory and have protection of silicon reversing(Doesn't exist yet(all implementations defeated through memory corruption))

[The Dark]

I could never get around it since from what I could see office 2010 (sp1) which is the newest uses license checks from the application and on the service side.  I would completely agree with the dongle and ISO protection on the Server side on what you list as really uncrackable.  I gave up on office 2010 and MAK activated it.

[chickenbutt]

The Dark, on 02 November 2011 - 04:49 PM, said:

I could never get around it since from what I could see office 2010 (sp1) which is the newest uses license checks from the application and on the service side.  I would completely agree with the dongle and ISO protection on the Server side on what you list as really uncrackable.  I gave up on office 2010 and MAK activated it.

Not even with server crypto, no disk no decryption. Modern protectors encrypt inline VM handlers and more using disk press data and there is no inline checks that reveal keys. ISO teams have private rebuild and trace tools for securom, byteshield, starforce, tanges etc..

Office has a lot of checks that's why everyone spoofs data, It's all common inline patches though, you don't even need crypto data.

The only thing existing that is potentially uncrackable is hardware based encryption isolation, and in the most advanced cases that's only till code execution dumps buffers..

My current philosophy is, wait till someone wanting recognition cracks it and go make money off contracts in the mean time. RCE doesn't pay as most people here can confess.

Edited by chickenbutt, 04 November 2011 - 01:58 PM.

[HACKAL]

Hi to everybody and specially for lena151 for this topic You're a legend lena and i'm one of your thousands rce students around the world.
it's an honor to write to you al least through here. well i want to give you my opinion and my point of view.
I think everything can be cracked, in the strict sense of the word, don't confunse with keygenning or patching only(will be discussed later).
Some targets can be a real headache even for the most hardcore and skilled crackers. just to put an simple example. They can use extreme and sometimes hidden(for the public eye) concepts to prevent reconstruction of the algorythm. can add also the populars RSA keys(just an example) 1024bits. All we know the RSA 1024 factoring it's not available and as a RSA Challenge it's frozen. For those who first time read some about RSA must to know that the factoring of RSA 1024 will take a lot and i said A LOT of years with supercomputers. The keys were chossen randomly and RSA keeps in secret and away from public eyes .
That's one of the reasons not the only. At this point we have a not keygennable Target. If this "Target" it's an headache itself imagine besides it's packed with an extreme Packer/Crypter like VMProtect, SafEngine, ZProtect, just to show a few some of the most hardcore packers. The most popular feature used by these packers are Virtualization, Mutation Engine, So. What we need to try to defeat them. Well an advanced knowledge of unpacking, vitual machines, crypto, and pe format specifiaction theory. However this it's not guarantee of sucsessful unpacking .
Another feature that makes not possible de keygenning its the server activation check.
By these 3 points exposed i think the Cracking(in the strict sense of the word) of a target it's 100% Possible, however when target it's more and more complex, we need more time( a lot lot lot lot of time ), more resources(pc, hardware, etc), more knowledge(coding skills, crypto skills, etc) and more effort to crack a target.
Keygenning by the exposed above it's not possible in all the cases. I think keygenning(in the strict sense of the word) it's possible in the 70~80% of the cases.
And finally patching i think it's possible in about 50% of the cases.

That's my opinion. Regards to all of you. specially for lena151
Hackal

[chickenbutt]

You can't just keygen where there is server checks and/or the algo is server side. Currently there are VM tech protectors you can't inline or rebuild without dongle or press data as well.

The crypto exists to make practically un-crackable DRM now, but it has to use buffer-safe hardware isolation and prevent silicon reversing on-chip storage(also doable via resin and mesh techniques).

Luckily for us everything is designed with emphasis on marketing and not security..I think vendors think if a doctor codes and/or designs it, it makes it more secure. All we have to do is read a text book to understand their knowledge XD