Before you publish something, you choose the option if it’s transparent (permanent) or dynamic (it can change). And client accessing these data know it. That would increase the credibility or decrease it depending what is your intention.
@happybeing I hate to bother you with something I should probably know by now. But, deduplication, I would think, must take place at the 1 MB chunck level. Right? And each chunck is encrypted uniquely depending on account that put file. Right? How will any chuncks be identical to be deduplicated?
Your 1 MB I don’t know but maybe at a very rare case to begin with it’s if I put multiple copy of the same file on different folder. Outside of that I don’t really know how it can be done…
Yes @Ghaunt, your comment " Has I am aware again, logically the de-duplication only work on each account." Made me think this must be true. As unique encryption for each account.
I suspect there is a lot of misunderstanding on this thread. Hopefully, I am not adding to it by linking self-encryption, which ties in with how de-duplication works: self_encryption/README.md at master · maidsafe/self_encryption · GitHub
In short, IIRC encrypting chunks is independent of how the data map is encrypted. This allows both private and public data maps to point to the same chunks (enabling De-duplication).
1 Like
Man common, are you serious. April 2014 and that README.md file you want me to read. I already read that.
I think you and @digipl are confusing about an improvement I was trying to make here. And talking about it and comparing it with the current system.
Tbh, it is difficult to know what you are talking about, as you seem to be hopping between existing functionality and proposed functionality. In the process, others look like they are being confused too. I wouldn’t like this to become FUD rather than constructive debate.
Perhaps create a new thread (or add to the old thread - SAFE Network storage abuse) with your proposal for adding new delete functionality instead of adding it to this one about existing system behaviour?
I just don’t understand what you are saying it’s just too confusing. Even if I read it thousand of times.
Tldr; create a new thread for your proposal.
No… the proposal is already done and debated.
I just understand sorry. So you don’t want a mix between the current and the suggested. Unless the post are transferred to an another topic by a moderator for me the discussion it’s finished anyway.
1 Like
Chunk encryption is done symmetrically, not asymetrically.
In symmetric encryption both encryption and decryption are done by the same key.
De-deplucation works network wide, because regardless who encrypts a file, the resulting encrypted chunks always end up the same if the source file is the same. It’s the principle behind self-encryption. The datamap contains the keys with which the chunks were encrypted, and because it is symmetric encryption they can also be used for decryption.
The difference between a public and a private file is that the public file’s datamap is uploaded to the network without encryption, and for private data you encrypt the datamap with your public key (so you can decrypt it with your private key. Here we do use asymetric encryption).
6 Likes
No. The chunks, are encrypted based in their own contents irrespective of the owner, so If two people save the same file their will access the same chunks.
This is a basic security of the SAFE network that don’t connects data and owner. And a great benefit to the whole network prevents maintain the same data twice.
Try to broke this security adding an owner to the chunks is remove one of the basic characteristics of the SAFE network making it more complex, bulky and unsafe.
5 Likes
@janeannford I think your question to me has been answered, along with the corrections to @ghaunt who needs to do more reading, including things he says he’s already read, but not yet grasped (e.g self encryption).
4 Likes
In my mind it’s a bit like there’s a colony (which numbers in the many billions) of data ants, crawling from hard drive to hard drive and when called upon will deliver their data gift and in the process reward the hard drive it was on at the time with a few safecoins, with additional kickbacks going to its parents (maidsafe and dev pods), and whichever nice people created the app which called the data.
When the data map is deleted this ant will have its xor location struck from the record at which point it becomes useless encrypted noise and will wonder the hard drives of the work like its ant brethren, but without ever being asked to perform another task. So whether an ant goes into retirement with its encrypted data no longer accessible, or is squished/rebooted instead (which seems mean) then there’s not much of a difference really in terms of real world impact to end user of true data deletion.
Does this make sense or do I have some details of the logic wrong (highly probable:)
1 Like
I think this is a useful thought experiment. Its hard to come up with an “accurate” story, but I think images like this help because they make more use of our cognitive capacity than just linear thinking, which gets stuck in a few limited perspectives (hung up on certain assumptions and beliefs).
2 Likes
Things are again more clear. @happybeing Reading in english is harder for me than usual. There are some term like asymmetric vs symmetric wasn’t clear so I forget it while reading and some others too. It happened that I read the same docs more than one time yet.
Thank you all.
2 Likes
No worries, it is obviously more difficult reading in another language, but to be honest we’ve all had to work hard to understand how this system works, and most of us have had lots of patient explanations on the forum, not least from David. I don’t understand it all by any means, so I can just point out the things I have learned and encourage deeper reading of docs and code, and then follow up questions, from which we all learn a lot.
7 Likes
I’ve been following this forum for over a year, which has become almost an obsession, and if I have learned anything it is that there are thousands of hours of hard thinking from very smart people behind the SAFE network. As a result, developers, especially Irvine, are light years ahead from most of us.
Think a newcomer, and that barely knows basic aspects of the SAFE network, could find structural failure or propose to change basics aspects of the network is, at best, extremely naive (this forum is, in a British style, very politically correct. In my language, normally, I use less kind words).
This does not mean that our duty is to act as bloodhounds trying to find the slightest mistake, criticize what we consider wrong or try to give new ideas, knowing the difficulty, that will help the development of the network.
Among the bad things, we not find a way, not too difficult, to explain the SAFE network as a whole. A precision machinery, as a mechanical watch, where every piece has a function and any changes, regardless the whole, can have disastrous consequences. Getting to understand the beauty of the global design is, perhaps, one of the remaining tasks that all can help.
13 Likes