Regarding Project SAFE development, we are going to do a Q&A video conference (8/10/14) where questions posted on this thread will be answered in the video. We are currently in TestNet1, see the Roadmap.
Feel free to post your technical question in this thread.
We will focus on answering development questions. This is to help current and future developers understand the project and the code. If you’re not a developer but want to understand how the platform/protocol functions you’re welcome to post your questions as well.
An easy way to phrase your question is to look at it from a construction point of view. How do I build X or attach Y?
Don’t be shy, the more questions we have the better it will help everyone. Even if you know the answer to a question, please post the question so the (video answers) will be made available for future inquiries.
@Viv will likely give more info, this week has been debugging like mad (visualiser and vaults), but you can check this run apps against it (again Viv can explain in the examples) and you should after today, hopefully be able to run a local network drive and play around. Basically this is a device that will login create account/present drive and let you work on the drive as though it were local. Frantic pace right now, but its pretty much all there for dev testing for sure.
When we get a bit more info and some rudp and routing fixes in place then we will let you know for sure. ITs pretty much there though if you want to play around and anyone who creates their own private network can use the visualiser to see what’s happening. ITs pretty cool to watch the vaults all bootstrapping and joining (well for me anyway )
The visualiser tools is about to have a few major changes merged to it’s master branch hopefully by today. Just waiting on the c++ reporting side to match the new requirements. Again, this should be done sometime tonight so I’ll add the full updates/features added info once it goes live
As for examples itself, bootstrap changes required tests in corresponding projects updated. With that out of the way( @prakash is working on this currently) we’ll have the authentication example out which you should be able to use and join the testnet with.
^^ for more info and steps to setup your own test net’s and have the tool monitor it. You do need a google mail id to be whitelisted as that’s the only oAuth we have integrated.
@Ross is getting the testnet up right now on the new version. To know when it’s done, do check up on the mailing list thread.
I’m sure you got better technical questions Happybeing so ask
Does the messaging on the network have OTR? Off The Record
Is my account not vunerable because I have to enter a PIN, Keyword and password on a
keyboard? “From the Lifestuff example”
Sorry, but I’ll be honest the user credentials is a disaster (people now have to memorise 2things and you want them to memorise 3things? No way) (2FA, SQRL or BITID is way better to login)
Can the Lifestuff account be used for every single app on the network or can I create my own login mechanisme?
Could cryptolocker encrypt data on my VFS Drive and take the data for hostage?
Does the meta-data that’s on the network not make it possible for someone to monitor my
interactions?
In the youtube “API General Discussion, time 32:00” you talk about apps taking “Time-Outs” is this solved? (It think you can find a solution to this if you look @ how data-packets work on the current internet, there is UDP and another protocol, I’m to lazy now to go and find it)
Could me and my friends have a private Maidsafe network offline on our wifi island?
Will the Maidsafe be available as browser plugin?
With Project Zero Google is improving the internet with hackers, are you guys currently
working with hackers to improve Maidsafe?
Can you report bugs right in the network? Does the system detect when it crashes, if
it crashes?
What does the system do with “The same file” but different formats (the
Netflix thingy)? Maybe the system could get recognise “The same files different formats” in the future and just display 1 format “VIDEO” on all devices and delete the others
If I like to develop for Maidsafe in a crossplatform language? Which language is best suited?
Please developers don’t get angry @ me for saying/asking question 3 and 7, you girls/guys are hero’s (strange that I don’t say the female version, but what the heck I’m a guy and it’s a mans world, but it would be nothing without a woman our a girl)
I’m not really asking technical questions (matter of fact, not @ all), but I have a strange feeling like the people who are suppose to ask it are not asking…
I repeat, these are great questions IMO. Agree we could do with others but am not filled with q’s myself. I can think of the attacks q’s (what stops double spend etc), but these are well known.
It’s limitless really, is this not just a copy of freenet or tor ot tahoe etc. are the most common. Also
What happens if a computer goes down etc.
Where are the servers, I know for sure you must have servers
How do we know this is not a scam
This is hard, what makes you think you can you do it
How did you come up with this idea
Can you guarantee no bugs
Why are you not producing apps
Can I just download and run this (but I dont know what it is)
Why self encryption, its too much or it rubbish (but I cannot explain it to you)
Why do we need privacy
Others have security so what’s the difference
Are you really the CIA/NSA etc.
Why is the code not open source and available
Why not show us the code
Why no documents
Website video too technical how can I understand this
Website video not technical enough how can I understand this
Why not use bitcoin as payment
Why c++ and not language XX
These are probably the main ones that I have answered an awful lot. There are a ton more much more offensive questions, but probably not worth listing really.
Added this question from another topic discussion above.
Please demonstrate a whiteboard process of how data (origin and storage location) is obfuscated starting from the uploading node and ending at the storing node.
How does a (relay) node know the next relay, without knowing the final destination address?
Would there ever be an event where the uploading node sends a 1MB chunk directly to their close neighbor for storage?
Most people assume nodes will talk directly to each other in the network space. A video answer would be useful. Would it be accurate to describe XOR addresses similar to PO Box addresses? Unlike the physical world, these PO Box addresses can be accessed from any physical location.
My original concern was that claims were being made about the anonymity properties of MaidSafe, such as ‘yes it will be as anonymous as Tor’ which made alarm bells go off…
Then I saw in multiple documents that MaidSafe was based on a modified Kademlia DHT. The alarms increased in volume.
So I started hassling you guys, due to laziness and that I found it hard to find the answers to my questions in the docs
So I read the MaidSafe-dht paper.
I watched a David Irvine talk about Self Encryption and DHTs at the seattle conference (via youtube lol)
So yes this convinced me that there are big issues with this scheme.
BUT it later transpired that in fact your DHT is not really a DHT at all and is just used for routing messages. Both the paper I read and the video I watched were no longer relevant.
The user’s machine never issues a request to store anything on the DHT
The DHT in our case (its pidgeon holing to a bad extent) is routing and vaults and yes you do store in the vault network.
At the moment the window of opportunity for an attack is the close nodes you are connected to who are the nodes with knowledge of your IP and port (current). This is the part we have said quite a lot in recent threads is an area to watch and possibly improve on. This is where I was discussing a scatter gather approach to storage, the gather part is on the net and nowhere near the IP:PORT of the uploader.
In terms of TOR, I really like it, but its a hugely different proposition. As anonymous as X is a weird way to look at things, we need to measure how anonymous has X been and there are measurements available for TOR, but not yet SAFE. So we will see. Say TOR exposed 48% of people on it then SAFE exposes 48% then its the same perhaps, the design is different for a great many reasons though.
Have you read the system docs yet?, I think they might help a lot. There is an opportunity to ask further detail there and we can fill in any missing bits. I am sure there are loads of undocumented features, but we cna document them and point to the bit of the code to help auditing.
My original concern was that claims were being made about the anonymity properties of MaidSafe, such as ‘yes it will be as anonymous as Tor’ which made alarm bells go off…
~ Me, on reddit. Not MaidSafe.
Apologies if this was not a reasonable statement, though in still thinking it might be!
)
