well done to all the team on another week of git hub madness I some times wonder when you all find time for sleep? !!
there was some mention of approx. 4 weeks till the next test net about 3 weeks ago do you think there is any chance of a cheeky test net next week @dirvine ? as I have manage to get one of the arm oracle VM’s as discovered by @happybeing up and running and standing by for test net
It is possible minus DBCs, an absolute flurry of simplification happening and will be nice to test it. Lots happening in background test suits as well so I expect much faster / better pre-merge code tests which will only make this even quicker.
Thank you Team for the news update. I see the excitement is building up. I and looking forward to the time when this Team will have their time in the “sun”.
Excellent. Private transactions are exceptionally important in a network and having them as default avoids falling into unknown pit holes that may appear.
Other things to consider (although im unsure if these apply to Safe Network) is that identification of clients could be possible based on network timing of participants to anonymise participants.
For instant if state level adversary was to run legitimate nodes but could see the requests coming across the network for such activity and can marry the two together (much like what folks were doing on Monero due to the network size of user base).
I believe the SAFE network design would make it prohibitively difficult to get a convergence of data points for this to be of much concern.
To get a single data meaningful data point as to identity, such an actor would have to control and be sniffing several random nodes, which nodes’ activities would be shifting randomly. Then identifying what any identity was doing is another problem, due to the self-encryption and other obfuscation factors. Capturing both ends of a communication/transaction multiplies the difficulty immeasurably.
The key defense is the constant mutation of factors engaged in by all honest nodes as default behavior.
That’s my assessment from a non-techie perspective, but I think it’s valid. Others please comment on my assumptions.
Maybe? This is sort of like sibyl-esque type monitoring is about the metadata of the packets being routed across the sections. If they are all blinded and look the same as everything else, it wont be an issue, if they can be differentiated then it becomes easier.
So it seems like moving on by attempting/adding DBCs outweighs the fixes (anti-entropy and maybe another?) that still need to be done from prior test nets? Maybe AE is more or less solved after some bugs, as well as it being easier to test more foundations at once during tests, is the reason for this continuation onward. So basically the rule is to squash most problems, not all, and then work on new features. Cool to know. Thanks for the work!
IMO, I think they are working on all problems at once now … the bugs aren’t always evident from any test, but they can’t hold back other work and divert all attention to bugs … so solve things in parallel as they progress.
Yes, diff guys are on diff bugs all the time right now. Also some on optimisation (significant savings, not profiling type micro optimisations) and some on DBC etc. as well as more work in CI by @chriso with our own @StephenC dropping in to check we are OK. Nice to be nice so it is! Very refreshing to see real humanity.
Though I don’t log in that often but following the updates daily. Its good to see rapid progress nowadays after so much. Still my expectation from the project did not change a little bit. A full fledged safe network is hard to imagine before 2025. I am expecting the same from 2017 when I joined. Best of luck to the team.
It’s incredible to me that Safe ended up using Mints, which were one of the possibilities that Satoshi described in his white paper. Although he did say that having mints reissue tokens means centralisation etcetera, but since this is the SAFE network, that issue has been solved