Update 28 October , 2021

I don’t know what the above post’s content means for program creation, but it sounds cool otherwise.

So, refactoring means less bugs. Sounds like a plan. Can never have enough refactor, then! As much as it sounds like that thing that’s already been done a lot of. Keeps the sanity in check.


Really appreciate these write-up as always—the gold standard in the space.

:ant: :ant: :ant: :ant: :ant:


It seems to me that there has been more than the usual amount of frustration these last few weeks amongst the followers of the project, with people crying out, in essence, for MaidSafe to make compromises. Be a bit more like these other projects, be a bit more focused on image, or price, or whatever.

That you continue not to compromise, there in the weeds with the magnifying glasses out hunting for these little and large bugs, getting closer and closer to clarity and simplicity and to the reality of this immense thing you’re making, is the reason I have such confidence in the whole thing.

That dedication and tenacity is rare indeed. Hats off, and thanks for a nice little read about bugs, fascinating to picture it even roughly.


Well that was a though luck… But big thanks for doing the dirty job in a way that is understandable for us less techie people as well!

And thanks to the whole team for one more week of good job!

Hmm… I don’t know, but this made me wonder about one thing: does the network have any autonomous activity, or is it so, that if nobody up/downloads anything, then nodes and network just sit idle?

I think a few weeks back there was some change in code with the idea that you wanted to promote some behaviour of the network even when there are no external stimulus for that? I just don’t remember what it was…

I’m wondering because I just recently learned, that the basic model of the function of neurons (in humans or any other species), that external stimulus causes action potential in soma, which then travels via axons to the synapses and from there to the dendrite of the next neuron etc. - is a bit simplified. Neurons actually fire all the time on their own and the external stimulus just makes them fire more often. This is at least a partial reason for the phenomena of brainwaves (Wikipedia - Neural oscillation).:

Oscillatory activity in groups of neurons generally arises from feedback connections between the neurons that result in the synchronization of their firing patterns.The interaction between neurons can give rise to oscillations at a different frequency than the firing frequency of individual neurons.

I’m just wondering if there would be some benefit for the network having some autonomous action? It could then for example know if some nodes drop offline before the node it asked to do anything? Maybe that would just generate network traffic for no good reason. But from a theoretical point of view you could kind of have time, but not as an overruling king of everything, rather just a not-so-exact pulse to dance to, if you don’t have anything better to do. Does Safe Network have pulse of it’s own, and would there be any benefit for it having one? In living organisms pulse is a rough measure of health, for example.


That’s a lot of improvement!

Man today I looked at coinmarketcap and top 40 starts at $4.8B and MAID is just at $241M. With all these latest updates getting us closer to beta network I cannot even imagine what we will see with full network launch!

Quality over quantity, thats what MaidSafe has always represented! Great job guys.


Thx 4 the update and your hard work Maidsafe devs

Seems like: if we help find bughunters we’ll have a stable network soon.

Keep hacking super ants


There have been some past discussions about auditing nodes to ensure all their chunks are secure, even the old chunks no one has requested in a long time. Ideally, every chunk on the network would be verified at some reasonable interval (ex. 24 hours).


This much is true, however surely when the network reaches a reasonable size and nodes may be storing tens of Tb, checking every stored chunk daily would be a intolerable burden?
or am I missing something?

1 Like

Happy Halloween weekend team keep plugging away :jack_o_lantern:


Is that highly advanced GIMP or neep carving skills?

OK,ok they are pumkins, no neeps but thats just US cultural appropriation of an ancient Scottish tradition. No our fault the Yanks canny grow neeps…

1 Like

Just a welder with a drill and a knife :joy:


That’s what I mean when I say I believe in expertise. A physician or a linguist would hardly stand a chance against you in that field.

1 Like

Welder eh? Wanna give me quote for the usual chassis suspects on a MX-5 NB? Im happy enough doing the sills etc but the chassis struts at the front are beyond me…

They wouldn’t stand a chance with me in many other fields either :wink:

Sorry I don’t repair cars (unless it’s mine) usually more hassle than it’s worth :woozy_face:

1 Like

Bugger… Anybody want an almost complete set of parts for a 2004 Mazda MX-5?

miles per gallon is iffy but smiles per mile is outstanding. If you are going to do it, do it sideways in a Mazda :slight_smile:

I need a multidisciplinary person who understands both welding and salt-water boats. (Rust developers who think welding aluminum is like welding “any iron” need not apply.)


Are you talking about sacrificial anodes? Definitely not in my inventory but did watch a good documentary the other day.

1 Like

Not really imo. For example, the nodes could be asked to verify a set of random 4k sectors. Given a 10TB disk means about 10,000,000 x 1MB chunks. One randomly selected sector from each chunk means only about 40GB of data needs to be processed per the given time interval. Presuming a gross processing speed of roughly 40 MB/s means that a node/vault verification could be performed in 15 minutes. Yes, this is an extremely rough estimate since we don’t know all the details yet.

As an example I can compute the sha512sum of a 10GB random datafile in 19.046 seconds on my computer’s ssd with page caching (that’s about 525 MB/s)… However, it is conceivable that most vaults, even low powered ones, could be audited one to four times per hour using this method. Eventually every randomly selected sector from every chunk would get audited. In the interim it is highly unlikely that a vault could pass the test unless they have actually kept all the data.


i love the regular update-posts!
keep it up


Thank you for the heavy work team MaidSafe! I add the translations in the first post :dragon:

Privacy. Security. Freedom