@bochaco was pretty clear that rewrites will be limited:
How big is a cusp?
This makes sense - but I must admit I also thought ‘oh no, another addition before testnet, what’s next week’s newest addition’
But always supporting the teams decisions take your time, get it right
Sound like a potentially useful sideffect could be for shared private data where one user pays and those who it’s shared with can update without paying, useful for example for new users who may not have any safe network tokens yet.
Yes, It was more of a response to whoever said they should be allowed to be unlimited but pay for them.
For payments between users yes, the client needs to initiate and have a colluding few Elders. Remember though clients can create addresses anywhere and transfer funds to any section to get colluding.
3 quarts, 5mins and a 1/2"
I thought it would be about that.
Just in case anyone’s struggling, in the metric system that’s just over 3 litres, 5 minutes and 12.7 mm.
Time? We are in 2021. The testnet will happen in 2021
The way I currently see it, the design is complete. An implementation of this design is complete. What remains is refining the implementation to bridge any small gaps in the design and any implementation flaws which incorrectly deviate from the design.
So, I can see why it feels just around the corner to the team. Just when they think it is ready, another bug pops up. Then they have to figure out which of the above it is, then address it. Next week they may find no more bugs. Maybe they will find several. How long does it take to find them all? That rather depends on how many are found.
Think of it positively - we have an implementation. There are bugs, but they are obstacles which can be removed. That is a good place to be in.
The optimist will always hope no more will be found. The pessimist will always think there are more to be found. The reality is probably somewhere between the two.
Is it just the owner that can mutate for free? If it is a public data item, it would be frustrating if someone else filled it shortly after you created it.
Each piece of content has a Policy, which is basically a set of permissions mapped to public keys, these can be BLS or Ed keys, which means you can set edit permissions for multi/single signature.
So this is only for private data (shared or not)? I’m just trying to grasp the limits/impact of the proposal.
For both public and private, you set the policy with the permissions.
Sorry but It looks like in month, two or 3 months, please remember about past
OK, thanks. I think I understand. I hadn’t realised there was so much control over public data items. I will take another look at policies in the code tomorrow.
Is it possible to have a public sequence that anyone can append to or is it always restricted to a defined set of keys? I thought it could be the former.
You’re being British, aren’t you? Or should that be English?
Is it possible to have a public sequence that anyone can append to or is it always restricted to a defined set of keys? I thought it could be the former.
Ah yes, that’s correct for public data that can be also set: https://github.com/maidsafe/sn_data_types/blob/master/src/sequence/metadata.rs#L202
This update also helped us identify a previously undiscovered issue that left streams open ultimately stalling network communications once the upper limit was reached. The Quinn team promptly assisted us and the issue is now fixed in
qp2p
.sn_routing
's communications are working flawlessly again! We expect all our crates to be updated in the next few days.
Great news! This sounds tricky to find and fix so thanks to the people involved for getting it done.
it’s felt like an uphill struggle with the challenges we’ve faced over the last couple of weeks, and that we are putting a seemingly never ending amount of temporary plaster over cracks that we know are comprehensively resolved by lazy messaging. So, we’ve decided to waste no more time here and move to using this pattern in
sn_node
Really pleased to see this. I understand people get frustrated about feature creep and timelines being extended and more ‘delays’, but there’s not much point putting out a testnet only for people to report many small ‘bugs’ that slow devs down from doing the fixes; especially if those bugs could be found and fixed right now without the extra / redundant reporting from the community. It would be a shame for the testnet to turn into community-doing-duplicate-bug-discovery. I definitely agree with fixing this now even if it means more delays.
Excellent; you’re on board now! If a testnet comes out in 3 months with a bunch of features we weren’t expecting to have with bugs smoothed out that is what I call an absolute, resounding success. And if we get that in a month even better!