We are all in that team
Can someone explain what’s happening with qp2p and the dropping of ConnectionPool? Sounds like not an easy change and with lots of ramifications…why was that necessary?
From dev update a couple weeks ago:
I see word “refactor”
Yes, we refactor something to fix every bug The connection pool is causing issues and is extra code that makes assumptions that it should not. The fix is to have each connection handle its own lifespan and not have a thing in the middle. I would hope to see the end of connection lost being as frequent and unexpected with this, or at least show us where and exactly why it happens, allowing that code flow to consider retry etc.
61 posts were split to a new topic: Spentbook
The agogometer is not readily available to me at this time but I can imagine it trembling gently in its lead-lined case in the desk drawer at home.
Got that Thursday feeling
Any bugs squashed this week?
Every day of every week for last 4 months
Lot of mentions of blocks, and chains… interestingggg
Not so sure what it is though really.
Its a linked list datastructure.
@jlpell Guess I’m most curious what it’s purpose is.
Seeing it separated out and such.
As it is in the code we have that kind of verbiage in the code with things like section chains which are what evolved from data chains I believe but this one is new and I know there are some things coming along and wasn’t sure if perhaps it was related…
Edit: Upon further inspection, I suspect it may be related to this refactor(routing): unified network knowledge for updating SAPs and section chain (Section->NetworkKnowledge name change) by bochaco · Pull Request #673 · maidsafe/safe_network · GitHub
To update Section Authority Providers and Section Chain in a cleaner fashion. Still neato if true. Just fun seeing new things develop and to follow along.
Yes that’s it @Nigel it’s all lingo from data chains and now the Section Chain. So SectionChain is a linked list where a key is signed by the previous key. We don’t identify who made up that key except in the current block and that is where we link the
SAP (Section Authority Provider).
The SAP tells us the members of the group (Elders) plus their key and this section key part as well as identifying the prefix (as the section keys are not related to prefix and require a mapping). This SAP is signed by the previous key.
Would be interested to know beyond SafeId for the individual, what might become possible for groups shared access to data. I don’t know, if that’s a matter of iterating a shared password on each change and whether there are simple options for this. I don’t know then if necessarily a member retains access to what they had access to, in the event they fall outside that group later, or whether this is less fundamental and more a layer on top managing this.
I’m just speculating without getting traction to be sure of atm… so, I don’t know if one option is some mutable list of members and individual knocks the door that is the list and is granted access or not.
Nice, of course
I mean, did you guys get a stable internal network with 11 nodes.?
I think I seen tests revolving around 45 nodes in GH recently
Getting that Thursday feeling again who knows what we are going to find out tomorrow.
I used to measure time passing by Saturday Nights now it’s Thursday afternoons that are my peak of the week.
With so much Github action, it does seem like something big is on the horizon.
But it’s impossible to know
GitHub going quiet is a more likely sign of a new testnet I think.
Yea it’s when it goes quiet get exited