Thought experiments by those who do complicated thinking seem to indicate it should theoretically be just fine.
Our job will be to prove them right or wrong 
1 Like
It intuitively seems like it should be sufficient, as long as the mechanism to replicate when one host goes offline is quick & robust. The chances of 5 effectively random nodes all disappearing in quick succession seems incredibly small, assuming a decent sized network.
1 Like
Not sure if this is of any significance to us, but a hole-punching related bug found and to be fixed in Libp2p:
13 Likes
Yay! Cashnotes were stuffing my cloud instance… hopefully this doesn’t have too much of an impact on node memory footprint!
10 Likes
“Either first attempt works, or we move on”
So be it!!!
2 Likes
Haha, missing context here is that we retry afterwards. Basically it looks like the odds of a retry being successful in this verification step are slim. (Normally the price has changed), so let’s carry on with the rest and then we retry/repay in a later step.
3 Likes
The devil is in the details!
2 Likes
No new testnet this week 
[cough]…
… [cough cough]…
Oh well, but another key feature just merged and ready to be tested along with all those scrummy bugfixes!
What’s next @joshuef?
7 Likes
He’s gonna fly dead fast round earth backwards and make it 2014 again 
4 Likes
Nooooo.
Anyway, I got muddled, have edited culprit.
1 Like
I know right, I have resorted to gluing little feet to my node machines in boredom if this carries on much longer who knows where it will end… paint?
Just kidding, normally a mad rush between tests.
5 Likes
I’ve downed tools. No more vdash announcements until the next testnet 
9 Likes
Strike!! Should we start a union?
8 Likes
It’s actually good to have a pause to run some OS and router updates for example. Sort out some test data as well. And maybe do a bit of gaming that I’ve been neglecting!
3 Likes
Heresy 
2 Likes
The Year of the Testnets was pulled over by the development Cops and fined for exceeding the testnet speed limits. Once the required number of PRs have been paid, testnets will be allowed to continue as long as the fast lane is used in future.
9 Likes
Fixing some mem issues we see in larger testnets w/ the royalties (progress coming along nicely there). If we get that sorted before PUT simplification that’ll be the one.
Otherwise we’ll see some testing of the PUT simplifications when that’s made it into main.
ORrrrrr if folk are really bored we could try out the latest patched libp2p and see how mem leaks are looking. That’s not that exciting from my POV and we can passively assess that via the other tests.
6 Likes
I vote against. Boredom is a problem, but exhaustion is too. I’d rather se us hungry until the more useful test cases are ready, and then go with vigour again.
4 Likes
Any chance the amount of logs we keep will be configurable in the next test?
1 Like
6 Likes