Rob,
now I was finally able to tithe (my coinage was enough ) so I decided to try "spam" test.
At the begginig there was dif around 8k and mincoinage was around 9 days.
I startet to spam with tx "tithe 259" and then it changed so "tithe 254"
I was able to made 26 x 259 and 31 x 254 txs. Totally 14613 BBPs in tithe within 5 blocks.
Then it stops because I havent more "packages" with corresponding coinage
At the end dif is around 29,5k and mincoinage 26,9. And my weight in pool is more than 0.6 (At the beggining was near 0).
Is it OK? Someone who has enough BBP can spam like this and others wouldnt be possible to tithe.
Or am I wrong?
I think we're OK as a lot of psychological thought has gone in.
Since we don't want to discourage legal tithes (IE a person with real coin age), as that would cause them to do sneaky things like try to create multiple wallets, and we dont want to deny legal tithes on a block switch (IE we quote a certain max tithe amount for the diff, block switches and user loses the whole tithe) - that would probably give us bad reputation, etc, I came to the conclusion that If a person has the ability to tithe it should count if they tithe it.
What will happen in prod is difficulty will shoot up (even if its 3 blocks late like in your case), and prevent everyone but the whales from tithing for a certain period.
The system is not 'broken' however, in that diff will drop as a few of the whale blocks drop off the 205 block lookback window.
So this type of thing will happen, even if we have something like this going on: lets say 8 whales took 30 days to build up enough coin age and they all mature at the same time. That means all 8 will start auto-tithing and diff will rise. So even though nothing nefarious was going on diff rose - another words we cant predict whats going to happen in prod at any given time whether auto or manual.
So basically the client will just react to this, and the user who did it really didnt get away with anything special, they relinquised their coin age and got into the pool - they choked out everyone else for about 12~ more blocks, then as diff drops others can get back in.
I do think that diff could be very high in prod due to this, but also remember right now we are only using 10% of our budget since PODC is still active. If we open the budget up diff would drop significantly.
On this subject in the opposite scenario, for a couple days in testnet my two whale nodes didnt have enough coin age to participate (RAndrews had an 80 mil balance) and couldnt get in the pool for 2 days now - and I observed low diff for a while;
So I think this situation will be similar to the cycles we see with BOINC rac, its interesting how we can observe elliott wave cycles even in apparently random sources.