There seems to be some common myths of how POS works and why attempts fail.
By design, PoS is random. It can be forced of course, by modifying code to try to stake as many of your coins as possible, but this will *not* produce more rewards for you and will only result you having to wait longer for the next stake. The code is adjusted so staking happens 'regularly'. (except, it's random <grin> so no guarantees when)
This is not exactly as interest in the bank, where you have contract at which date exactly the bank pays back interest.
It is common misconception, that being 'away' from other nodes impacts your PoS negatively. While *in theory* it is so, one has to do some math to understand why it's not an issue *at all*. For example:
The current Internet has a maximum RTT of under 200ms worldwide. (ok, maybe 300?) If you are connected via two way satellite connection, that should go up to 600ms. This is .. 0.6 seconds for your block to reach the other node(s).
Compare that with the designed PoS interval of 600 seconds per PoS block. It is sometimes less, sometimes more. We talk statistics here, so it's essentially 600 seconds in the long run. This means, you have 1:1000 chance to miss a block in case you are connected via high-latency satellite link to the rest of the network. Should happen once every 1000 PoS blocks you discover, in the long run. (again, we talk statistics here, not someone's bad luck, or.. see below)
Now, if you are lucky, within Europe we have < 60ms RTT, so that gives "us" 1:10000 or 10 times better chances. This means, we could lose a PoS block every 10,000 blocks we generate, instead of "you poor satellite connected guys" who will lose one every 1000 blocks generated. Of course, next time your wallet will PoS, it will (statistically) succeed and because the coins were sitting there longer, you would generate the same profit (in percentages, more in absolute numbers).
By the way, there are many more factors, that influence the block acceptance and in general, they do introduce much larger delays, so the above examples are a bit extreme. Life is actually better with regards to RTT influence on PoS.
The real issues, besides doing PoS in a fork are different. It's one of the typical human sins, greed ;-)
More is not always better.
I am talking about the number of peers here!
There is a reason, why the wallet opens no more than 12 such connections by itself.
The reason is, the more connections you have, the more time you spend servicing them. Further, peers can request you feed them blocks, and if you have say 50 connections, each just booted, you spend pretty much all your time sending them blocks. This leaves you no time for PoS.
So even if you found a PoS (or PoW, same, or even greater problem) block, it will sit in the queue waiting to be sent to other nodes, while you feed them existing blocks (sometimes having to read from disk, do a lot of hash calculations etc).
In summary, PoS and large number of connections are a big no-no.
So those of you, who are having such difficulties
[teacher hat on]
Write 100 times the following (your choice where to write):
"I will not use addnode=1, noirc=0 or listen=1 if I am generating blocks, PoW or PoS. I Promise!"
If, after writing it 100 times, you still do not understand why, try writing it 100 more times. It helps, with time.
[teacher hat off]
Oh, and by the way, you will help reduce the network forks. Thanks!
