I love thisproject however I have a problem with the witnesses choose mecanism, aka "12 catpains of industry" or well reputed equivalent.
First (and biggest one) : who is to coopte such "captains". In a supposed decentralysed network it means a central authority somewhere or at least a consensus voting system.
Second : what if no such "captain of industry" joins in? Should we bless project founders for eternity ? (we all die some day).
Three : 12 is a too limited number. This is intentionaly set so that we can humanely follow the witnesses quality. But I do not want to ask periodicaly myself whereas witnesses are good quality or not. I just want to feel confident with the network any time.
I would suggest the following :
- no more limited witnesses number
- all wallets could optionnaly run as witnesses
- if a end-user wishes to act as a witness is has to back some "good" amount of GBYTE (prevent spam and initial amount for serial post).
- each time a witness serial post he is awarded more than its post cost (aka end users are prone to become witnesses for profit)
- wallets choose their witnesses in a hard-coded random fashion among all witnesses pool.
First of all, I remind you that following the witness lists of "captains of industry" is just a suggestion how witness lists could be updated, it is not a protocol-level thing.
That said, I don't think the question of captains of industry is that hard. If you ask me about captains of auto industry today, I would name GM, Ford, Toyota, Renault, Michelin, and a few others. In Bitcoin industry today: AlphaBay, Coinbase, Localbitcoins, Overstock, Polo, and a few others.
But I do not want to ask periodicaly myself whereas witnesses are good quality or not.
You don't have to. If you don't want to or feel unable to make decisions yourself in this particular area, you can always follow the opinion of someone you trust.
To clarify, there is no such thing as "wish to act as a witness", you can only "wish to be named as a witness". Anyone can be named a witness by anyone.
The above was 100% human, 0% technical. Regarding suggested protocol changes, all such changes should be carefully analyzed with respect to potential consequences, in particular whether they could lead to consensus failures such as forks and deadlocks.