RE: The History of Delegated Proof-of-Stake (DPOS)
I had no idea of your history extending to those formative days! Although not sure I totally understood the fine points about POS, you explained clearly enough the motivations to seek an alternative in the form of the new dPOS that a layman should be able to grasp how it came about. Quite ingenious, actually, though we're still in somewhat "early days" of figuring out how to make the system totally trustless and resistant to different attack vectors.
Like other commenters, I wonder if we can explore in a major way various alternative options for witness voting — 1 SP, X votes (X being 1-30), the ability to vote for as many witnesses as you want, where the % of VESTs is split up between them, etc. What about an exponential curve of witness voting? Perhaps in the context of this series you might be able to enlighten us as to the decision to go with the 30x multiplication if you were around for that?
Anyway, grateful you are finding the time to keep the solid posts up in these hectic circumstances — there is still such widespread ignorance as to the basic functioning of blockchain that it's really important to have concisely-written articles like this available for people who are interested—I guess we are really all still early adopters? It feels weird to say that when we're a decade deep into BTC's life, but in the grand scheme of things that's really not so long. Be well @blocktrades :-)
The original design of Steem had unlimited votes. The idea was that you vote for all the witnesses (however many that is) you think will do a good job and not harm the system in any way, and so does everyone else. The elected witnesses are the ones that the most stake views as doing a good job and not malicious, which is likely to be very strong and broad support indeed. In this system you don't need to choose or rank your favorite witnesses, just rate them as good (vote) or bad/unkown (no vote).
It was recognized after Steem launched that votes do not only choose the top 20 (was top 19 at the time) witnesses but also determine the weight for backup witnesses. By voting for more backup witnesses (and recall that the number of votes was unlimited), someone can direct more of the backup witnesses weight, which is arguably unfair and also exploitable. Therefore a limit of 30 was put on it, which was seen as a balance between all the good stakeholders being able to vote for enough "good" witness candidates to fill all the top slots, and making even more votes which would upset the backup weights.