You are viewing a single comment's thread from:

RE: Proposed changes in the Calibrae fork

in #calibrae7 years ago

You think I haven't been campaigning to improve this platform since I arrived last year? You don't think I have already got to know very well exactly what their SOP is?

I resisted making a fork for a very long time, once the licence was changed to MIT. A very long time I was just going to leave and build something from scratch. But I am a big believer in code reuse, and I know that I can make something better out of it, so I am.

Good luck with your campaign. When Calibrae is ready, you will be no less welcome than anyone else to come along and put in your 2 cents on anything and everything. When the second phase, with the total rewrite is done, the code itself will be like the forum, where programmers can earn by making contributions to it.

The fork is just phase 1 of the process. Phase two will be forking SporeDB to make it technically far more robust, SporeDB can remain functional down to only one non-byzantine node (in theory, according to the whitepaper), whereas only 12 of the top 19 witness nodes have to be compromised here and you can completely control the chain.

If you think that Calibrae is going to be a centralised platform, you obviously haven't read enough about what is being proposed to really have a valid basis to comment.

There is more than a few things in the design of Steem that make it so liable to become centralised, the witness election, with its multiplicative 30 votes per account (each vote counts with 100% of stake), combined with the huge stakes that the developers and a little band of outsiders have built, and this is also why the platform is constantly wracked with paranoia, distrust, and anger. The developers don't have quite as much control as they would like, but they are fighting with the outsiders.

Have a look at the witness leaderboard now. For the first time since the beginning, furion and smooth are out of the top 19. I would expect that this is for a reason, as Smooth caused a big problem that led to HF19, and furion just let the cat out of the bag about how incredibly resource hungry steem RPC nodes are. By my commenting on the proposal post about how they are going to multithread the RPC function, I won myself some more flags, even though I am right - it is now impossible for anyone without a big budget to run a Steem RPC node, and we are all expected to trust and rely on Gandalf or Steemit's RPC nodes. Some decentralised platform... not even 15 months old and it's impossible to run your own RPC to develop apps unless you got hundreds of dollars a month to run the thing, or can afford to spend over $3000 on a server with over 128Gb (and growing requirements) of memory.

The guys at the top have no concept of how far out of touch they have got with the userbase, and you have no idea just how far gone this thing is. I give the platform 3 months at best. The only way they can save themselves from failure due to engineering error, is to plough probably at least $200,000 bux into development and paying a lot of coders. They are probably now doing that, but I'm abandoning a platform that gives lip service to decentralisation.

It was vindication to have furion confirm that I was not incompetent, but rather that the platform has grown so humongous that my 50gb memory 10 core SSD VPS cannot replay the chain. I have had just such a VPS running the replay for over a week now, and it's still nowhere near to finished.