• 목록
• 답변
• 글쓰기
• 게시판 리스트 옵션
  • 수정
  • 삭제

I'm glad that the terms "message passing" and "shared heap" appear to have caught on on the subject of analyzing the technical differences in strategy between these techniques. But for me personally, that is a strange set of decisions in terms of "non-profits/communities can host giant infrastructure!" Well sure, but not as a result of they do not value loads. With that in place, it's time to analyze the "message passing" structure vs the "shared heap" structure in terms of how they perform when scaling. I believe (I may very well be incorrect) that "shared heap" is a time period I launched right here, though I did not really state that I was doing so. Skip forward if you do not care about this kind of thing, but I think there are interesting things to say about the discourse and meta-discourse of expertise analysis each typically and the way it has played out right here. But I nonetheless suppose "credible exit" remains a "invaluable worth" for Bluesky.

So I tried to be empathetic, however I nonetheless did not want to hold back on technical critique. I still do truly imagine that "credible exit" is a worthy goal. It is my opinion that leaning into "credible exit" is the smartest thing that Bluesky can do: maybe a large corporation or two all the time have to take a seat at the middle of Bluesky, however perhaps also it will likely be attainable for individuals to go away. People had been asking me what I thought of ATProto and i stated on a public fediverse thread that I had been "holding my tongue". Traditionally when traders realize users are doing this and eradicating a income stream, that is the point at which they start pressuring arduous on enshittification and eradicating issues like public access to APIs, etc. What's going to occur in Bluesky's case? Bluesky now has much bigger pressures than decentralization, particularly to fulfill the large scale of users who wish to flock to the platform now, to satisfy traders which can increasingly be excited by whether or not or not they will see a return, and to attain sufficient income to keep their workers and servers going.

Rearchitecting in direction of meaningful decentralization will be a big pivot and will possible introduce a lot of the problems that Bluesky has touted their platform as not having that other decentralized platforms have. There may be an inherent linear cost to customers participating on the network, insofar as for n users, there will all the time be an O(n) price of operation. The cost of working a full-network, totally archiving relay has elevated over time. User control of personal data: Some users wish to have autonomous management over their own social data, and share their knowledge selectively throughout various systems. And, fact be informed, individuals perhaps mostly read the social media threads that summarized it in bitesized chunks, but there have been so lots of those bitesized chunks. If you are going to learn any part of this writeup, if you're going to quote any section, this one is the necessary one. If you want to see more pointed, specific responses, leap forward once more to the subsequent section. Anything you see quoted in this part comes straight from Bryan's article, so take that as implicit. However, the very subsequent part is titled "Centralization Could be Helpful"!

The ability to modify between completely different operate providers is a core mechanism to manage centralization. Bryan Newbold, core Bluesky engineer, wrote a thoughtful response article: Reply on Bluesky and Decentralization, what is control cable which I consider price reading. So I guess it's price seeing that this is an instance where that is definitively not the case. This will straightforward to get lost about; the instance above of stating that "gossip" can enhance things indicates that speaking about message sending is complicated the matter. In different phrases, as we make our programs more decentralized, message passing systems handle issues fairly effective. While Mark weighs each, the doc reads as a one that authors standards doc who would like the internet to be extra decentralized the place it's potential, but in addition operates from the "pragmatic" perspective that issues are going to re-centralize most of the time anyway, and after they do this finally tends to be useful. Much speak of expertise tends to treat said tech to be as socially neutral as a hammer.