Topic: No BIPs made since 2021-06-27 (Read 234 times)

I mean if I had to guess I would say bitcoin is pretty disproportionate in wallet sizes and not much really moving in blockchains just a guess though

LN is entirely separate from bitcoin core. It uses existing consensus rules to implement a layer 2 protocol. Further, the software is entirely separate.

There are a number of BIPs that made LN possible. The BIP related to SW is a major one that comes to mind. There are others going back years before the concept of LN was ever made public. The BIP related to nTineLock is one example.

Apparently one of them is Luke dashjr, because he gave me a full stop on my BIP with a recommendation to improve BIP322 instead.

I'll have more to say about this once I open the new discussion thread for it.

They are as transparent as possible. If you have major changes like Segwit or Taproot, then you can see them publicly. But as the whole protocol is more and more stable, next features could be implemented even without touching the protocol. Also note that the way Bitcoin was created, was never 100% public. You can get version 0.1.0, but you cannot get versions that existed before, they were shared only with some people, if they were lucky enough to ask Satoshi directly. And the same is true with many other features. For example, P2P communication require no consensus changes at all, things like that can be handled in a fully backward-compatible way. And there are many things, where no consensus changes are needed. For example, Silent Payments are publicly discussed, but they require no consensus changes, those things will probably be purely informational. The same with HD wallets, no consensus changes are needed. The same with signatures for different address types, again, it will be purely informational. And you will see more such changes.

Yes. Many things are prepared upfront, tested, planned, carefully designed, and then shared, when authors are convinced enough, that their ideas are unstoppable enough, and they are worth sharing in the first place. Coding and testing takes time. And if every single commit would be shared, then some ideas could be misused. For example, imagine that Bitcoin could use SHA-1 as a hash function, or could use RSA keys for handling public key cryptography. Or imagine that we could have a lot of transaction types with no Script, so there would be "an explosion of special cases", for example we could be forced to use OP_CHECKMULTISIG, even if there is only one public key.

Also note that there are many existing things with no BIPs. For example, where are BIPs for the Lightning Network?

I agree, but the question we may ask is: if we have "public"/"community" project, however you want to call it, should development decisions be transparent?
Because what was said we may imagine that github (or other repository) is just a place for end-user to download the ready program. All the development process is then done behind closed door with "me and my friends".

Taking the part I bolded I will say that for a lot of projects now that is true. 10+ years ago it might have been different. But with so many other methods of real-time digital communication now being used by so many people (telegram / discord / etc.) that for a lot of the preliminary and even mid-term programing work it is easier to do it using those services then github. Bitcoin uses the bitcoin-dev mailing list, but I know of many (non crypto) projects that by the time it's on what we see as the github for the project it probably spent several months on the programmers private^* git space while being discussed in 3 discord channels.

-Dave

^* It's not private like you can't see it, it's private like if you don't know to look or ask where it is you probably will not see it.

Development of bitcoin is not driven by programmers' sense of appeal, neither it is going to halt because of superstitious feelings of ignorant users. Ideas and new technologies emerge, deficiencies and shortcomings appear, there is a chance for subsets of each group to match, development happens. The match may be accidental, or simply because that one group has triggered the other. Sometimes new technology is perceived by users as something important, even necessary to ask for, sometimes the need for fixing problems solely lead to research and development.

It is a distributed system, resistant to single/limited points of failure, unlike what conservative guardians are advocating, there is little chance for an update to cause fatal damage, with very few preliminary measures taken cautiously, it never happens, ZERO.
Otherwise, what would be the advantage of having bitcoin replicating everything?

Pro-tip (copy catting Ethereum's lead scammer now ):
Upgrade to software engineering and architecture, programming is the most trivial part of the deal, the part that tilts your head down, we need it to be up with our eyes wide.

The threshold to submit a PR is low, so some of them may not be seriously considered, and/or are a low priority.

The development of Bitcoin is different from most other software because no one entity is responsible for it, and also because there is so much commerce that involves bitcoin. The later gives an incentive for many to try to influence the development of bitcoin for various reasons. As achow101 said above, much of the work regarding BIPs is done outside of github.

It is not about now being submitted - there are BIPs not merged. There is quite huge list of pull requests: https://github.com/bitcoin/bips/pulls

I do not know who is responsible for processing that requests, but indeed, it looks like a quite big backlog.

I know we have different opinions on this, but I actually find the news that development may have slowed down to be quite calming [1].

As a programmer, sure, I can see the appeal of constantly trying to improve the technology.

But as a user, I'd be much happier if the technology behind Bitcoin basically stopped being improved [2].

It's a fringe and depressing view, but unless Terence Tao is reviewing BIPs, then it's only a matter of time before some serious security flaw sneaks into production (by way of some "improvement") and (maybe only months or years later) causes havoc.

[1] At least, I did until @achow101 reassured everyone that it's not true.

[2] Excluding things like making Bitcoin "post-quantum" which would be negligent to ignore.

First of all, BIPs are not numbered in order; while BIP 386 is the highest number BIP, it is not necessarily the most recently added - that would be BIP 326. The dates listed for each BIP is also not the date it was added, these dates are often the date of some draft (often the first draft).

BIPs are often discussed on the bitcoin-dev mailing list prior to being added to the BIPs repo. Off the top of my head, there are at least three BIPs in draft on the mailing list - MuSig2, Package Relay, and Multipath Descriptors.

I don't think there's any "loss of interest". Spec work and protocol development takes time, and a lot of low hanging fruit has been picked off, so I think it's expected that there are times that fewer BIPs are proposed, and that fewer BIPs would be proposed over time.

Also many of the BIPs that are currently in the repo are not final. They are still being worked on, with wording changes, test cases, semi-major overhauls, etc. constantly going on. So it's not as if people are not interested in BIPs, there's still a lot of interest and people are still working on them.

I find it quite disturbing that there hasn't been a single new BIP submitted to the github repository at https://github.com/bitcoin/bips, ever since 2021-06-27 [a bunch of BIPs about descriptors were made at that time]. What happened?

Have the bulk of the people lost interest in this kind of stuff?