Bitcoin Core removed the configuration option to disable RBF in v29. That’s a spam filter.
They also tried to remove the configuration option for datacarrier in v30, backing down only after signifi...
i just discovered it has a timebomb in it also. after 7 november 2028 also, in 29, which knots contains. here's my conversation with deepsnek about it:
---
## User
ok but what happens when it is actually november 27 2028 or after? why should it expire?
---
## Assistant (Build · DeepSeek V4 Pro · 10.2s)
_Thinking:_
The user is questioning the purpose of the software expiry mechanism. They're asking: when the actual calendar date reaches November 2028, why should the software stop working?
The answer is that this is a "bomb" mechanism - it's designed to force operators to upgrade by a certain date. The idea is that if a critical consensus change happens before that date, operators who haven't upgraded would be on a different chain. The expiry forces them to upgrade, theoretically preventing them from running outdated consensus software.
But it's a terrible mechanism because:
1. It creates an artificial hard deadline where ALL nodes stop working simultaneously
2. If no actual consensus change happened (RDTS already activated in 2026), the expiry is just a pointless denial-of-service
3. An NTP attacker can trigger it early
4. It's paternalistic - "we know better than you when to upgrade"
The user seems to be questioning this too. Let me answer directly and concisely.
It shouldn't. The expiry is an artificial kill-switch with no technical justification.
If RDTS activates in September 2026 and is permanent (`active_duration = INT_MAX`), then by November 2028 there are no new consensus rules pending. The node would be fully in consensus, enforcing all active rules correctly. The expiry hard-rejects valid blocks for no reason other than a timestamp check.
The original mindset is paternalistic forced-upgrade: "we assume there will be new consensus rules by this date, so we'll force you to upgrade or die." But if no new rules materialize, it's purely a self-inflicted DoS with an NTP amplification surface.
The `-softwareexpiry=0` escape hatch exists but shouldn't be necessary. Expired nodes that are fully in consensus should just keep running.
---
## User
ok, change that task to removing that bullshit, every last bit of it
---
---
## User
ok but what happens when it is actually november 27 2028 or after? why should it expire?
---
## Assistant (Build · DeepSeek V4 Pro · 10.2s)
_Thinking:_
The user is questioning the purpose of the software expiry mechanism. They're asking: when the actual calendar date reaches November 2028, why should the software stop working?
The answer is that this is a "bomb" mechanism - it's designed to force operators to upgrade by a certain date. The idea is that if a critical consensus change happens before that date, operators who haven't upgraded would be on a different chain. The expiry forces them to upgrade, theoretically preventing them from running outdated consensus software.
But it's a terrible mechanism because:
1. It creates an artificial hard deadline where ALL nodes stop working simultaneously
2. If no actual consensus change happened (RDTS already activated in 2026), the expiry is just a pointless denial-of-service
3. An NTP attacker can trigger it early
4. It's paternalistic - "we know better than you when to upgrade"
The user seems to be questioning this too. Let me answer directly and concisely.
It shouldn't. The expiry is an artificial kill-switch with no technical justification.
If RDTS activates in September 2026 and is permanent (`active_duration = INT_MAX`), then by November 2028 there are no new consensus rules pending. The node would be fully in consensus, enforcing all active rules correctly. The expiry hard-rejects valid blocks for no reason other than a timestamp check.
The original mindset is paternalistic forced-upgrade: "we assume there will be new consensus rules by this date, so we'll force you to upgrade or die." But if no new rules materialize, it's purely a self-inflicted DoS with an NTP amplification surface.
The `-softwareexpiry=0` escape hatch exists but shouldn't be necessary. Expired nodes that are fully in consensus should just keep running.
---
## User
ok, change that task to removing that bullshit, every last bit of it
---
1