Can the spec relax/remove the requirement on minimal bit length? about vc-bitstring-status-list HOT 4 CLOSED

@thehenrytsai What's the use case you are trying to address? Although the 16KB limit is arbitrary, it provides a quantifiable floor for herd privacy.

The sense of the group is that far too many developers will just do the minimum or choose the default, so to empower users across the board, we need a floor to assure that the quick-to-market developer ships something reasonable.

It was claimed that compressing a 16KB bitstring typically reduces to a few hundred bytes (in scenarios with relatively few flipped bits).

If you have a use case that needs a smaller data size, we might be able to help find alternative solutions that maintain that herd privacy characteristic.

from vc-bitstring-status-list.

The issue was discussed in a meeting on 2023-09-15

• no resolutions were taken

from vc-bitstring-status-list.

@jandrieu, the desire at the time was to have a "nicer round" number so that it is arguably "friendlier" to end users when they read implementor's documentations and think about "orders or magnitude" of protection; it also makes it easier to have an accurate description on how the status list work when verbally describing the mechanism as opposed to giving an "approximate" number, this is something that I encountered at the time having to explain the mechanism to various people.

ie. "we aggregate 100,000 status into one list to a achieve herd-privacy..." instead of an seemingly odd arbitrary number: we aggregate 131,072 status into one list..."

But ultimately this is a nice to have IMO, not a deal breaker. Changing the default to 100,000 is also a viable alternative, a nice round number is "more meaningful" in this day of age IMO.

from vc-bitstring-status-list.

ie. "we aggregate 100,000 status into one list to a achieve herd-privacy..." instead of an seemingly odd arbitrary number: we aggregate 131,072 status into one list..."

You can always say "We aggregate roughly 100,000 statuses into one list." <-- just change the way you convey the information to be less precise. Only developers implementing will probably care about the details; they'll be the ones that look at the specification and see the 16 KiB figure.

I'm going to close the issue without a change to the specification since this problem can be addressed by just presenting the information differently to people that don't need to understand the details. Please re-open if you disagree.

from vc-bitstring-status-list.