NIP-88, Polls on Nostr #1346
NIP-88, Polls on Nostr #1346
Conversation
abhay-raizada
force-pushed
the
nostr-polls
branch
2 times, most recently
from
b78802e to
76bc670
Compare
abhay-raizada
force-pushed
the
nostr-polls
branch
from
76bc670 to
21587e2
Compare
118.md
Outdated
|
|
||
| ### Poll Content | ||
|
|
||
| The poll event is defined as a `kind:1068` event. Major tags in the poll event are: |
There was a problem hiding this comment.
I think defining the new kind is a good point.
| The poll event is defined as a `kind:1068` event. Major tags in the poll event are: | ||
|
|
||
| - **label**: The accompanying text of the poll. | ||
| - **option**: The option tags contain an OptionId(any alphanumeric) field, followed by an option label field. |
There was a problem hiding this comment.
I don't understand why you need the OptionId. The label can be the ID.
There was a problem hiding this comment.
what if there are two labels with the same name? i think this is an easy way to get past that. I don't like the idea of tallying same labels together.
There was a problem hiding this comment.
I feel like a label with the same name but different meaning is not a practical option.
There was a problem hiding this comment.
I agree with @AsaiToshiya. If clients are only showing the name and not the option ID, the same name should mean the same vote. Not having an ID removes the need to disambiguate names when they duplicate.
There was a problem hiding this comment.
I disagree, the client is not doing anything in this special case.
Its the poll creator that is just having fun, maybe they want to see how the option order affects the poll results, which they should be allowed to do.
There was a problem hiding this comment.
I agree with abhay-raizada. I might make a poll that says: Who will you vote for? 1) Trump, 2) Trump, 3) Trump
and it should collect the three answers separately.
|
Does this have 2 implementations? I was just about to comment on it:
Not using a replaceable event for the response is especially bad if the response is meant to be edited. |
|
I will actually implement polls. I have a UI for it and could do it in a couple hours. But counting votes is going to be such a pain because we don't use existing replaceable semantics. |
I asked exactly these two questions myself, but I decided it wasn't a big deal and not worth changing the way Pollerama has been doing for months and is quite good already. We could definitely change it if @abhay-raizada agrees. |
It will cause issues with expiration, this way we can discard any responses after an expiration time. (NIP-40). Although someone could just send in lower created_at values. But at-least this method keeps open the scope for expiration, where as there's no scope with parameterized replaceable events. |
| "pubkey": "dee45a23c4f1d93f3a2043650c5081e4ac14a778e0acbef03de3768e4f81ac7b", | ||
| "sig": "7fa93bf3c430eaef784b0dacc217d3cd5eff1c520e7ef5d961381bc0f014dde6286618048d924808e54d1be03f2f2c2f0f8b5c9c2082a4480caf45a565ca9797", | ||
| "tags": [ | ||
| ["label", "Pineapple on Pizza"], |
There was a problem hiding this comment.
implemented NIP-88 on getwired.app
super easy to implement and works well. don't really see the point of the label tag on the poll event, but other than that no complaints
There was a problem hiding this comment.
it's an artifact from before the label was moved to the content field. Will update
This NIP takes the ideas evolved from
and tries to combine them in a spec for a simple polls implementation.
Working version: https://pollerama.fun/
cc: @staab @vitorpamplona @vicariousdrama @alexgleason @gocedoko @fiatjaf