Instagram feeds

[vitorpamplona]

Adds a simple kind for picture-first clients building Instagram-like feeds, where the picture should be attenuated and the descriptions are less important and might not even be displayed in the main UI.

No, I don't think Kind 1s must be reused for these clients.

And no, NIP-94 became too generic for this.

Read here

[fiatjaf]

Looks good at first glance.

[vitorpamplona]

I am starting to wonder if NIP-71's "vertical videos" is too broad. They should be short vertical videos. No one wants to see a movie in these feeds, not even vertical movies.

[sepehr-safari]

NIPs like these are essential to move the ecosystem forward! Relying on a few broad NIPs just doesn't cover all advanced use cases. For example, with Pinstr, I had to create custom events that other clients don't recognize, which disrupts compatibility and limits potential growth for those kind of clients.

[vitorpamplona]

@sepehr-safari which kinds do you use? Would you switch to the new kind here or should we just use your kind?

[derekross]

I'm a fan. Text only kind 1 events are clearly different.

[sepehr-safari]

@sepehr-safari which kinds do you use?

Pinstr uses kind: 33889.

Would you switch to the new kind here or should we just use your kind?

I think none of the above! Pinstr needs more options than this NIP and it's not a replacement of Instagram.

[vitorpamplona]

ohh I see, so a pinstr only shows boards and each board is basically a list of links that can be sorted in any way the user sees fit. Cool. Yeah, they are different.

[frbitten]

Wouldn't it be simpler to have JSON data where there is the description and the imeta? After all, they are the content of the post. I think putting them in tags will complicate the parsing of this in the relay and in the clients.

[believethehype]

I think it's great and would be a great addition to galleries, replacing their content with these.

[flox1an]

Tagging with a geo location might be good. NIP-52 uses:

// Location
    ["location", "<location>"],
    ["g", "<geohash>"],

The annotate-user in the list of imeta-fields looked a bit strange to me, but I guess we need it at that level.