I read the thing, but that’s still no reason not to seperate nsfw category further. It makes no mention of a technical limitation that prevents creation of more categories, it’s just a person suggesting how a tag system could look like.

there he explained that it won’t work like that

The entire reason I’m commenting is because I don’t want it to work like that. I have no idea how to reply to a git commit, if I did I’d object there too.

1. CW and NSFW are too broad to be the only 2 types of tags.

2. The suggestion to use tag urls as ids don’t work with the idea of using ids to share tags across instances. I don’t want the entire threadiverse to decide lemmy.world/t/nsfw is the global nsfw tag and smaller instances being forced to use that or risk people facing unsolicited nsfw from their instance.

3. There is no mention of filtering posts based on their tag category, which would make it impossible to block nsfw posts across instances without using the shared id. There needs to be platform level tag categories (especially for things people might not want to see) so content filtering works in a consistent and decentralised manner. Then individual instances or communities can create their their own tags based on those categories to fit their needs.

4. The author seems to want to replace the spoiler mark with a (cw) tag called spoiler, but things can be both spoilers and fall under other tags (ie an image post, a leak, a discussion etc). Also this would pointlessly filter out regular spoiler content if a user blocks tags with the cw category.

5. The way I’d want things is tags and spoilers / content filters being completely seperate from each other at the platform level. Content warnings let people filter out content they don’t want to see. They aren’t a content discovery tool. And they should be nestable. There should be platform wide default ones like

• Nsfw *Artful nudity
  • Nudity
    • Porn
  • Shock
    • Injury
      • Gore
  • Slurs
• Spoiler

. Users would be able to choose whether to show, blur or block content with these warnings in the settings. Instances and communities would be able to make their own content warning as well, managed by the user in the settings of a community. A user creating a post would see a menu to pick the content warnings (multiple if they want) that describe their post. Tags describe content to people who want to find such content. There could be generic platform wide ones like news, qna etc, but the majority would be community specific. Tags could also include default content warnings (eg ‘leak’ tag could activate the spoiler content warning).

I think this would be the most elegant way to do it, but even what we have now feels better than what is proposed in that link