1. 4

I’m wondering what people think of a decentralized tag. I’m not sure that the current tags adequately categorize these kinds of links. The closest thing I can think of is “distributed”, but lots of software is distributed but not decentralized.


  2. 7

    I would rather add the word ‘decentralized’ or ‘federated’ to the description of the ‘distributed’ tag and call it good.

    1. 5

      I think that might be going too far.

      1. 4

        A good way to make this case would be to come up with a list of existing posts that should have the tag.

          1. 2

            The distributed tag would seem appropriate for almost all of them, no?

            1. 1

              Sure, I guess it just seems to me that distributed might be overly general. In the same vein, most of those could be tagged “programming” too but that doesn’t really capture what’s interesting about this stories (in my opinion). When I hear the word “distributed” the first things that pop into my head are architectures where you generally trust your servers and maybe you want to account for things like failure, maybe you’re interested in stuff like master/slave replication, consensus algorithms, or whatever. That’s not say decentralized stuff isn’t distributed but what’s interesting about decentralized stuff is that you generally have to deal with trust (like byzantine generals) on the one side, and on the other generally there is some democratizing aspect to the technology (low barrier to participation, etc) on the other. Both of these things in my mind make decentralized stuff kind of categorically different from the more common concerns of distributed computing.

        1. 2

          Honestly, the current set of tags is already getting too large for me. I end up having to scroll through all of them everytime anyways because I’m never exactly sure what the creator of the tag decided to name the concept I have in my head.

          1. 1

            What tags do you feel are misnamed?

            1. 3

              It’s not that I think anything is misnamed, but rather that there are multiple names something could have and I don’t know what ones the creator chose.

              The language tags are the most inconsistent to me. We have haskell, elixir, erlang, c, go, etc. All languages. But then sometimes we have ml (a class of language), and dotnet (a platform for languages that posts related to those languages go under). There doesn’t not exist an FP or imperative programming tag incase the post is functional-but-not-language-specific. And then you have all of the posts that are needed before a tag for a language gets created so now searching by that language tag doens’t mean you’ll find all the posts related to it. I understand why it is this way, and I don’t necessarily know of a superior solution that is also so simple, but it basically means over the course of a few years the tag set explodes with entries that mostly don’t matter.

              1. 1

                I kind-of feel like the tag UI is missing an opportunity here… something that presents relationships between tags, like “sibling tag”, or “parent tag”, or even “frequently tagged with”.

                Also, “ml” is an interesting example because there’s a conflicting use, “machine learning”. If there were a tag subsumption graph, the tag-entry UI could autocomplete “ml” to something like “technology > programming languages > ml”, but that would also use a lot of screen space and might not make a lot of sense if it wound up making the UI clumsy. I also haven’t seen a great deal of mis-tagging. Incomplete tagging is the immediate problem.

                Hierarchical tags are also a fair amount of work on the back-end, which I’m not offering to do. :) So I’m just mentioning this in case anyone else wants to run with it.