CWz wrote:I think that the clip public server should be excluded from the public server list due to being anarchy server, it gives minetest a bad image and a lot of young players will hate minetest because of how bad things are there. it is probably the cause of most the griefing and misbehaving players on the servers.
I don’t know if your claims about the server are true. But it does not matter for now. I am strongly against delisting any server solely for its policy, or in this case a lack thereof.
Of course one is free to delist any server one wants to but you go a dangerous road of quasi-censorship. Yeah, you can’t
prevent that people visit servers like that, but you propose to make it just
inconvenient. This decision is authotarian because indirectly you are softly forcing servers to have some anti-griefing policy. Because you wish to have servers without that policy to delist. This makes these servers much less visible and you are effectively punishing them. But punishing servers, no matter which justification, would result in an even more biased list. This thread is about how to
get rid of the bias, not to introduce more bias.
And I don’t care about Minetest’s image. If people think Minetest is the griefer game just because of one server, they are being irrational. If people don’t want to play Minetest, I am completely fine with that. Popularity should not be a main goal (if at all) of Minetest. For me, the main goal of Minetest is to make a game / game engine which works.
I have a compromise suggestion: Do not exclude servers from the list itself. Instead, servers could be able to publish some additional standarized flags about which policies they have or the lack thereof. Some of these flags could be:
- Don’t grief
- Don’t use nerd-poling (funny, I actually saw a server with that rule)
- Don’t <do whatever>
- No rules (of course this flag makes no sense when other policy flags are used)
If there are no policy flags, it does not neccessarily mean there are no policies. It just means the server does not say weather or not there are policies.
With such flags, everyone can easily filter on their own needs. It is important that by default, no filtering should be applied, anything else would still result in a biased list, which would be morally questionable. And you could of course filter all those “anarchist” servers if you don’t like them.
Besides: Welcome to the Internet! The place where kids, anarchists, terrorists, fuckwads, trolls, cat-lovers, mothers and programmers (sometimes all in one person!) meet. ;-)
Besides 2: Generally I also think you are very rude by suggesting to exclude a server from the list without at least talking to or just even informing the server operators or the server users. Even if you have something against the server, you should have the guts to express yourself in
the server’s thread instead of just attepting to “backstab” it.
------------- (imagine a long horizontal line here) -------------
Back to topic: I agree that the current system is biased towards popular servers. This is bad. Popular servers are likely to be rewared by even more popularity, just because they have a good list place.
So a reform is a good idea.
But I am not very happy with the initial proposal either. The poll did not consider that opinion, so it is a poor poll and won’t help with the issues.
Vazon wrote:Weighted votes, age, lag, and alpha
I don’t think votes would help, especially I don’t think
weighted votes would help. This is likely to put even more bias into the list, and, if the votes are also weighted, some influential voters could basically “tailor” the list as they wish. The bias would be perfect. I am also skeptical of an equal votes (aka 1 player = 1 vote) system. Other players have other needs and wishes for servers and so they’d rate them based on different criteria. I doubt how the average of those ratings, which are all based on a lot of different things, is going to help ME to find the servers I want.
I am unsure about age. How is this defined? If a server X steadily run for 10 years, then the server was turned off (intentionally or not) and returns after a hour or so, it the age resetted to 0 again? If so, then age alone would be a very poor criterion. A better criterion would be the uptime/downtime ration plus the time it last started
Lag (aka latency) is very important to know and definetely should be visible to the client.
I have no idea what you mean with “alpha”.
One comment about age:
What about letting the server publish planned downtimes for the server list. This would be incredibly useful for servers which are not up 24/7 but are still visible in the list so they can say “We are down now, but we will be up again on 8 ’o clock”. Maybe the public list should delist servers which are down for more than 7 days or so, because what would be the point of a server if it is down for so long?
I think the raw number of connected players should still be available as a data point. It is simply interesting to know.
Generally, I think the in-game list should expose much more information about the servers. Especially the latency. It would be a helpful feature to filter out servers based on user-selected criteria.
To make the server list as neutral as possible, I suggest to initially show the servers either in random order (full neutrality) or sorted by latency (not neutral, but I justify this by latency being technically very important; high latency servers are unplayable and deserve less attention). And that no filters are applied by default, but the user can enable them.
Just my 2 cents on the Public Server List Reform. :)