headless server names should not change

[titi]

Currently we have good and not so good headless servers. When people join one of these servers they overwrite the original name of the server with something like "titi controls".
I vote to not rename headless servers, because by this players can see what server it is for real. Most people know which server works for them and which makes trouble and by this we can give them a chance to decide.

And the headless servers should use the name to provide some information too like:

myServ 1CORE 512MB
titis 4CORE 8GB
...

[tomreyn]

Currently we have good and not so good headless servers. When people join one of these servers they overwrite the original name of the server with something like "titi controls".
I vote to not rename headless servers, because by this players can see what server it is for real. Most people know which server works for them and which makes trouble and by this we can give them a chance to decide.

I concur. While it may have seemed nice to point out who is controlling and connected, this information should be transferred separately and does not belong into this field. Currently, it replaces information (the server name) which is relevant in itself, and does so semantically wrong.

And the headless servers should use the name to provide some information too like:

myServ 1CORE 512MB
titis 4CORE 8GB
...

I strongly disagree. This information does not belong into the server name / title field either.

I do agree that it may be nice to provide such information, though. If we do, we will need to make sure this information is of any use. A server with 8 GB physical RAM and 4 cores can be much slower than a single core server with 1 GB RAM if the former also runs a bitcoin miner and two build VMs and a web and database server. So how to tell what's slow / fast? On POSIX systems the 1/5/15m average server load might help, but this won't work on Windows (unless someone wrote a compatible implementation). At the same time, the architecture can make quite a difference, too, think of Intel i7 vs. some ARM system (of which we will see more soon, especially for VPS / cheap dedicated servers I think).

[titi]

I see no point in seeing who is controlling the server. At least I think this is no important information.
About the hardware thats true, but our space in the server list is really limited.

[Omega]

I don't think it's a good idea to just display things like RAM and the number of cores. There's too many factors influencing performance (as Tomreyn pointed out).

Instead, why not show a number from some benchmark? I'm not sure what benchmark you'd want to use, though.

[tomreyn]

I see no point in seeing who is controlling the server. At least I think this is no important information.
About the hardware thats true, but our space in the server list is really limited.

Limited space in web design is a secondary matter, this is just about how we present the information recorded to the database. In the first step, we need to ensure we record every piece of information we gather into separate storage areas, so database columns. Then for presentation we can decide to display two or more database columns into one table column if we wish to do so. And someone else can poll the CSV or JSON output and create yet another presentation from it, assembling the various components in a different way.

[Omega]

A side note that we could always put the specs/benchmark/whatever in a tooltip.

I don't like the current tooltips, and would like to see a dedicated documentation page for the masterserver API, which should make the current tooltips obsolete.

[tomreyn]

I concur with Omega's statements.
Also using CSS layers, a smaller font and lesser margins are options to stuff more info into the presentation.

[titi]

OK, back to topic. Initial topic was not about giving more info about the headless servers hardware/performance or html layout problems in first place. It was just about replacing the info who controlls the server with the original info which server it is, just to show to the player to witch server he connects. If you drop out from one server 3 times in a row you know that this server is not good for you and you don't want to connect again. But if you cannot see its name its hard. And who controlls the server does not give much important info. The field is called servername so the "who controlls info" isn't correct there anyway.

[Coldfusionstorm]

I agree, Giving the server a distint server name is a good idea. instead of who controls it.

[tomreyn]

It seems those who took part in this discussion so far are in agreement that the server title should not be overwritten by other information.

I do think the fact that player X is currently controlling server Y (be it a headless server or not) would be good to preserve, probably in the form of a "controlling" (or similar) SQL column which just stores a boolean. That's because other players may not want to connect to a server which is controlled by a player they know tends to create game setups they do not like, but might prefer to connect to another server then.