Equilibrium for Honours

Rispok

Seems like this is going to be a hot-button issue for a while, unless Tecton immediately redacts. If you're a coder or divine, feel free to enlighten the unwashed masses.

Vadimuses

When I was introducing NameDB in 2013, I gave Tecton a heads up that it was coming, so it wasn't a complete surprise to the game admin. He did voice a concern about Achaea's performance, but it turned out to be okay, and worked just fine for the masses for Achaea.

Just recently, I gave him another heads up that I was planning to make it quicker - as, you know, it was working fine for two years in production. He came back with a suggestion to use a method that was far slower than the current one - which didn't make a whole lot of sense, as Achaea wasn't experiencing performance issues already with the current setup.

Then, this happened.

I'm quite fine with a server admin being cautious for their game's performance, but in this case, it seems to either be a knee-jerk reaction, or if there really was performance issues, bad monitoring for two years that has not discovered it.

If Achaea starts being massively faster for everyone then this change is warranted - but I really doubt it will be. All this has done is break existing functionality that makes Achaea nicer to play.

Klendathu

My guess is that it's a counter for people using Svo / ndb where their system is spamming honours when there's a new person on qw.

Tahquil

I thought maybe it's got to do with the introduction of server side everything. I hell of a lot more stuff is going to be injected/stored (even if for a few seconds) on the server

Sanek

I have no idea of the load of the honours speed on the game's performance, but how it was without eq cost at the moment seems to have no impact. Maybe speeding it up with the new system would have a negative impact, and should not happen, that is not something I can judge.

This equilibrium cost really comes over as @Tecton slapping down @Vadimuses for informing him about something, and from the info in this thread, it seems like it could've been handled a lot better.

Moral of the story: Next time, ninja change, don't inform? Don't think that is the attitude that is desired.

Rangor

I wasn't using auto honors for a while, and you'll quickly realize that you will manually honors the people you need to honors and leave everyone else grey. It's not a biggie.

Rispok

As a point of interest, I just checked to see how long it would take for me to NDB UPDATE, and it's just over 13 hours, considering I have slightly over 20,000 entries.

Jonathin

I hardly use ndb anyways. It's more of an inconvenience for me than anything

Lorielan

@Rispok, you do realise that before this change, that 20,000 entries would check within about 10 minutes? And you are just one player.

Ruth

Does repeated honors'ing lag the game or something?

Merador

This poll seems entirely pointless, to me. Of course people would rather have no balance loss using honours if they were answering selfishly. The important consideration is whether it's for the good of the game, and that's not something we can answer at all.

Avto

Well, there was a reason why people were using ndb checks, if you are removing that aspect due to performance issues, you could firstly set in basic serverside ndb type commands. Like color flagging different citizens, this is already in place (CTF), qwho check with list ordered per cities etc..

ndb was in game for years, before vadi created mass ndb script, people were using other scripts (be it not as complex as vadi's).

Mishgul

i mean, the first question to ask is, can we verify if there is a serverside ndb in the works? I mean I have my own personal one, but it organises by non mhaldorian, shoot on sight and asked questions then seem to be pretty okay people which is simple to program and doesn't require honours, so this change doesn't really bother me, and I feel like it was done with some good for the game plan in mind.

Sarapis

Ruth said:

Does repeated honors'ing lag the game or something?

All commands 'lag' the game, as when one command is being processed, all others are held in a queue until that one is done (for the technical folks out there, we're a single-threaded application). Most of the time, there's more than enough processing time to make that essentially irrelevant, but when people spam the server with many commands in a short amount of time (sending 20,000 commands in a short time is the very definition of spamming the server) it causes lag hiccups. 

Sarapis

Vadimuses said:

If Achaea starts being massively even the tiniest bit faster for everyone then this change is warranted - but I really doubt it will be. All this has done is break existing functionality that makes Achaea nicer to play.

FTFY!

Sarapis

Name highlighting on the server isn't going to happen for reasons already explained in detail in other threads. Viewer-relative colouring would be a huge performance impact of the type that gets very bad when a lot of people are in the same room.

Austere

Sarapis said:

Name highlighting on the server isn't going to happen for reasons already explained in detail in other threads. Viewer-relative colouring would be a huge performance impact of the type that gets very bad when a lot of people are in the same room.

Does this mean infamywho is out as well?  I swear,  poor Vadi has had the idea wrote down since ndb came out and he JUST released it. . Don't do us like that. .

Jacen

Could you provide a QW sorted by city then? And a QW that lists soldiers for each city?

Aerek

I've honestly never liked the datamining tendencies that really started catching on few years ago. Parsing WHO for target locations still feels like cheating to me, (never understood why WHO needed to show room location with thirdeye in the first place) and instantly knowing the infamy/Mark status of everyone online so you can build a gank list always felt like something that couldn't be stopped, but didn't need to be encouraged.

So yeah, if the game performance can be improved even slightly by removing the ability to effortlessly keep tabs on everyone online at all times, I'm not going to shed a tear.

Merador

Sarapis said:

Name highlighting on the server isn't going to happen for reasons already explained in detail in other threads. Viewer-relative colouring would be a huge performance impact of the type that gets very bad when a lot of people are in the same room.

Is that still the case if colours aren't customisable? For example, capture the flags have consistent colours for cities that are the same for everyone. Would this sort of colouring where Mhaldorian players are always red still be an issue? 

Sarapis

Jacen said:

Could you provide a QW sorted by city then? And a QW that lists soldiers for each city?

Maybe. 

Trilliana

something akin to this?

  rogue     (15): Amranu, Artanis, Bemanigisc, Bluef, Ermot, Jarrod, Juree, Kadrim, Killian, Makaela, Pharue, Saramy, Shirszae, Suzzette and Taarkos
  Cyrene    (14): Adaele, Asmodan, Doina, Edena, Farran, Haeley, Halgar, Ilen, Keorin, Mazel, Morganbarl, Rashmi, Sanek and Shianne
  Mhaldor   (13): Carmain, Chiam, Crixos, Elazar, Etias, Jaybles, Kitiara, Kray, Locki, Multon, Natashia, Orzaansyn and Xinna
  Ashtan    (12): Alarke, Arralana, Baliuk, Foehn, Hexan, Hydrogen, Jonesey, Khaibit, Merador, Seraya, Vyn and Zadie
  Targossas (10): Anedhel, Azelhu, Daitya, Invictus, Kaiu, Kellonius, Lianice, Mel, Shunsui and Szareine
  Eleusis   (9): Aeceon, Ajo, Kriphi, Myna, Seftin, Slink, Sylarana, Synnendir and Zakhar
  Hashan    (3): Andregor, Rekhyr and Torrent

Sarapis

Merador said:

Sarapis said:

Name highlighting on the server isn't going to happen for reasons already explained in detail in other threads. Viewer-relative colouring would be a huge performance impact of the type that gets very bad when a lot of people are in the same room.

Is that still the case if colours aren't customisable? For example, capture the flags have consistent colours for cities that are the same for everyone. Would this sort of colouring where Mhaldorian players are always red still be an issue? 

Well, they still have to be customisable in the sense of being on or off via a config option, which CTF isn't. It'd be the simplest version of viewer-relative coloring. In this case, it's more of a huge pain in the ass to implement than it would be a major performance issue. Since our name generation routines aren't viewer-relative, we'd need to go through the codebase and alter every single call to a name generation routine and, in some cases, minorly rewrite how entire routines work. I don't have the code in front of me, but I'm sure there are multiple tens of thousands of calls to them. In terms of reward vs. effort, that's not going to be a high priority.

Jacen

Aerek said:

I've honestly never liked the datamining tendencies that really started catching on few years ago. Parsing WHO for target locations still feels like cheating to me, (never understood why WHO needed to show room location with thirdeye in the first place) and instantly knowing the infamy/Mark status of everyone online so you can build a gank list always felt like something that couldn't be stopped, but didn't need to be encouraged.

So yeah, if the game performance can be improved even slightly by removing the ability to effortlessly keep tabs on everyone online at all times, I'm not going to shed a tear.

Its a very nice feature for raid leaders though. Being able to see how many people in a city are online can be very influential in your decision to raid a city, whether for the good or bad. Moreover, you can check their WHO list locations to see if they're all grouped up at a non-gatehouse/crossroads location, to see if maybe they have some event going on.

Jovolo

You could do it old school (Cain, Rukimoro et al) and just roll in. Then leeroy jenkins out when things hit the fan

Leaving the rest of your group behind of course

Sarapis

Sarapis said:

Jacen said:

Could you provide a QW sorted by city then? And a QW that lists soldiers for each city?

Maybe. 

Switched to yes for qw sorted by city.

Bean

Seems like a better solution would be either a balance specifically for honors, or a limit of up to 10 within 10 minutes with no cost or limit at all - both of which would completely address the performance concerns without negatively impacting other aspects of the game.

It definitely shouldn't affect your characters actual balance - this just seems like overkill. If I want to to honors my opponent in a fight, I shouldn't have to completely stop fighting for any period of time.  It also breaks the in-game/OOC barrier as equilibrium is an in-character concept, and the honors interface is entirely OOC.  Same logic that was used for sending messages, IIRC.

Sarapis

QW <city/rogue> is in btw.

Kellonius

Why equilibrium though? Why not its own timer? I dont even care about NDB that much, but I do often honours people on the fly, like in a group fight, and being off eq has certain implications for that. Also its forceable.

Shirszae

I suspect having its own timer would not be big enough of a deterrent for people not to queue or spam it continuously

Kresslack

Aerek said:

I've honestly never liked the datamining tendencies that really started catching on few years ago. Parsing WHO for target locations still feels like cheating to me, (never understood why WHO needed to show room location with thirdeye in the first place)

It's a workaround for those who don't have skills which reveal a player's location, such as Occultist, Magi, Blademaster, etc, so we can still maintain situational awareness.