I think what they decided on is a series of strategies.
They range from minimal work minor impact > minimal work major impact > maximum work major impact.
As they said they have to test on live...so they have to try lowest risk strategies 1st.
Obviously if they have to tear the engine down and rebuild it in the end..thats what they'll do...piecemeal
It appears to me that they have finally come to the conclusion that what must be done is large scale risky engine/client/server & netcode restructuring ...and tinkering simply wont cut it.

I'm impressed with the strategy they adopted a couple of months ago where they decided to go through everything with a fine tooth comb and leave no stone unturned.
The slow thorough methodical approach...ripping every packet and function apart as well as redesigning client/server offload.
The reports we have been seeing recently add proof to that strategy.

This comprehensive and slow approach is infuriating for us that just want to play a working game.
But at the end of the day, its the only way to guarantee the best fix that can possibly be achieved.
You cant fix a problem without analysing it first and gathering as much data as possible.
I really wouldn't be surprised if a lot of the +999 is down to live ninja testing/analysis/sniffing.

Does ZOS appreciate being hated on for pulling every aspect of the game to bits in an effort to fix PvP ?
Probably not. Its probably only their professionalism that keeps driving them forward.
Their secrecy on the process comes across as a lack of interest in PvP though.
Its really only @ZOS_BrianWheeler offering tidbits on whats going on that gives us any reprieve from the frustration.
Perhaps they feel the player base wouldn't understand the tech anyway, so no point spouting geek stuff.

It could have been easy for ZOS to just say screw it, it aint worth the hassle.
Indeed I would agree that they did try to ignore the problem as long as possible to get other content finished.
But where would that leave the ES franchise if they couldnt overcome problems ?
They have been in bad places before.
They didn't get where they are now by sticking their heads in the sand and running away from issues.

Don't be surprised if the game runs slicker than snot when finished now that they are taking the issue seriously.
[And I agree..there is no way they owned the code.
If they did they wouldn't need to investigate how everything works to fix it.
I think they are taking ownership now though]

I honestly enjoyed Matt's work in Warhammer Online, I hope the admins don't ban me for advertising even though the game no longer exist.


The PvP on Warhammer Online is what made me quit another game because it's PvP was so goodddd, ESO it's PvP is alright but it's unique since I've never played an MMO that didn't have cooldowns so that must be difficult to balance.

Rune_Relic wrote: »

I'm impressed with the strategy they adopted a couple of months ago where they decided to go through everything with a fine tooth comb and leave no stone unturned.
The slow thorough methodical approach...ripping every packet and function apart as well as redesigning client/server offload.
The reports we have been seeing recently add proof to that strategy.

This comprehensive and slow approach is infuriating for us that just want to play a working game.
But at the end of the day, its the only way to guarantee the best fix that can possibly be achieved.
You cant fix a problem without analysing it first and gathering as much data as possible.
I really wouldn't be surprised if a lot of the +999 is down to live ninja testing/analysis/sniffing.]

You seem to be very knowledgeable of Z. devs, coders and management actions. Could you please try to explain to someone who knows very little about coding what were those actions that you call "going through everything with a fine tooth comb and leave no stone unturned"?

Could you also explain what were their actions that were aimed at reducing problems without big changes in code?

And last thing - in what way ESO is different from other AAA titles that manage to handle those problems in time frame lesser than 1.5 year?

Septimus_Magna wrote: »

Now they've addressed this issues they cannot leave it in the current state thats it in.

They "address" the issue all the time.

There is a big difference between addressing an issue and doing anything noticable to that issue.


Concerning the proposed changes about to hit the PTS and the ones post that with the Thieths guild launch. I am optimistic that they may help.

Its rare to get much detail from Zos and when they do post such, then generally they know that it wont come back and bite them in the backside.

Cherryblossom wrote: »

Messy1 wrote: »

I've been critical of the game lately, but only because I enjoy playing it and want to see it get better, which has been my stance all along.

Anyway, I thought it was very commendable that Matt Firor in his Road Ahead outlined a plan to improve game performance in Cyrodil. Performance issues hinder the players enjoyment in PvP and it discourages other players from joining PvP.

So good work ZOS team and keep striving for excellence. Polish the game, make it better, and it will live long.

So after nearly 2 years they are starting to work on the Lag!

and adding content that should of been in at release! Woohoo!

I swear some players can be easily fooled. Wanting to and actually fixing a problem are too different things. In ZOS's case, they have had two years to fix the ever going problem in Cyrodiil and have done nothing. Sure they continue to say we are working on it which to me says they are stalling to get around the truth that with the servers they have at present they cannot.

In my opinion the only viable option to fix Cyrodiil would be for ZOS to add additional servers dedicated to just supporting the PvP environment. However, the cost of doing that type of upgrade would be enormous so I really don't think that will happen. Anyways stay tuned for more of "We Know There Is A Problem and We are Working On It"

But there are much simpler steps that can be taken. They could have done so much sooner to at least bring the game to a playable state.

How about some temporary "solutions" that might actually work?!!

Instead of removing some deer, why didn't they stop people being able to group in over the Population cap?!?

Why in the name of Jebus havent they reduced the numbers allowed into Azura's at any one time? This seems to be the most obvious. Sure, they want the combat to be of a very large scale, but there is no combat at all right now, just a slide show where you can be lucky if some spells register. It would also spread the population amongst those so called buff servers and give people more choice of where to play.

Anyway, IMO reducing the amount of ticks from abilities to reduce server load is an unproven concept. We could very easily be waiting another year and a half for any real difference.

I am waiting to renew my subscription thats been up since mid December, but only when I feel the steps being taken have a real chance at resolution. So far I am not seeing it.

@Septimus_Magna

I am actually saying the same as you.

Addressing an issues just means "talk about it", therefore Zos have addressed this issue multiple times over the last two years.

The road ahead articles are just a basic wish list. Nothing listed in one of these can be expected to apear, nor is it in anyway binding on Zos.

Like someone else said, dont expect anything unless its in the patch notes.


My personal view is that their code is so badly designed that nothing will fix it. Design ideas like a single zone (Cryodil) in which all calculations affect all parts of that zone and can therefore affect all players are just terrible and should never have left the whiteboard in the design stage. Games from 25 years ago didint make this mistake.....

The bits I find optimistic are those listed in the forum posts over the last couple of days. Only in that they gave us some real detail which makes it more likely that it is a "thing" rather than more whitewash.

They still only talk about optimising the mess we have now, and nothing that really points to a fix for lag and the terrible performance in Cryodil

Matt Firor has not earned even the slightest bit of respect. Especially not after the ESO live episode which just showed how it has always been with ZoS. Much blabla with no substance.

Tors wrote: »

My personal view is that their code is so badly designed that nothing will fix it. Design ideas like a single zone (Cryodil) in which all calculations affect all parts of that zone and can therefore affect all players are just terrible and should never have left the whiteboard in the design stage. Games from 25 years ago didint make this mistake.....

Twenty-five years ago, computers and networks were not advanced enough to be able to do a full graphical game zone with a Campaign sized server population in the same confined area. They could not have pulled it off. There are too many differences in computer and networking technology over the last 25 years to even make a reasonable comparison. This would apply only if ZOS was using servers built in 1990 instead of in 2014.



Are we at a place where we can have 1000 people standing on the Alessia bridge spamming AoE, today? No, I do not think so. Are we at a place where we can have 1000 players in the same Cyrodiil Campaign at the same time? Yes, I do think we are there.

As far as "all calculations affect all parts of that zone and can therefore affect all players" goes, that applies to all servers today. To one degree or another, you cannot do anything on the internet without being in a situation where your calculations could affect others using the same server. In ESO, there is a limit where, up to that point, you are unaware (in terms of players consuming server resources) that there are other people in the same Campaign. That limit is considerably higher today than it was even 10 years ago, but it is something that all online services have to deal with. ESO is no different from any MMO in that the actions of any one player, or group of players, can adversely impact all of the other players on the same server.

Is ZOS being ambitious in trying to have a very large number of AoE spamming players together in a relatively small space? I think they are and that it does come with the risk of failure. The extent by which ZOS is willing to succeed is something that remains to be seen. That is what will determine my respect.

NeillMcAttack wrote: »

Instead of removing some deer, why didn't they stop people being able to group in over the Population cap?!?

Why in the name of Jebus havent they reduced the numbers allowed into Azura's at any one time? This seems to be the most obvious. Sure, they want the combat to be of a very large scale, but there is no combat at all right now, just a slide show where you can be lucky if some spells register. It would also spread the population amongst those so called buff servers and give people more choice of where to play.

Well, I think the answer to the first question is more to do with not wanting to break up groups. The answer to the second question is that they have done that already, and the best guess is that they may have done it several times already. There was one point over the summer where one of the fixes improved Cyrodiil performance and they may have nudged the cap up in response. This performance gain may have been eaten up by more people partaking in the lag-inducing behavior over the last few months.

lordrichter wrote: »

Tors wrote: »

My personal view is that their code is so badly designed that nothing will fix it. Design ideas like a single zone (Cryodil) in which all calculations affect all parts of that zone and can therefore affect all players are just terrible and should never have left the whiteboard in the design stage. Games from 25 years ago didint make this mistake.....

Twenty-five years ago, computers and networks were not advanced enough to be able to do a full graphical game zone with a Campaign sized server population in the same confined area. They could not have pulled it off. There are too many differences in computer and networking technology over the last 25 years to even make a reasonable comparison. This would apply only if ZOS was using servers built in 1990 instead of in 2014.



Are we at a place where we can have 1000 people standing on the Alessia bridge spamming AoE, today? No, I do not think so. Are we at a place where we can have 1000 players in the same Cyrodiil Campaign at the same time? Yes, I do think we are there.

As far as "all calculations affect all parts of that zone and can therefore affect all players" goes, that applies to all servers today. To one degree or another, you cannot do anything on the internet without being in a situation where your calculations could affect others using the same server. In ESO, there is a limit where, up to that point, you are unaware (in terms of players consuming server resources) that there are other people in the same Campaign. That limit is considerably higher today than it was even 10 years ago, but it is something that all online services have to deal with. ESO is no different from any MMO in that the actions of any one player, or group of players, can adversely impact all of the other players on the same server.

Is ZOS being ambitious in trying to have a very large number of AoE spamming players together in a relatively small space? I think they are and that it does come with the risk of failure. The extent by which ZOS is willing to succeed is something that remains to be seen. That is what will determine my respect.

NeillMcAttack wrote: »

Instead of removing some deer, why didn't they stop people being able to group in over the Population cap?!?

Why in the name of Jebus havent they reduced the numbers allowed into Azura's at any one time? This seems to be the most obvious. Sure, they want the combat to be of a very large scale, but there is no combat at all right now, just a slide show where you can be lucky if some spells register. It would also spread the population amongst those so called buff servers and give people more choice of where to play.

Well, I think the answer to the first question is more to do with not wanting to break up groups. The answer to the second question is that they have done that already, and the best guess is that they may have done it several times already. There was one point over the summer where one of the fixes improved Cyrodiil performance and they may have nudged the cap up in response. This performance gain may have been eaten up by more people partaking in the lag-inducing behavior over the last few months.

Fair enough. Now let's compare it to it's contemporaries and see how it stacks up.

Septimus_Magna wrote: »

@Tors

The performance in Cyrodiil was pretty good up to 1.4 when they decided to shift client processes to the server side (to reduce farming bots I believe). From this point it has been going downhill performance wise.

@Septimus_Magna yup, again I agree. The anti-cheat code (or Lighting patch) definitely destroyed performance.

I would love some numbers on the amount of CPU's Zos gives to a campain, now and in the past. It would be interesting to cross reference that with the time that the game went F2P and see if they were forced to reduce the total at that point

Septimus_Magna wrote: »

@Tors

The performance in Cyrodiil was pretty good up to 1.4 when they decided to shift client processes to the server side (to reduce farming bots I believe). From this point it has been going downhill performance wise.

I dont think the current problems are all the cause of bad programming, it looks like pushing too much to the server side has been somewhat fatal for the performance in Cyrodiil.

Its good ZOS took measures to exclude farming bots etc but the consequences are too severe for the server performance.

I would definitely vote for the removal of harvest nodes in Cyrodiil (to make it impossible for bots to farm) and push more processes to the client side (similar to 1.4) to relieve stress on the server.

Besides that I believe they are investigating skills/buffs that have a huge checking radius which cause the number of calculations to spike when there are many players in roughly the same area.

I think it was mainly because of duping exploits, not bots?

But yeah. Client was smooth as a licked kitten.