https://www.youtube.com/watch?v=nO-NLBjSO1Urelentless_turnip wrote: »I think it's too early to tell and pc EU is still really bad, though instantly better when ball groups log off... It should definetly be part of the overall solution, but I still think we need a full skill audit and make as many single target as possible.
Please reduce the effectiveness of proc set stacking as well, its not fun at all...
TX12001rwb17_ESO wrote: »So Healers are the cause of the lag?
Well good thing that ROTPO is coming out.
TX12001rwb17_ESO wrote: »So Healers are the cause of the lag?
Well good thing that ROTPO is coming out.
[Quoted post has been removed.]
EdmondDontes wrote: »Can the OP explain why things worked so much better from 2014 to 2018 with cross healing and population caps at least double what they are now?
I'm pretty sure they just need to put the game back onto servers that can support it.
Czekoludek wrote: »[Quoted post has been removed.]
Yeah but this test was his proposition and after he mentioned that on stream forum was full of ppl who spam about this new test possibility. So you can laught about worshiping fengrush but it is probably because of him, zos didn't even consider this in their first tests that were announced.
Definietely the best test lag and gameplay wise but performance is still bad. So I think this group thing is one of the reasons why performance is poor but there is a lot more.
Tbh at this point i'm not even shocked that idea proposed by player is way better then what zos proposed at the beginning as devs clearly don't know how their spaghetti code works and what makes the game successful and healthy
luen79rwb17_ESO wrote: »Aww so cross heal is an issue because those heals save many casuals and solos from their small ball-group...! Cross heal is an issue, big groups were an issue, AOE Caps were an issue... Anything that doesn't fit their style or counters it, is an issue! What's even worse they want to force everybody into small groups too.
Thevampirenight wrote: »EdmondDontes wrote: »Can the OP explain why things worked so much better from 2014 to 2018 with cross healing and population caps at least double what they are now?
I'm pretty sure they just need to put the game back onto servers that can support it.
The problem is the game is growing more players are playing it and we just have two mega servers for each platform.
They moved so many calculations to server side to remove cheating and I'm guessing as well to add Stadia in. So the servers are doing all the combat calculations and not the client. lag has been a problem its just its gotten a lot worse over time.
People think the old days of lag are better then todays. I suspect a lot of contributors to why the server can't handle as many calculations is the servers have to do way to much, it has to load in monsters, it has to load in the landscapes, it has to do all the storage when it comes to stuff like ui stuff like the costumes, outfits and stuff like that. Then all the achievements that are not account wide but more character wide.
The new collections for the account would remove so much unneeded inventory clutter allowing you to safely delete stuff. Which should help when it comes to inventory management and allow the servers to store less on the database side.
The gist of what needs to be done.
1. Reduce Calculations obviously, and the cp system to be less impactful on performance.
2. Making achievements account wide and also make all those books and letters you collect for Eidetic memory account wide. This will remove a lot of achievement data that needs to be stored per account and by making the memory Eidetic account wide as well they can reduce the number of books in the database that has to be calculated by eidetic memory. Doing this would allow for them to add in many more character slots too. As it would reduce all that stuff that needs to be stored on their databases.
3. Condensing down the materials needed to craft things. This means like each style has their own unique way of building it. What they need to do is this, the more Daedric type armors that are crafted use daedric hearts. Reachman, Bosmer and other types of armors that use more of a bone theme use bones. By removing all the unneeded regents and condensing that down so each motif uses a shared regent one needs to craft that style, they can in fact reduce the worst of the clutter.
4. Making the older dlcs just base game after so long especially the dungeon dlcs. Since they tie into the group finder and that causes issues I'm certain. That would help remove unneeded calculations and help with number 1.
5.Outfits, instead of making them just one per character then buy ten each character, which is expensive they have one for each character slot you have, plus the ability to unlock ten more for your account. Why do this is that would help with not having to store all that information. So basically if they add more character slots and you buy one you get one more outfit slot. So got 17 characters you have 17 outfit slots plus the ability to make ten more. Since its account wide you can have the same outfit for multiple characters.
6. A cap on the Eso Plus crafting bag, a high one like 5000 or 7500. This would at least help make it so its not an infinite storage bag with unlimited storage for regents you forget about. Plus there is possibly people with tens of thousands of regents just sitting in their left abandoned. That would either get people to storage it somewhere on alts or just delete the excess or sell it.
7. Account wide mount feeding removing the need to have that data for each character and just having one version of the mount feeding system then one for each character that would not only be beneficial to reducing space but also be very alt friendly.
So if they were do all that I think it would help the servers and allow them to perform better then they are now. Plus it would allow the servers to handle many more players that are playing the game now.
Thevampirenight wrote: »EdmondDontes wrote: »Can the OP explain why things worked so much better from 2014 to 2018 with cross healing and population caps at least double what they are now?
I'm pretty sure they just need to put the game back onto servers that can support it.
The problem is the game is growing more players are playing it and we just have two mega servers for each platform.
They moved so many calculations to server side to remove cheating and I'm guessing as well to add Stadia in. So the servers are doing all the combat calculations and not the client. lag has been a problem its just its gotten a lot worse over time.
People think the old days of lag are better then todays. I suspect a lot of contributors to why the server can't handle as many calculations is the servers have to do way to much, it has to load in monsters, it has to load in the landscapes, it has to do all the storage when it comes to stuff like ui stuff like the costumes, outfits and stuff like that. Then all the achievements that are not account wide but more character wide.
The new collections for the account would remove so much unneeded inventory clutter allowing you to safely delete stuff. Which should help when it comes to inventory management and allow the servers to store less on the database side.
The gist of what needs to be done.
1. Reduce Calculations obviously, and the cp system to be less impactful on performance.
2. Making achievements account wide and also make all those books and letters you collect for Eidetic memory account wide. This will remove a lot of achievement data that needs to be stored per account and by making the memory Eidetic account wide as well they can reduce the number of books in the database that has to be calculated by eidetic memory. Doing this would allow for them to add in many more character slots too. As it would reduce all that stuff that needs to be stored on their databases.
3. Condensing down the materials needed to craft things. This means like each style has their own unique way of building it. What they need to do is this, the more Daedric type armors that are crafted use daedric hearts. Reachman, Bosmer and other types of armors that use more of a bone theme use bones. By removing all the unneeded regents and condensing that down so each motif uses a shared regent one needs to craft that style, they can in fact reduce the worst of the clutter.
4. Making the older dlcs just base game after so long especially the dungeon dlcs. Since they tie into the group finder and that causes issues I'm certain. That would help remove unneeded calculations and help with number 1.
5.Outfits, instead of making them just one per character then buy ten each character, which is expensive they have one for each character slot you have, plus the ability to unlock ten more for your account. Why do this is that would help with not having to store all that information. So basically if they add more character slots and you buy one you get one more outfit slot. So got 17 characters you have 17 outfit slots plus the ability to make ten more. Since its account wide you can have the same outfit for multiple characters.
6. A cap on the Eso Plus crafting bag, a high one like 5000 or 7500. This would at least help make it so its not an infinite storage bag with unlimited storage for regents you forget about. Plus there is possibly people with tens of thousands of regents just sitting in their left abandoned. That would either get people to storage it somewhere on alts or just delete the excess or sell it.
7. Account wide mount feeding removing the need to have that data for each character and just having one version of the mount feeding system then one for each character that would not only be beneficial to reducing space but also be very alt friendly.
So if they were do all that I think it would help the servers and allow them to perform better then they are now. Plus it would allow the servers to handle many more players that are playing the game now.
1. This would help, but IIRC Zenimax looked into performance with CP enabled vs disabled, and CP didn't contribute much to the performance difference. Obviously CP contributes to performance (everything related to calculations will negatively impact performance), but it may not be a large contributor.
2. Anything related to storage will not impact performance, only how much data is needing to be stored in the database. Even then, I doubt this will help much to reduce the amount of data needing to be stored. Lorebooks are probably represented as just unique ID's in the database, with a boolean (true/false) paired with it that signifies whether the player has discovered that particular lorebook.
3. As said above, anything related to storage will not impact performance. All this does is streamline crafting.
4. I doubt the accessible dungeons contribute much to the performance of the group finder. If it's done smartly, they're probably cached on the player so that the group finder literally just has to go "is this in the list". Maybe even cache the DLC ID's themselves, instead of the dungeons.
5. The 10 additional outfit slots will still be stored within the database, so this would do nothing. All this does is streamline outfit slot purchases.
6. Size doesn't matter for the crafting bag, because of how it works. In the database, the crafting bag would just store the item ID of whatever material is being stored, with an integer paired with it that signifies how many of that material is stored. Doing so would only take up 4 bytes (an integer is typically 32 bits in length, there's 8 bits in a byte, so 4 bytes in length), which is insanely small in comparison to other things being stored in the database. Not to mention that an integer taking up 4 bytes can represent any whole number from 0 to ~4.295 billion, assuming it's unsigned (ie doesn't support negative values).
7. Like the above, this probably wouldn't matter. This can be stored in just 3 bytes (a byte can represent any whole number from 0 to 255, assuming it's unsigned), potentially with some extra space to fit another value if you only care about the whole numbers from 0 to 60 (0 to 63 can be represented in just 6 bits, leaving 2 extra bits free per byte, there's 3 bytes, so that's 6 extra bits, where another 0 to 63 can be stored). 3 bytes. That's 3/4's of a whole integer!
Thevampirenight wrote: »Should Test 5 be added in as the perma solution? is this poll.
Basically Fengrush he's part of the stream team and a major pvper. He's really good at it and is skilled enough to know many moves players make and knows exactly what causes the lag.
I think he's the one that got Zenimax to do this type of test, and so far everything he has said about healing seems to be correct. It is the biggest cause of Cyrodiil lag. Reducing Group size and removing so much cross healing by limiting it to groups has done a great job in fixing it going by forum threads here. Should Test 5 be the one that is added as the perma solution? As it is the fix, and its the lag fix that is needed to fix pvp Cyrodiil.
Should the Fengrush solution be implemented?
Here is a video where he talks about and shows off examples of why cross healing causes the lag and why it causes more strain then damage abilties.https://www.youtube.com/watch?v=nO-NLBjSO1U
second poll option I should have said test go live permanently. The reason why I worded it that way because those that say no to this is basically asking them to game performance killing calculations caused by cross healing. Removing the cross healing looks to be what is and the only solution to fixing the majority of lag issues. Fengrush was 100% on the ball when it came to that issue and guess what he was 100% correct that it was a big factor, along with ball groups.
They really should upgarde the servers and, its the feeling of some folks and I've seen this talked about aby folks and Fengrush himself. There is a possibility they have in fact switched, changed or downgraded the servers. Possibly to save money, which makes things worse for performance.Thevampirenight wrote: »Should Test 5 be added in as the perma solution? is this poll.
Basically Fengrush he's part of the stream team and a major pvper. He's really good at it and is skilled enough to know many moves players make and knows exactly what causes the lag.
I think he's the one that got Zenimax to do this type of test, and so far everything he has said about healing seems to be correct. It is the biggest cause of Cyrodiil lag. Reducing Group size and removing so much cross healing by limiting it to groups has done a great job in fixing it going by forum threads here. Should Test 5 be the one that is added as the perma solution? As it is the fix, and its the lag fix that is needed to fix pvp Cyrodiil.
Should the Fengrush solution be implemented?
Here is a video where he talks about and shows off examples of why cross healing causes the lag and why it causes more strain then damage abilties.https://www.youtube.com/watch?v=nO-NLBjSO1U
second poll option I should have said test go live permanently. The reason why I worded it that way because those that say no to this is basically asking them to game performance killing calculations caused by cross healing. Removing the cross healing looks to be what is and the only solution to fixing the majority of lag issues. Fengrush was 100% on the ball when it came to that issue and guess what he was 100% correct that it was a big factor, along with ball groups.
This poll is obviously bias just by looking at the questions. haha
The problem is their servers can't handle the load. That is the problem. It also makes absolutely no sense why it would just be healing calculations causing lag but not all the other so-called calculations. So no - I don't believe this is the "major" source of the lag. Another excuse to nerf healing? Absolutely.
They can trim around the edges to try and make it less of a problem. But until they upgrade their server infrastructure you're always going to have lag when it gets super crowded. And that Fengrush guy should know that - since a massive zerg usually floods into Cyrodil when ever that guy shows up.
This happens in PvE too when a zone gets extremely crowded. West Skyrim was unplayable the lag was so bad shortly after Greymoor was released. And I don't think that was because everyone was just cross healing each other and over burdening the system with calculations either.