And another block bug...

Pyvos

I assume the big feature that'll be announced for Q4 2023 is the actual fix to the block bug, right?

K9002

I'm wondering if the QA department fails to catch some bugs not because they're doing their job poorly, but because they're doing it on the in-house test server. It's nothing like live megaservers or even PTS, it operates under virtually no stress compared to the live environment. I'm saying this because I've seen some Korean games which featured timed and rhythmic mechanics that worked well on Korean servers but utterly failed when these games were published on global servers, due to both much larger population and longer/more variable latencies. Maybe ESO's combat is coded around the reality of internal servers and it fails on megaservers due to time-sensitivity of queuing action inputs.

Jaimeh

MajorProblem wrote: »

ZOS_GinaBruno wrote: »

[...]
We are working on a fix now and want to take some time to thoroughly test over the next couple days to ensure things are stable;
[...]

Whatever they had promised you, it must feel like a shot in the back now.

They are closing down threads asking about how testing is carried out, so be careful around this topic.

React

ZOS_GinaBruno wrote: »

Hi everyone,

While we can confirm we have fixed the original issue with block, we identified the cause of this new blocking issue that went live with yesterday’s incremental patch. This issue most commonly occurs when attempting to swap weapons while weapons are being drawn or sheathed, preventing you from blocking again until zoning or relogging.

We are working on a fix now and want to take some time to thoroughly test over the next couple days to ensure things are stable; unfortunately, the fix for this issue does require another new client build and is not something we are able to hotfix. It will also be present after the console incremental patch on Wednesday.

We understand how frustrating this is for everyone, including us. We’ll let you know as soon as possible when to expect the fix but expect it to be on or around December 19.

It certainly wasn't my intention to spread "misinformation/conspiracy", so let me go ahead and provide the context for my previous claim.

The above quote from Gina suggests that the fix which released today was being "thoroughly tested". This does not appear to be true, because the exact same steps to reproduce the bug from this thread still work.

What exactly does "thoroughly testing" mean in zenimax terms?

Gizit

React wrote: »

ZOS_GinaBruno wrote: »

Hi everyone,

While we can confirm we have fixed the original issue with block, we identified the cause of this new blocking issue that went live with yesterday’s incremental patch. This issue most commonly occurs when attempting to swap weapons while weapons are being drawn or sheathed, preventing you from blocking again until zoning or relogging.

We are working on a fix now and want to take some time to thoroughly test over the next couple days to ensure things are stable; unfortunately, the fix for this issue does require another new client build and is not something we are able to hotfix. It will also be present after the console incremental patch on Wednesday.

We understand how frustrating this is for everyone, including us. We’ll let you know as soon as possible when to expect the fix but expect it to be on or around December 19.

It certainly wasn't my intention to spread "misinformation/conspiracy", so let me go ahead and provide the context for my previous claim.

The above quote from Gina suggests that the fix which released today was being "thoroughly tested". This does not appear to be true, because the exact same steps to reproduce the bug from this thread still work.

What exactly does "thoroughly testing" mean in zenimax terms?

Wasn't QA on some sort of unionizing deal?

KlauthWarthog

They really have gone the extra mile on avoiding that fabled Q&A... and accidentally avoided the regular QA as well.

tomofhyrule

I'm wondering if the reason this seems so hard to fix is that everyone's looking in the wrong place. This might not be a block bug, it might be a barswap bug.

If they reverted block to pre-U36 state, then that means that this problem either isn't caused by block or this issue was extant before U36 dropped, but we just never noticed it. Granted, there aren't many points that people would unsheathe/barswap/light attack all at the same time in combat, and if barswapping reset it to its default state, then the only way we'd notice it is if we got caught by a meteor or something right after it landed (at which point everyone - including the logs - would say that the block just didn't come up in time).

Come to think about it, I have felt that barswap was a little clunky recently and more than once I've ended up on the wrong bar by accident even though I know I hit barswap.

But in the end, if this is just "Block Bug v2 but fixable by barswapping," that alone downgrades this from 'gamebreaking' to 'annoying,' so I'll take it. My progs are on hiatus for the holidays now anyway, but even with the last bug we got a nice win last night, and if we can just do that again but with people being able to fix themselves from block bug in the middle of a pull, that's great.

I do like the suggestion that incrementals be pushed to PTS though, and I would hope ZOS takes that into consideration. The PTS doesn't do much between patches, but if the incrementals went up there and people could test them, that might help to catch a lot of these issues earlier.

PingTheAwesome

My progs are dying.
My friends are leaving.

We gave you time. We accepted the guars. We tried to play nice, to watch our language, to emphasize with the people on the other side of the screen who likely could do very little to fix longstanding, well-known bugs.
We gave you your two weeks. We tried work arounds like testing block before each pull. We talked to each other to ensure all teams were aware of the progress of the bug and the best ways to adjust to a core game function being so supremely unreliable.

And now we find ourselves here with yet another block bug. The good news this time is that we maybe might be able to get out of it with bar swaps? This isn't a fix -- we should not be working around the team's lack of double checking before rolling something out. We should not be adjusting to your team's errors while also experiencing radio silence and snipped comments (on that note, the forums are for more than just the mods or various teams. It is where players can communicate to solve problems -- please let us do so without shutting things down. Since we have to play catch-up with workarounds and guars we didn't ask for, please let us have this one inconsequential thing).

U36 almost hurts worse than U35. U35, we knew that people would leave. The outcry was massive, the backlash swift, and the wallets cinched. But now it's like a slow motion papercut -- watching teammates leave because they've grown angry, disheartened, and turning to other valid options in a marketplace saturated with other games just waiting to be enjoyed (even if they are not MMORPG's).

But I am a sucker. And I will be here, yet again, waiting for a fix. At this point, this is now the fourth? fifth? attempt to fix this. Instead of spending your energy and time censoring a very unhappy playerbase, please just fix the bug. Just fix it.

Thank you. Happy holidays, if you celebrate them.

RupertVandy

tomofhyrule wrote: »

I'm wondering if the reason this seems so hard to fix is that everyone's looking in the wrong place. This might not be a block bug, it might be a barswap bug.

If they reverted block to pre-U36 state, then that means that this problem either isn't caused by block or this issue was extant before U36 dropped, but we just never noticed it. Granted, there aren't many points that people would unsheathe/barswap/light attack all at the same time in combat, and if barswapping reset it to its default state, then the only way we'd notice it is if we got caught by a meteor or something right after it landed (at which point everyone - including the logs - would say that the block just didn't come up in time).

Come to think about it, I have felt that barswap was a little clunky recently and more than once I've ended up on the wrong bar by accident even though I know I hit barswap.

But in the end, if this is just "Block Bug v2 but fixable by barswapping," that alone downgrades this from 'gamebreaking' to 'annoying,' so I'll take it. My progs are on hiatus for the holidays now anyway, but even with the last bug we got a nice win last night, and if we can just do that again but with people being able to fix themselves from block bug in the middle of a pull, that's great.

I do like the suggestion that incrementals be pushed to PTS though, and I would hope ZOS takes that into consideration. The PTS doesn't do much between patches, but if the incrementals went up there and people could test them, that might help to catch a lot of these issues earlier.

Without knowing the "spaghetti code" exactly, but knowing how client/server interactions work, this is my educated guess. There's essentially action queues stored in memory and time stamped. These resolve actions for combat and movement etc. They're also used to prevent the client from cheating and saying its doing such and such or has moved 100 steps in 0.5 sec. So the server must resolve this queue before anything meaningful (interaction with another entity) can occur. This also means that any screw up on the timestamps or status of an object/event/skill/BAR/shield etc can result in a situation along the lines of a status being set like "currently blocking" at the client level but the eso server recognizes no such status. What needs to likely happen is a periodic resolution step where the server and client verify their current states. Its likely a "weapon isnt out yet" or "character is mounted" state is being recognized by one and not transmitted to the other.

alternatelder

PingTheAwesome wrote: »

Instead of spending your energy and time censoring a very unhappy playerbase, please just fix the bug. Just fix it.

Thank you. Happy holidays, if you celebrate them.

The people censoring are not the devs, they are moderators of this forum.

RupertVandy

alternatelder wrote: »

PingTheAwesome wrote: »

Instead of spending your energy and time censoring a very unhappy playerbase, please just fix the bug. Just fix it.

Thank you. Happy holidays, if you celebrate them.

The people censoring are not the devs, they are moderators of this forum.

So the development and Q/A staff are perfectly staffed and funded? Could the salary of say, a forum mod, instead be used to hire someone who knows how to do more than reskin mounts?

tomofhyrule

RupertVandy wrote: »

tomofhyrule wrote: »

I'm wondering if the reason this seems so hard to fix is that everyone's looking in the wrong place. This might not be a block bug, it might be a barswap bug.

If they reverted block to pre-U36 state, then that means that this problem either isn't caused by block or this issue was extant before U36 dropped, but we just never noticed it. Granted, there aren't many points that people would unsheathe/barswap/light attack all at the same time in combat, and if barswapping reset it to its default state, then the only way we'd notice it is if we got caught by a meteor or something right after it landed (at which point everyone - including the logs - would say that the block just didn't come up in time).

Come to think about it, I have felt that barswap was a little clunky recently and more than once I've ended up on the wrong bar by accident even though I know I hit barswap.

But in the end, if this is just "Block Bug v2 but fixable by barswapping," that alone downgrades this from 'gamebreaking' to 'annoying,' so I'll take it. My progs are on hiatus for the holidays now anyway, but even with the last bug we got a nice win last night, and if we can just do that again but with people being able to fix themselves from block bug in the middle of a pull, that's great.

I do like the suggestion that incrementals be pushed to PTS though, and I would hope ZOS takes that into consideration. The PTS doesn't do much between patches, but if the incrementals went up there and people could test them, that might help to catch a lot of these issues earlier.

Without knowing the "spaghetti code" exactly, but knowing how client/server interactions work, this is my educated guess. There's essentially action queues stored in memory and time stamped. These resolve actions for combat and movement etc. They're also used to prevent the client from cheating and saying its doing such and such or has moved 100 steps in 0.5 sec. So the server must resolve this queue before anything meaningful (interaction with another entity) can occur. This also means that any screw up on the timestamps or status of an object/event/skill/BAR/shield etc can result in a situation along the lines of a status being set like "currently blocking" at the client level but the eso server recognizes no such status. What needs to likely happen is a periodic resolution step where the server and client verify their current states. Its likely a "weapon isnt out yet" or "character is mounted" state is being recognized by one and not transmitted to the other.

This makes sense.

In my (absolutely uninformed) opinion, addressing something like a periodic check that the server and client are in the same state may also fix other things, notably things like 'am I in combat or not?' for the perennial 'stuck in combat' bugs.

I haven't been playing as long as many people, but I have heard anecdotes that combat got a lot clunkier once several of the calculations were moves server-side. Could it be that we're ending up with a lot of desynch? If so, could a periodic 'resynch' call fix (or at least address) a lot of issues?

blktauna

Ok well if ZOS does want to do maintenance, revert to before U35 and just maintain.

We know there will be no rewrite. Just admit it. Just be honest with your playerbase. Yes there will be [Snip]. Lots of it. But have some spine and deal with the results of your choices. People will whinge but at least you will have gained some respect in our eyes. We shouldn't have to beg to receive what you promise. We shouldn't have to beg to have a game that works properly. We a4re not the enemy. We have the same goal, a healthy and playable game.

Be honest with us.

[Edited for profanity bypass]

CapnCrunchYT

can confirm, crown store is working

Carcamongus

I argued for patience on previous posts, but the 19th is here and many people report the bug endures. We were promised "thorough testing" two weeks ago. If that promise was kept, then there is something wrong with the bugfixing procedures that might also explain other problems. @ZOS_GinaBruno , do we have to wait another fortnight for another fix or can things be sped up this time?


Mark Wahlberg is disappointed. This breaks my heart.

SaintJohnHM

I was able to verify block was still bugged within twenty seconds of logging in and testing.

I'm glad we waited two months for this "fix".

How many raiders will we lose before the next patch?

cyclonus11

cyclonus11 wrote: »

So no dungeons, trials, or PVP for at least another 2 weeks.

Aaand another 2-4 weeks.

FantasticFreddie

SaintJohnHM wrote: »

I was able to verify block was still bugged within twenty seconds of logging in and testing.

I'm glad we waited two months for this "fix".

How many raiders will we lose before the next patch?

At least one from my teams is quitting after tonight.....

Jaraal

tomofhyrule wrote: »

I'm wondering if the reason this seems so hard to fix is that everyone's looking in the wrong place. This might not be a block bug, it might be a barswap bug.

If they reverted block to pre-U36 state, then that means that this problem either isn't caused by block or this issue was extant before U36 dropped, but we just never noticed it.

The whole situation occurred after the Syrabane's Ward set was released to live, which is a passive AOE block mitigation synergy for allies. If they reverted the changes to block back to pre-U36 standards, and the bug is still present, then maybe there is something with the coding for this set that is breaking it? Just a guess.

ZOS_GinaBruno

We want to apologize because the patch note for today’s block fix wasn’t as clear as it should’ve been. While we did roll back code to make it so block is in a state from before Update 36 launched, we missed a known issue that should have been included in the patch notes.

As some of you have noted, your character can get into a temporary bad state by performing any action that cancels sheathe or wield animations; weapon swapping should resolve the issue. If you cancel the sheathe/wield animation multiple times, you will need to bar swap the same number of times to fix the issue. While we do know that this issue pre-dates Update 36, we are investigating whether any changes in Update 36 resulted in how frequently it can occur.

For next steps, this issue requires further investigation then we will have internal discussions about the plan. We’ll follow back up with everyone as soon as we have more information to share. Again, apologies for missing this critical piece of information in the patch notes for today’s fix.

danthemann5

Most infuriating is the complete absence of any meaningful response from ZOS. The line about being "aware and investigating" really doesn't fly at this point, as the (alleged) awareness and investigation has been ongoing for 7 weeks. This behavior strikes me as contemptuous.

Edit: Of course there's an update while I was writing this post. This still isn't satisfactory, however. What have they been doing for the last 7 weeks? It's not like no one had mentioned the issue(s). So we're going investigate some more and have a few meetings to discuss a plan of some sort. Meanwhile, a fundamental part of the game is not functioning correctly. How much more investigation? How many more meetings? How many more days? How many more weeks?

And what about next time?

code65536

We raided for 2 hours today and nobody complained about block issues.

The repro that people are talking about today is a bit contrived--you generally don't sheathe your weapons in combat, and as people often cast skills to prebuff before combat, people generally don't start combat by light-attacking to unsheathe.

It's a scenario that I imagine most people will never run into in actual gameplay. And the only reason we're seeing it is because it was being used to reliably replicate the more serious form of the bug that is indeed gone after today.

It's fine. Try it in actual gameplay rather than a contrived set of repro steps; the bug that existed before today was legitimately bad since it could happen in actual combat from just bar swapping. The issue people are talking about today is nothing like that.

FantasticFreddie

code65536 wrote: »

We raided for 2 hours today and nobody complained about block issues.

The repro that people are talking about today is a bit contrived--you generally don't sheathe your weapons in combat, and as people often cast skills to prebuff before combat, people generally don't start combat by light-attacking to unsheathe.

It's a scenario that I imagine most people will never run into in actual gameplay. And the only reason we're seeing it is because it was being used to reliably replicate the more serious form of the bug that is indeed gone after today.

Yeah, while I understand the frustration, it doesn't actually sound like a big hurdle to raiding, so there's that at least.

BeeKing

The current state of the block bug is mildly inconvenient at worst and can be remedied by bar swapping if you happen to get it. The sky is, in fact, not falling.

It can be frustrating that it is not completely fixed, but it is in a much better state after today's patch and should not prevent anyone from doing any content.

Jaimeh

FantasticFreddie wrote: »

Yeah, while I understand the frustration, it doesn't actually sound like a big hurdle to raiding, so there's that at least.

Everyone's conveniently forgetting PvP exists...

BeeKing

You can bar swap now to fix it. So the stuck in combat bug shouldnt prevent you from fixing the block bug in pvp, unless you are using oakensoul.

RupertVandy

Yeah, while I understand the frustration, it doesn't actually sound like a big hurdle to raiding, so there's that at least.

The issue isn't the bug or its workaround. The issue is the way this entire thing was handled, the mismanagement, and the ability for a AAA game studio to somehow get a series of very specific and known bugs through QA not once, not twice, but likely more than 3 times. This series of events should and will scare away existing and future players. This game has hemorrhaged its player base in the last 12 months despite how many accounts were created in china. Towns are already ghost towns. I've run across zones without seeing another person. It takes forever to fill a raid now. They are killing this game through negligence and dragging down the Elder Scrolls name and reputation. The big hurdle to raiding isn't this bug, its zennimax driving away people to actually raid.

Ralamil

ZOS_GinaBruno wrote: »

We want to apologize because the patch note for today’s block fix wasn’t as clear as it should’ve been. While we did roll back code to make it so block is in a state from before Update 36 launched, we missed a known issue that should have been included in the patch notes.

As some of you have noted, your character can get into a temporary bad state by performing any action that cancels sheathe or wield animations; weapon swapping should resolve the issue. If you cancel the sheathe/wield animation multiple times, you will need to bar swap the same number of times to fix the issue. While we do know that this issue pre-dates Update 36, we are investigating whether any changes in Update 36 resulted in how frequently it can occur.

For next steps, this issue requires further investigation then we will have internal discussions about the plan. We’ll follow back up with everyone as soon as we have more information to share. Again, apologies for missing this critical piece of information in the patch notes for today’s fix.

@ZOS_GinaBruno: Weren't you JUST begging us for patience, and promising thorough testing after more than a month of this bug being present?

The issue here isn't that you "missed a critical piece of information in the patch notes for today's fix." It's that there is STILL a major, game-breaking bug around block, so your teams haven't really FIXED anything.

Do you think folks in PvP have time to futz around with whatever janky *** removes the bug temporarily while in the midst of already-laggy, under-performing combat? Do you think folks who wipe in PvE will be content knowing that they might not have wiped, had their tank bar swapped 1s earlier, preventing the one-shot that then wipes the raid?

ZOS is doing precious little to earn patience and understanding from their player base.

The_Lex

Jaimeh wrote: »

FantasticFreddie wrote: »

Yeah, while I understand the frustration, it doesn't actually sound like a big hurdle to raiding, so there's that at least.

Everyone's conveniently forgetting PvP exists...

That’s always conveniently forgotten…by the PvE-only crowd AND by ZOS.

Chef42

So I'm guessing the feature for Q4 next year will be a blocking skill line you can buy for crowns? At least then we would know it would work.