Are we ever gonna see a true engine overhaul?[Merged]

[Lord Trejgon.2809]

13 minutes ago, KrHome.1920 said:

Besides that the game performes exactly the same like before.  Just go to a low population area with a low amount of objects on the screen and the fps of DX9 and 11 are identical.

Personally I see a quite a sizeable difference in graphics quality between the renderers - and I had side by side comparision as I was testing if one glitch was overal glitch of dx11 glitch so I disabled dx11 and rebooted the game and everything went so blurry.....

FPS on both renderers seems roughtly similar.

[Cuks.8241]

9 minutes ago, Mungo Zen.9364 said:

This is far more common than you think.  Almost every software team I have worked on has, at some point, had to invest time into learning and disseminating old code that no one understood.  This usually happens when the developers and coders who broke ground on a project are making design decisions based upon the tools and knowledge they have at the time.  As those team members move on, tools change, methodology is updated,the knowhow of what was built, how it was built and why it was built is also lost.

Oh I know this is common. Its not just software development. I see this in other industries also. Just patching over problems instead of getting to the core.  I've been working for 15 years now as a process engineer in development also helping different manufacturing sites resolving different issues. To some extent it happens to everyone.

But at some point you have to tackle these issues or you get left behind forever. Often this will also require some overhaul in the management team and senior expert staff. But I don't have internal insight in Anet and can't say where the issue is, it doesn't even matter.  The point was you don't  give public statements that basically state "we don't know our product and are unable to further develop it". And it's not just the engine. The UI overhaul is long overdue. These are parts of the game that all the users will be in contact with from day 0, all the time. 

[Redfeather.6401]

Matt Firor and Rich Lambert over at ZoS are actually really great guys and if someone at Arenanet asked them about it, it's possible that they'd give some insight on what ZoS did to give ESO that engine upgrade that increased the game's accessibility on older hardware. My fps went from 35fps in Silent Halls to 80+fps. There is an interview with Matt talking about how important accessibility is to them. And in that same intervew Matt talks about how in the past other game's devs would come to him to talk about stuff. To get some of his experience and wisdom since he's been in mmos for decades now.

[pRogz.2301]

Came to forums just to find if they have made any performance enhancements but looks like they did not.

 

Man, I wish they could, kitten I would even pay for an expansion that only included performance enhancements.

[AusarViled.7106]

I rather not have the engine reworked and have it end like NGS’s new world with bugs and having people duplicate legendaries and sell them for like 1 gold…. Actually I do not have any legendaries yet, so nvm…. Rework the engine. 

[Masego.4850]

17 hours ago, Cuks.8241 said:

I've been working for 15 years now as a process engineer in development also helping different manufacturing sites resolving different issues.

...

But at some point you have to tackle these issues or you get left behind forever. 

You're extrapolating your expertise where it doesn't apply here.

 

That is relevant for systems you expect to continue to be reliant on. Games rarely function that way. GW2 is unlikely to expand beyond EoD, and GW3 is unlikely to continue to use the GW1 engine. With a limited shelf life, those issues don't need to be tackled, they just need things to remain stable as it's adequate for current requirements.

[Sirius.4510]

In case folks forgot, they explained in the article that the transition to DX11 was not expected to yield massive returns by itself (users often reported moderate FPS improvements, regardless). It's the subsequent work that the transition enables that is supposed to pay the dividends. We're still waiting on that, and yes, it is still planned.

[Einsof.1457]

if you play any other mmo you would realize how polished and great gw2's engine is. it's weighty, it's impactful, it's dynamic. What more do you want?

[Cuks.8241]

3 hours ago, Masego.4850 said:

You're extrapolating your expertise where it doesn't apply here.

 

That is relevant for systems you expect to continue to be reliant on. Games rarely function that way. GW2 is unlikely to expand beyond EoD, and GW3 is unlikely to continue to use the GW1 engine. With a limited shelf life, those issues don't need to be tackled, they just need things to remain stable as it's adequate for current requirements.

I doubt it doesn't apply to some extent. I'm not a software developer so I could be wrong in some cases but a decade is not a short lifetime for a product. Besides improvement extends lifetime.

But first of all it doesnt look like they intend to stop GW2 development soon since they announced so many additions recently. Also this has been the case for at least half of the game's lifetime. 

And secondly I think it's in every company's interest to keep at least some internal development and knowledge of it's product key systems. I'm quite sure that like in other industries expertise and solutions can be transfered to future products even if it runs on a different engine (and I'm not speaking strictly about engine anyway, there are other aspects that lack development). 

[Vayne.8563]

On 2/10/2022 at 8:09 AM, Cuks.8241 said:

What  baffles me with this game's development and PR are these strange excuses about spaghetti code or all the original developers left so there is noone to fix it and so on. It's basicaly stating you are incompetent and unable to develop your products.

It really doesnt matter in what industry you are working. I just can't imagine someone working in development giving such an excuse and keeping his job. And to actually give such public statements.

And like vinterberg said, this is not some amateur indie developer, they have decent revenue and this has been going on for years. 

Show me some quotes.  I'm relatively sure those words aren't from Anet but from posters who are trying to excuse Anet.

It's a 9 year old game. The cost of a complete engine overhaul is high.  It's nice to say that Anet said specific stuff, but without quotes no one should take your word for it.

[Mungo Zen.9364]

4 hours ago, Cuks.8241 said:

And secondly I think it's in every company's interest to keep at least some internal development and knowledge of it's product key systems. I'm quite sure that like in other industries expertise and solutions can be transfered to future products even if it runs on a different engine (and I'm not speaking strictly about engine anyway, there are other aspects that lack development). 

The cost overhead for documenting software development falls way down the priority chart for most companies.  What works in a perfect world, does not apply in practice unfortunately.

Some of the most organized teams I worked for used AGILE variants to determine work load and priority.  This is where the team breaks down the workload into small chunks and assigns it to devs/coders to work on.  Even in a well structured environment, no cycles/hours were given to documentation, just to the building.  These teams, with millions of dollars in payroll, could not afford the time to document the work they were doing.

Smaller teams, teams under budget or time constraints, the first thing to go is stuff like documentation.

I actively pursued this with one team, after we sank hundreds of man-hours into reverse engineering modules for our product.  I kept raising the concern that, if we didn't document the work we had done, in a few years, the same situation would arise.  The response was that it was a waste of time and money to prepare for something that may never be needed.

A lot of companies do not care about the long game, rather the here and now.

Edited February 10, 2022 by Mungo Zen.9364

[voltaicbore.8012]

On 2/9/2022 at 3:37 PM, Mungo Zen.9364 said:

This is far more common than you think.  Almost every software team I have worked on has, at some point, had to invest time into learning and disseminating old code that no one understood.  This usually happens when the developers and coders who broke ground on a project are making design decisions based upon the tools and knowledge they have at the time.  As those team members move on, tools change, methodology is updated,the knowhow of what was built, how it was built and why it was built is also lost.

I'm not a software dev by any means, but many of my clients are. When we are discussing why their products look the way they do and are so hard to fix in some respects, this is exactly the discussion that comes up. If the current team can even figure out what exactly from the old buried spaghetti code is responsible for a certain phenomenon, it's usually traceable back to a founding-era developer having to make a very inefficient compromise just so a certain feature can make it into launch.

Yes, there are notes and logs and whatnot, but for projects and teams of a given size, it's like any other human endeavor of that scale - things get lost, forgotten, or not properly understood. Finally you get someone way down the line that does figure out what the problem is and what should be done, but by that point it no longer makes business sense to actually fix the issue anymore.

  

On 2/9/2022 at 3:54 PM, Cuks.8241 said:

The point was you don't  give public statements that basically state "we don't know our product and are unable to further develop it". And it's not just the engine. The UI overhaul is long overdue. These are parts of the game that all the users will be in contact with from day 0, all the time. 

This is also why I'm constantly disagreeing with people blindly clamoring for a steam release, or for ANet to please please vacuum up all the people leaving WoW. GW2 is simply not ready for the kind of attention FFXIV got, and this "we can't really handle our own product" attitude has everything to do with it. Just too many day 0 rough edges don't get fixed.

Edited February 10, 2022 by voltaicbore.8012

[Cuks.8241]

10 hours ago, Mungo Zen.9364 said:

I actively pursued this with one team, after we sank hundreds of man-hours into reverse engineering modules for our product.  I kept raising the concern that, if we didn't document the work we had done, in a few years, the same situation would arise.  The response was that it was a waste of time and money to prepare for something that may never be needed.

A lot of companies do not care about the long game, rather the here and now.

Yeah maybe my perspective is skewed here because I primarily work in an industry where detailed understanding of your product and processes is s requirement that is also audited by external regulatories. And development and changes have to be documented. 

[Astralporing.1957]

On 2/9/2022 at 12:53 PM, vinterberg.2783 said:

It cannot take years to improve.

And yet they have spent years to improve it already. Literally - they were continuously improving it practically since launch.

Seriously, all the easy fixes they could do? They've already done those.

 

[voltaicbore.8012]

5 hours ago, Astralporing.1957 said:

Seriously, all the easy fixes they could do? They've already done those.

I agree this is probably true. There's a big difference between what can be done and what should be done. A lot of the things we still want fixed are possible, but at what cost? People already complain about the slower/erratic pace of content release from ANet, then turn right around and demand that ANet add even more to their list of things to do.

This is not to say that there's no room for improvement. We really don't know anything about how they run themselves as an organization or what their workflow looks like, so we can't really say either way. All I can say is that between my various long term in-game goals and the content ANet puts out, the pace has been enough for me to stick with GW2. I'd of course welcome engine improvements, but my aging potato chugs along at a stable 60 fps for the vast majority of what I do, so it's not my highest priority.

[SoftFootpaws.9134]

Its not going to happen because there's no market for newer graphics, literally. Cards that are five years old and cost $200 on release now cost $400 due to cryptocurrency, current-gen cards cost around $1,000. The devs aren't going to write an entirely new game engine for hardware that no one can even afford to buy anymore.

 

The average user still has a GPU from a decade ago, which is what the game targets. They keep data/metrics on all this stuff and have talked about it many times.

 

They've said if the DX11 Beta does well they'll port more CPU code to GPU shaders, that's it.

Edited February 12, 2022 by Hannelore.8153

[Kiro Kobra.6478]

Devs, 

 

Thanks for bringing DX11 to GW2!

That being said, the game still requires a MONSEROUS processor if you want to set the Character Model Limit to anything above Medium. 

 

I'm over here with a 6-core i7-9750H that boosts to 4Ghz and this game STILL kicks the crap out of it... Meanwhile my GPU is only being utilized at 13%, sipping whine and laughing at my CPU like Leonardo DiCaprio. 

 

So my question is...

With EoD right around the corner, is there any possibility that we will see further improvements to CPU utilization and optimizing multiple characters on-screen? While the current DX11 beta is appreciated (and I don't mean to sound like a spoiled brat), there are a plethora of other online games that handle multiple units on-screen much more efficiently so it really concerns me that we're not seeing this with GW2. 

 

Edited February 14, 2022 by Kiro Kobra.6478

[Ashantara.8731]

Yes, a sensible request. However, I'd prefer them to fix the severe texture display issues and client crash triggers first before addressing further technical improvements, because at its current state the DX11 beta is unplayable.

[Kiro Kobra.6478]

13 minutes ago, Ashantara.8731 said:

at its current state the DX11 beta is unplayable.

I have some texture issues but have not crashed in over a month, it's very much a playable experience. 

 

Unless of course I'm in a crowded area, then my CPU starts crying....

[solstice.1847]

1 hour ago, Kiro Kobra.6478 said:

I have some texture issues but have not crashed in over a month, it's very much a playable experience.

Consider yourself fairly lucky then.  I tried the DX11 beta again just this past weekend, and 3 crashes later (during Pinata, after killing the octovine, and one other that I forget) I had to disable it again.

I'm not using any add-ons (waiting for DX11 to stabilize before I add ArcDPS back) and I've disabled the discord overlay.  Because I had formerly tried a few add-ons, including the GW2 Unofficial Addon Manager (which doesn't seem to work well for disabling add-ons, in my experience), I did a completely fresh re-install of GW2 a couple months ago, to ensure that no add-ons were still hanging around.  I'm on a fully up to date Windows 10 Pro with fully up to date Nvidia GPU drivers.

As long as I don't enable DX11 beta, GW2 is very stable.  Not long after enabling DX11 beta, my game is certain to crash.

For a while after the DX11 beta was added, the developers would include notes about fixes related to DX11 in the new build release notes.  There's been no mention of DX11 in months, though, which kind of makes it seem like the developers believe all problems have been addressed?  At least when we were seeing the occasional "fixed a DX11 crash because of gemstore outfit X..." line in the release notes I knew that the developers still knew there were problems that needed fixing.  Right now, it's not really clear.

[Ashantara.8731]

1 hour ago, Kiro Kobra.6478 said:

I have some texture issues but have not crashed in over a month, it's very much a playable experience. 

Unless of course I'm in a crowded area, then my CPU starts crying....

The texture issues increase drastically in crowded environments, like WvW, over time: https://imgur.com/a/TJH4Prc

Many of the crashes are bound to certain elite specs and skills used, also in crowded situations (like the Drakkar meta, for instance).

If you don't play such content often, you are unfortunately in no position to give an accurate feedback on these two issues and the playability of the DX11 beta. 😐

[Kiro Kobra.6478]

25 minutes ago, Ashantara.8731 said:

The texture issues increase drastically in crowded environments, like WvW, over time: https://imgur.com/a/TJH4Prc

Many of the crashes are bound to certain elite specs and skills used, also in crowded situations (like the Drakkar meta, for instance).

If you don't play such content often, you are unfortunately in no position to give an accurate feedback on these two issues and the playability of the DX11 beta. 😐

I've done Drakkar several times, many other large-scale events, and participated in pretty much every crowded festival event post-DX11 beta release. 

So luckily by your logic, I am indeed in a position to provide feedback on the DX11 beta.

 

Some hardware configs may contribute towards the frequency of crashing, of which I experience very little. So perhaps I am lucky, but that does not make my experience with the beta invalid.  

 

2 hours ago, Ashantara.8731 said:

However, I'd prefer them to fix the severe texture display issues and client crash triggers first before addressing further technical improvements, because at its current state the DX11 beta is unplayable.

Man, what's up with this community?

Almost every time I see a request, question, or suggestion, there's ALWAYS a response like this. "However...[insert some other feature] should be worked on instead." Is that what the topic is about? No. Am I suggesting this is an "either or" scenario? No. Should other aspects of the DX11 port be improved? Absolutely. There's nothing here implying otherwise.

Just because someone has a concern about a particular feature does not imply that other features of the game should be ignored or put lower on some priority list. 

 

Then you have the audacity to tell me my experience with the game is not valid. Send me all the confused-emojis you want, but man, you sure come off as rude. 

 

You could learn a thing or two from Solstice.1847 over here, at least they shared their experience with DX11 in a respectful manner. 

 

ANYWAYS...

My question still remains for anyone who might be "in the know"...  Can we expect some performance improvements related to the number of characters on-screen? 

Edited February 15, 2022 by Kiro Kobra.6478

[Ashantara.8731]

17 minutes ago, Kiro Kobra.6478 said:

I've done Drakkar several times, many other large-scale events, and participated in pretty much every crowded festival event post-DX11 beta release.

Okay, I misread that your CPU started "crashing", when it actually said "crying." 😅 My bad, please forgive me. 😉

[Lord Trejgon.2809]

Personally the only places I crash with DX11 beta renderer are smae places game would crash on me with dx9 anyway. It's playable, but still very much beta (as seen by variance in stability between people, and graphical artifacts dx11 rendered sometimes puts up).

But once it get's out of beta, I am all up for further optimization - after all AN themselves has stated, that while they do not expect DX11 migration alone to fix the performance of the game, it will open new possibilities in optimisation 🙂

[Dawdler.8521]

17 hours ago, solstice.1847 said:

There's been no mention of DX11 in months, though, which kind of makes it seem like the developers believe all problems have been addressed? 

Which is really the most disappointing thing about it. You'd think at least the very obvious texture bug would have been fixed by now. Or something like fullscreen being broken. Personally, I am mostly just plagued by the kitten black frame skip every other minute that is annoying as heck.

Hoping they are doing this intentionally to release a big patch to solve most of it is probably hoping too much, sigh.