General discussion about depreciated assets and so on..

Mod note: moved posts out of UWS system due to offtopic. The topic in that thread is supporting each other with UWS.

This thread is about general asset store moaning, which is a summary of the threads I moved.

--snip--

Exactly, there are lisencing issues around this. I’m inquiring as to buyer protection. I can disimilate an argument like this however we should now be able to literally sell that code however we please as the original seller has totally revoked responsibility. This is our money, this is our code, this is our business. Unity should consider better handling of this as should all market places.

There is nothing morally wrong with recouping our investments in what has been incomplete Internet services. Due to the amount of money being between €100-€200 there is an obvious discrepancy. The usage of a forum and GitHub implies adequate support. The support has not been provided for whichever reason so the full purchase promise has not been fulfilled.

What I’m saying is buyers should be better protected in what has been a Seller’s market.

The full amount of actual value is not received by users, we can agree on that. If something isn’t achieved to win back buyers then the Asset store format fails, becoming far less appealing as actual business.

If the Asset store doesn’t protect buyers then we are all better off 100% holistic.

 

I actually think there is a reasonable market for such water systems. Nothing earth-shattering but not bad. The problem is more likely along the lines of the ratio of support & updates required compared to profit. By support I mean people asking the same silly questions, people asking reasonable questions, the number of platforms within Unity to support, needing to change things when Unity bring out new versions that 'break rendering' in some particular way. And in the case of this particular asset, the burden of these things may be made much worse by the fact they inherited the code from someone else. I've followed all the forum talk about assets like this one and similar, and its not something I would want to support, thats for sure. Unless I was highly motivated, unlikely to have my own circumstances change for many years, and had come up with an elegant solution that didnt become a nightmare.

I hope that in the fulness of time someone has a crack at a fancy water system for the HD render pipeline. But I would place no bets on this happening and if it does, it will take some time to judge whether such efforts stand a better chance of longevity than assets for the standard unity render pathways have. I suspect the chances of such a thing happening and being a reasonably safe bet depend in part on what water Unity themselves provide for HD pipeline eventually - if they do something that could be used as a stable base that takes away from some of the rendering-related nightmares, and people just have to code some additional stuff elegantly on top of this base, the picture may change for the better one day.

 

The thing is that the new HD pipeline seems more like a marketing thing than anything else, i dont see what more it could do than the current low level vertex-fragment shader approach that could make things much faster. Probably is mostly a library to access more ready to use features than anything else, which is crazy bad because this should have been an addition to the current shader library than a new path with many different new semantics that requires remake of shaders, which may be impossible to do in a single life time and with the zero documentation about how to do it.

I guess that is why major assets may leave the store. If Unity does go ahead and deprecate all my current custom shaders and scripts, then even though i would love to keep my assets going forever, it might just not be possible. My idea is that if i have to support the HD pipeline, i would create a separate asset version, since it could take a year to port or more (or much more, given it is possible at all).

Generally the new approach of destroying everything that is there and create it anew that Unity has adopted lately is crazy bad and i dont see how the store could thrive on such an environment.

I am working on a water system as well currently

and will be with the standard shader and may or may not go to HD pipeline, but i try not to think about it right now, as otherwise i would simply not make the asset at all, as it is totally impossible to create it for 3 different rendering platforms, since it is 3x the work for the same benefit, which is totally crazy, given the new HD pipeline brings nothing new and could just be a library on the current shader path.

 

I am a huge fan of the HD pipeline but I dont intend to explain why here, given what you seem to think of HD pipeline this would just be a waste of time. I do appreciate the issues it causes for Asset Store developers, especially at this particular point in its development, but it is certainly not simply a marketing thing, its great in a variety of ways already.

 

Thanks for the insight in this, given i have been working all this time the lowest level possible with pure vertex-fragment shaders, i fail to see how it could be any different than a library on top of that, but i would love to know if this is not the case for sure.

The worse thing of all is the zero programming documentation for the new pipeline, which makes it impossible to even try to have a better educated opinion on it, without actually going in the code and do trial and error.

I guarantee that if the new pipeline has something that is not reproducible in the current shaders, then i would be extremely interested in it and how this might work. The problem is how can i know that without spending weeks on it, since Unity does not support this pipeline with any documentation.

This is the only resource i have found (https://docs.unity3d.com/Manual/ScriptableRenderPipeline.html), which is a joke and shows that this system is rushed and without any support and also i don't see any shader features that are not readily possible in the current shaders.

Also i see this (https://github.com/Unity-Technologi...wiki/High-Definition-Render-Pipeline-overview):
HDRP does not support OpenGL or OpenGL ES devices.

So i will be working on a black box system that by default does not support mobile. This is a deal breaker, as i am
fond of having my shaders made ground up and i have control over features, thus i can have a mobile version and
a version augmented with some extra code for cutting edge visuals, with minimal code changes, than change all semantics and functions for another different system altogether.

All in all, if there is no features that cant be made with SM3.0 (or SM4.0 for tesselation) in basic shader language and a new library to support the new lighting models, then it is really hard to make something for such a extremely specific range of systems, which is also undocumented on how to go about it to convert the current shaders.

That is why you see me skeptical on that. Should Unity had released this at a mature and fully documented state with a converter for all shaders (not just the standard one), then would be a whole different deal. Supporting the assets for all platforms is already a huge deal and takes away time that is becoming inproportional to the benefit, so adding even more changes like that all the time, makes it impossible to follow if they are in a rushed and undocumented state.

That said, i do plan to convert everything for the new pipelines eventually, but this only when the system is mature and finalized and with many examples to go with, otherwise would be a huge waste of time i could save by going with it later.

Also to clarify, when i say it is a marketing thing, i mean that is rushed for marketing purposes and could have been done in a different way, both in implementation (merge - replace current shaders for example than need new functions and shaders altogether) and in deployment, not that necessarily does not offer anything good (i wont say better here, because already is a system that requires to keep 3 versions of assets for a "unified" project, which is unbelievable to say the least and a vast step back in the make one for all project case).

 

I am not enough of a graphics pipeline programmer to be able to delve into all the details I'm afraid.

I certainly dont think it is for everyone, especially at this stage. Personally, I am a big fan of open development so I am pleased that they decided to develop it 'in the open'.

I do understand documentation frustrations, and that there will be lengthy periods at this stage where it makes people unhappy for various different reasons. My eyes are mostly on the long term goals of it, and I try not to get too carried away. Certainly the Unity marketing department has made the most of it and have not always been totally clear about what stage of development the pipelines are at, and also some of the marketing for 2017 Unity could easily mean that expectations are about 12 months or so out of whack with the actual reality. 2018 has seen far more progress though, but with much left to be done.

Personally a lot of my focus on it so far was in connection to the volumetric lighting/fog system, but I have also been a big fan of quite a number of Asset Store and github volumetric solutions in recent years, so I know this is not something completely new that could only have been done with the scriptable render pipelines.

Also, it is quite likely that some of the stuff I appreciate about the HD pipeline is stuff they might have been able to achieve with the existing render pipeline, had they not decided to focus on the new pipelines instead. But I also appreciate that one of the reasons they did this is that their existing rendering code is somewhat bogged down by legacy issues and mess from the past, and I am happy to see them have a fresh start, even if it causes pain in the intermediate years.

Perhaps some of the advantages will be easier for a broader range of people to appreciate once we have it combined with other new stuff that is going to start coming relatively soon, such as new terrain system and new VFX (particles) node based system.

Certainly I can understand the pain in terms of the platforms that HD pipeline does not support. But at the same time, I appreciate the modern approach to using compute shaders for a variety of rendering engine tasks, and again I have no problem with them having to make a clean break from the past and some tough decisions along these lines. I know that for example WebGL people might not be happy, but at the end of the day its up to that side of the computer/graphics API world to get up to date and offer compute shader support which Unity can then make use of.

 

And certainly for some of those reasons and others, I am not someone who is going to be happy if some users make 'support HD pipeline' demands of Asset Store creators before the time is really right for that sort of thing. If there is one thing I dont like about the scriptable render pipeline approach, its that some users of Unity dont seem to really appreciate how big a change this stuff is in a number of ways that are not easy for asset store creators whose products touch the rendering/shaders/lighting side of things to adapt to easily in various cases.

 

Had to merge a bunch of really off-topic stuff, so it landed here. Do not reply to UWS thread unless you intend to support those users with code or are asking questions about that asset.

 

Sadly I can't agree.

Builtin can't hope to reach the performance HDRP reaches at the same quality level. It's simply impossible. It uses compute shaders and techniques that aren't available or feasible on older GPUs. The SRP backend is way faster than builtin *because* it doesn't do what you want. It isn't lazy - it's removing the performance bottlenecks.

For people who need compatibility there is the LW pipeline. Asset authors who build water, or shaders need to support 3 pipelines:

builtin
LW
HD

If you can't support all 3 then you have to simply live with the fact you'll have less sales. Can't stop progress unfortunately. I also don't recommend asset authors build anything for custom SRPs, only the Unity ones. Because otherwise you'll never survive. 3 pipelines are manageable with minor tweaks between each however.

Sounds like your market target should be:
builtin + lightweight with the augmented version being hd...

I agree, but you know, it's not released yet, right? it's heavy WIP and listed as experimental... Also a converter exists but it will only convert the standard shader. This is because the extra code is too much for it to deal with.

Anyway I do understand why you're upset. A lot of asset authors that specialise in shaders are upset. And that's going to have to be the bitter pill to swallow because it's the price of progress. And it is progress. If Unity didn't do this, your market would slowly shrink as Unity bled users to rival engines.

And truth be told, there shouldn't really even be a market for shaders. People should be able to cobble this stuff together with shader graph and have it run almost if not the same as hand coded shaders. The fact people had to rely on workarounds, hacks and asset store indicated that Unity really did need to upgrade their whole rendering pipeline. And they did.

 

Great choice, we can discuss this in more focus here.

 

Indeed, i hope the demand for the HD one be much later, at least give the time to try to adapt.

The thing is that Unity has promoted this and so many users want support for the latest greatest thing, which makes things very hard for asset developers and puts extra pressure, making it impossible to keep up (considering things like Legacy particle deprecation, Javascript deprecation etc have already taken a huge amount of time to keep up with already, and those are trivial stuff comparing to shaders).

 

I agree a fully featured graph based shader tool (which is definitely not what we have now) could replace much of the work, but i cant be sure all i have done in my shaders could have been recreated with that, even if it is at a godly state.

Also graph based has its draw backs, one is versatility and the other is that code is much easier to use to perform certain tasks (e.g. volumetrics, loops etc)

I suppose a node to write shader code could also help, i think some tools had that functionality.

The other thing is that i dont understand the purpose of the LW pipeline, why for example not keep the HD and current pipelines and use the HD for the DX11+ and the current for everything else. This would directly solve 90% of the problem, since would not have to deprecate anything thus causing the minimum damage. Or in the very least not deprecate the current pipeline at all and keep all three.

 

I cant say much about LW pipeline because thats not my target so I havent paid attention to it. But as I did mention in a previous post, I do know the legacy render pipeline has become a bit of a nightmare for them to keep improving. And I think there are supposed to be advantages to the LW pipeline that they couldnt easily achieve with existing standard pipeline, but I cant speak of the detail of this.

I dont think they will deprecate the existing built in pipeline for a very long time though, and in the various pipeline specific threads they have been clear at various times that people should stick to the standard existing pipeline if the new ones arent meeting their needs. Eventually I suppose that will change, but it wont be soon.

 

Because it's not just about you. SRPs affect the entire pipe-line, not just the shaders. This is something they couldn't do in the current setup, so they had to start to move towards SRPs. Although not today and not in a year (if I have to guess in 2-3 years), but eventually they will drop the current render pipeline and they will remove from the engine altogether.
This means in the future there will be only scriptable render pipelines.
If they keep it, they would block the improvement of Unity and make the code cleaner, easier to maintain, which means more and more bugs and longer iteration when they actually do something.

You should not support officially HDRP just yet no matter what the users say. Using it in production is bad for you and bad for them. Your job would be to explain this to them.
Because if you start to support it, you will have a lot of work to rework your asset over and over, since it's only a preview and nowhere near a released, production ready product.
And if they use it in production, they will have to do the same.

Although I agree with you on one thing: the ability to write shader code for the SRPs is crucial especially for AS developers like yourself or @jbooth.
In place of Unity I would make that clear and easy to do and then would build the ShaderGraph on the top of that. But I guess they will open up the possibility eventually to script shaders as well. Because there will be demand.

 

I've not had a single person ask me about LWRP. I've had dozens and dozens and dozens asking about HDRP and why my assets don't work and when I will upgrade them. This is because, as @nasos_333 has said Unity is using a 'preview' feature for their latest and greatest marketing purposes. This is giving a false impression to many customers who are simply swayed by the possibility of glitzy new visuals, but Unity isn't upfront about explaining the problems people will face.

In the end I don't blame Unity for doing what they're doing (progress is messy), nor do I blame customers who want the latest and greatest. But it is falling on Asset Developers to educate these people to the situation, some are reasonable about it and some aren't. The whole situation is unfortunate. I do feel, as an asset developer, that I've become Unity's cannon fodder, pushed to the "front lines" while they lumber on slowly bringing up the rear.

 

The marketing dept != the developers of HDRP. @SebLagarde has been very gracious explaining it is a preview time and time again to mitigate some of the perceptions people might have.

On the other side of Unity there are many people who do not develop HDRP or SRP, and these people push the information out, they are marketing and evangelists, and this causes pressure on middleware providers like yourselves and I think that's really unfortunate.

I try to do what I can to keep people informed and not annoyed, but yep - most people are going to gun directly for HDRP and builtin will die pretty fast once HDRP is officially released. Only projects that are on LTS releases will be using legacy/builtin after a while.

(edit: clarified some text in bold that I got wrong)

So the 3 year estimate is not a bad one, and middleware authors are in limbo because Unity has officially stated that there will be no shaderlab concept for a while yet (they are looking into it). I think the best bet is being able to code your own nodes and sell those as plug ins for the graph.

It's a new kind of thinking but I guess, inevitable and when people rely on asset store for their living, it can be quite stressful.

Just remember the people developing HDRP do not deserve anger, as they aren't the same people doing the marketing and evangelism.

(Just shout at me or buy a stuffed hippo to kick around if it helps)

 

I'm gonna get me a stuffed hippo and chop it's head off and pull out all it's stuffing and burn it's carcass in the fireplace, just let me answer all these "where is HDRP support?" emails first!

 

It's that bad huh?

 

State on your asset pages and on your initial forum posts that you're only supporting production environments or that you aren't supporting preview and beta programs. Probably a lot of emails would go away.

 

First law of selling assets: people don't read stuff.

 

That, actually, isn't true. There are people who do. (I haven't said the emails will disappear, I said it will lower the load)

 

Is there a word missing there or the wrong word used or is my brain just having a bad day? Because obviously Shader Graph exists, its alternatives to it that are lacking.

 

Well no it's not THAT bad, I did exaggerate for effect... and I don't even own a fireplace.

 

Some emails might go away, I agree. I try to refrain from putting negatives on my asset descriptions though. I find it contributes a bad psychological effect. Always focus on what your product does, not what it doesn't. Not that I want to hide important information from buyers, but descriptions are selling tools. Relevant info is available for those who are interested and want to dig a little deeper.

 

Currently the pipelines are a beta anyway, so there is no need to mention anything about them yet i suppose. I hope Unity will be kind enough to remove all those 1/5 star reviews though that may come out of this new deal.