Steinberg are going to kill off your VST2 plugins.

[Kingston]

VST SDK 3.x comes with wrappers to VST 2.x and AU formats, has done for years already.

oh, ffs. I KNOW THAT. IVE SAID I KNOW THAT. MORE THAN ONCE.

What was the problem again? That Steinberg will pursue legal action against developers when they remove their direct download link?

The last time VST SDK 2.x was updated was 2006. Hasn't stopped anyone from using it.

[Compyfox]

This really depends on the point of view...

I mean... let's be serious. How much on the freeware market currently is actually naive code, and how much is SE/SM shell with custom code? SE/SM is locked to VST2 unless both development platforms are able to be ported over to VST3.

And most freeware providers these days also offer VST3 as well. So they are getting more and more.


The thing is, and here I understand you totally, the host providers need to up to VST3 as well. And I think this is the move that Steinberg is pulling now: "finally update!". As mentioned several times: Presonus, Image Line and Steinberg are the only host providers that have VST3 support - with others still sitting on the fence.


So there are several fractions to blame:
Steinberg with their SDK, the plugin developers that do not provide VST3 plugins (or can't), the Coding Platform providers (SE/SM) for not implementing VST3 or even AU, and the host providers that refuse to implement a more modern VST technology.



And we have to admit, 5 years running VST2 and VST3 SDK in parallel with VST3 SDK being backwards compatible, is way better than what AVID (drop RTAS for AAX with iLok routines) and Apple (32bit support) pulled.

And as long as host offer backwards compatibility (read: down to VST2.0, which VST2.4 and VST3 is capable off!), everything is fine.



I am no developer, I'm just a beta tester - I can't go more in depth. But I do understand each side.

[Krzysztof Oktalski]

There are a lot of smaller VST companies not producing VST3 plugins, so I'd suggest that they're not using the VST3 SDK.

Sure and they will carry on as normal. New devs will have the "benefit"/curse of a broader SDK, which hopefully will nudge Steinberg to fix the issues with it. Nobody is forced to do anything at all and there's not even an inference of dropping VST 2.x.

Having set a precedent with the VST2 SDK, who says that Steinberg wont deprecate support for VST2 in the VST3 SDK?

It's easier to support a single SDK than multiple ones. If Steinberg stopped supporting VST2 entirely they'd lose a lot of leverage because the other platforms would then settle on a different standard. They know this, so it's not going to happen - sleep easy.

Removing it completely is an attempt to force VST3 on develoeprs whether they want it or not.

No it isn't, or if it is, it's a pretty poor attempt! If Steinberg were trying to force everyone off of VST 2.x they'd surely stop support for it within their software. Since that hasn't and nearly certainly won't happen, I'd worry about something more pressing. You could just as easily be talking about the end of the world, the poles flipping, solar flares taking out all electronic devices, etc. There's simply no drama to be had here AFAICS.

This is unfounded sensationalism, and any considerations towards such a notion seem quite counter-intuitive. If ever such a thing happened there would be out and out war between many different factions and it would never get to a point where it effected end-users.

You know a day is messed up when you start it by defending Steinberg! But, the main thing here for normal DAW users is to point out that Steinberg are not going to kill off your VST2 plugins!.

[MFXxx]

This really depends on the point of view...
.....


And we have to admit, 5 years running VST2 and VST3 SDK in parallel with VST3 SDK being backwards compatible, is way better than what AVID (drop RTAS for AAX with iLok routines) and Apple (32bit support) pulled.

And as long as host offer backwards compatibility (read: down to VST2.0, which VST2.4 and VST3 is capable off!), everything is fine.


I am no developer, I'm just a beta tester - I can't go more in depth. But I do understand each side.

I totally agree. Further more it would be great to see more vst converted the the vst3 platform, help a lot with ASIO/CPU overheads. After all it's a challenge and business opportunity for the existing non vst3 products. Commercially who would upgrade to Cubase 7.5 or 8 should vst2 support be totally dropped?

[Krzysztof Oktalski]

And we have to admit, 5 years running VST2 and VST3 SDK in parallel with VST3 SDK being backwards compatible, is way better than what AVID (drop RTAS for AAX with iLok routines) and Apple (32bit support) pulled.

Bang on the money. Anyone that's seen anything of the RTAS->AAX transition can see that Steinberg are being comparatively reasonable, slow-moving and considerate lovers. There's nothing to suggest they're going flip you over and skullfunk you any second now.

[Compyfox]

Even if there will be a VST4 SDK by (say) next year and with Cubase 8, I still think that it will remain backwards compatible down to 2.4 at least!

Else they'd shoot their own foot. We had enough format wars already. Or does anyone even remember DX? Now it's VST, AU and AAX (rarely still RTAS) - and not VST, AU, AAX, RTAS, DX.


Now the host providers need to play along as well. Meaning: implementing VST3 rather than sticking to VST2. Then every developer can sleep better and focus on less provided standards.

Like: VST3, AU and AAX in native form. And AAX (according to a lot of developers I have contact to) is a PITA already.

[tony tony chopper]

Steinberg should have done this earlier. Most programmers don't wanna adapt, if they want the better VST3 to be adopted, they have to force it.

There is hardly any (well, maybe very specific professional stuff) software domain using as old & retarded technologies as what the music software domain is. (& I'm mostly talking about MIDI here). Just pick a video on 3D modelling & watch how much more advanced today's 3D apps AND users/artists are.
Meanwhile in the music domain, someone wanna change a note's parameter, but can't, because of a 30 years-old protocol that says that you only get velocity. With VST3 you can, this should already be the past, I hope it will be the future.

[whyterabbyt]

Removing it completely is an attempt to force VST3 on develoeprs whether they want it or not.

No it isn't, or if it is, it's a pretty poor attempt! If Steinberg were trying to force everyone off of VST 2.x they'd surely stop support for it within their software.

That's not quite what I meant there, FWIW. Im sayig that if a developer has to build for VST2 via the VST3 SDK, they're building VST3 plugins by default, when that wouldn't necessarily be happening before. That's 'artificially' increasing the VST3 developer pool solely because of the removal of the VST2 SDK.

[GaryG]

Umm, maybe I'm going mad but isn't that a double negative?

After that point in time no one who doesn't already have the VST2.x SDK will be unable to legitimately obtain it

So no one will be unable to obtain it?

Where's the problem!

[Compyfox]

I still try to understand why dropping an over 6 year old SDK in favor for a more maintenanced one, which still drops VST2 versions(!) is a bad thing?


It's like saying "you know what - HTML5 is out but I still code in HTML1", even though most browsers are backwards compatible and dropped certain faulty routines from that language in order to tidy up code.

[Kingston]

That's not quite what I meant there, FWIW. Im sayig that if a developer has to build for VST2 via the VST3 SDK, they're building VST3 plugins by default, when that wouldn't necessarily be happening before. That's 'artificially' increasing the VST3 developer pool solely because of the removal of the VST2 SDK.

If someone has deep investment on some framework they built around VST 2.4 then it might be a problem for writing new plugins. But having a business rely a on such ancient and problematic framework is not entirely wise either so that's probably a very minority group.

For the rest of the developers it's a good thing. VST3 forces better code and resolves most antiquities (ambiguities) of the old SDK. There's no development overhead writing a VST3 plugin compared to old SDK.

[whyterabbyt]

Umm, maybe I'm going mad but isn't that a double negative?

After that point in time no one who doesn't already have the VST2.x SDK will be unable to legitimately obtain it

So no one will be unable to obtain it?

Where's the problem!

pthththt. will edit to 'able to'

[Krzysztof Oktalski]

Im sayig that if a developer has to build for VST2 via the VST3 SDK, they're building VST3 plugins by default, when that wouldn't necessarily be happening before.

They're not making you support VST3 at all and it takes no more time/effort. We've been doing all of this stuff in a single mouse-click for years, it's really not a big deal!

[FabienTDR]

I still try to understand why dropping an over 6 year old SDK in favor for a more maintenanced one, which still drops VST2 versions(!) is a bad thing?


It's like saying "you know what - HTML5 is out but I still code in HTML1", even though most browsers are backwards compatible and dropped certain faulty routines from that language in order to tidy up code.

No, it is an incredible amount of work and testing. It asks most developers to completely re-think and re-work their whole build system. This is (much) more than just changing the doctype in a html.

The problem I see is that Steinberg practically never gave any kind of support on 2.0 anyway. Why trust them? For their market "power"? lol

[Kingston]

No, it is an incredible amount of work and testing. It asks most developers to completely re-think and re-work their whole build system. This is (much) more than just changing the doctype in a html.

To their defense Steinberg does meet the developer halfway. There's now a fully documented test host and a somewhat formalised support for unit testing.

The old SDK had neither, having you rely entirely on 3rd party hosts and their differing and ambiguous implementations/interpretations of what being a host means (VST 2.4 API is not always clear about that). Unit testing was also completely up to developer and if I had to guess most hobbyist developers wouldn't even know what unit testing means.

Again, this only concerns a developer doing porting. For any newcomers these are completely positive things.

As far as "trusting" Steinberg. It's not like you have a choice.