About CLAP

[mystran]

There's also resistance to the use of Turtle to define plugin metadata - that's a more legitimate objection, even though it's really 6-of-one, half-dozen-of-the other with respect to any actual alternative. Sure, more people know JSON these days (because the Semantic Web effort failed), but is it actually better for the purpose?

Nobody cares what the file format looks like, what we care about is that there is a separate file for the metadata in the first place. As long as I'm supposed to dump some metadata into a text-file, I don't even care what the rest of the API looks like.

[Dewdman42]

I was expressing regret that instead of unifying around a single plugin API, people who care about all the things that LV2 and CLAP are about are now going to be split into two communities.

Trying to get everyone to "unify" is not realistic. Its like herding cats. There is definitely a chicken or the egg problem in terms of getting some open source plugin format to take over VST where Steinberg has left us short or created unbearable licensing problems. There is not that much motivation for most host developers to spend much time at all adding a new plugin format...or assuming any risk on such a thing. Time is money. LV2 didn't make it easy enough to do or else they would havec adopted it already. The CLAP team seems more aware of this chicken or the egg problem and is trying to make it as easy as possible for existing mainstream host and plugin devs to switch to it. But most devs won't do it until they are very COMPELLED to do so. So what will it take to compel mainstream devs to adopt either LV2 or CLAP or anything else of similar open source ilk? it will require lost revenue without it. Chicken or the egg,.

Something along those lines will have to be the case for ANYTHING AT ALL to take hold. Otherwise CLAP will suffer the same fate is LV2, no adoption by anyone that matters.

And sure, what will happen will happen, but I'd rather not see people making false claims about LV2 as part of an explanation of why CLAP must exist.

this thread is about CLAP. I don't think anyone said CLAP has to exist because LV2 can't do something. People said CLAP should exist because Steinberg sucks. Someone else said, "hey what about LV2" and numerous people have tried to explain to you already why LV2 didn't work out that way (so far).

I think the authors of it don't really care about wide mainstream adoption

This is categorically untrue.

So say you. I have a differing opinion. This is often the case with open source projects. LV2 has too many cooks in the kitchen. I think Its a dead end, but hey like I said....let hte best format win, I don't really care who does it...but see above for the chicken or the egg problem which the LV2 crew simply has not addressed.

[dawhead]

LV2 didn't make it easy enough to do ....

Reaper didn't seem to have any problem with it. Neither did JUCE have any issue adding support for it in their plugin framework.

make it as easy as possible for existing mainstream host and plugin devs to switch to it.

Nobody is going to "switch to CLAP". People are likely to add support for CLAP, which is very different than dropping VST2/3, AU or any other format. The chances of Steinberg hosts supporting CLAP is extremely low to zero, and no plugin developer is going to say "nope, no VST3 for all you Cubase/Nuendo users".

what will it take to compel mainstream devs to adopt either LV2 or CLAP

The proprietary devs that have adopted LV2 either did so long ago, because they saw the point (e.g. Modarrt), or they switched once JUCE made it trivial for them to provide plugins in that format. The same will likely be true of CLAP, with U-he adopting early because they see the point, and others adopting once JUCE or some other plugin framework makes it trivial to do so.

I don't think anyone said CLAP has to exist because LV2 can't do something.

Actually, they've said more or less precisely that.

So say you. I have a differing opinion.

I am one of the (original) authors of LV2 (and also LV1, aka LADSPA). I am also a friend and colleague of all the people who have worked on LV2. It's not my opinion.

[Dewdman42]

ok well you just seem to want to argue. Do it with someone else.

[schwa]

CLAP does cover a few things that LV2 does not (and vice versa), but the distressing part is that LV2 could easily cover them too, if people had not simply done the NIH thing, and reinvented yet another plugin API.

I agree about the overlap in design. Actually I agree with most everything you've said here. Everything CLAP is intended to do could easily be accomplished as a fork of LV2, where most of the changes are removing things rather than adding them.

Simplicity is good, though, so I understand the appeal of a smaller, cleaner interface. From a host perspective, LV2 would be easier to support if it had *less* flexibility.

[pdxindy]

Is this NIH-ism? Is it a genuine reflection of problems with the LV2 "process"? Is it both? I don't know. I'm just really depressed that someone like Urs would simply embrace CLAP while having given up on LV2, when there is every chance that problems of similar magnitude will arise (for someone) with CLAP. What are they supposed to do then, say "we tried to release PluginMatic with CLAP but it just didn't work, so we're now supporting the CLOP effort!"

Like I said, good news is some respects, but so depressing in so many others.

I can see why it might be depressing since you are invested in LV2

As an end user, who knows very little about plugin formats, I am excited about CLAP cause my favorite DAW and Plugins (Bitwig and u-he) are and will be supporting it. Assuming things go along as they appear to be, I will soon see real world benefits with CLAP.

I don't care about LV2 because it has no meaningful relationship to me making music. I can say I support it in some abstract way since I like the idea open source software, but otherwise I have never given it any thought cause it is invisible. In terms of what I do, it does not exist.

Some things succeed and others don't and it is often just the luck of timing and not a measure of worth. If CLAP succeeds where LV2 hasn't, it may be as simple as that.

From your comments I can see that you have put a lot into LV2 so you feel defensive of it. Understandable, but unnecessary.

[dawhead]

From your comments I can see that you have put a lot into LV2 so you feel defensive of it. Understandable, but unnecessary.

Actually not so much. I was around at the start of LV2 (because I was one of the primary authors of LV1/LADSPA). But most of the work in Ardour (the DAW for which I am the original/lead developer) regarding LV2 has been done by others. I haven't written any plugins in my life, and Ardour supports AU, VST2 and VST3. I actually put far more effort into ways to support VST2 on Linux than I ever did into LV2.

What I'm unhappy about is the potential for wasting resources when efforts are repeated/replicated/duplicated. CLAP looks and feels a lot like that. The problem is of course is that sometimes improvements come from relatively mild changes, and it is possible that CLAP is a revisit of many of the ideas in LV2, but in a just-different-enough way to really make a difference. I don't know.

[glokraw]

Competition is always needed. From the musicians end-user viewpoint, Vst has stagnated for ages, and makes it needlessly difficult for the next generation of coders to start a carreer that includes creating plugins. Lv2 never grew to meet the needs of professional musicians. The good instruments available as lv2,
had to be coded to make up for lv2's deficiencies, which only bogs down developers, hindering their end results.

I've never read anything to indicate that coders loved the inner demands of vst or lv2. Leave them in their dust, implement superior ideas in the open, and move on. An open format levels the playing field, invites experimentation, which leads to innovation, which at the end of the day, means more and better music.
Monopoly interest$ always fear and fight outsider innovation.
mi dos centavos

[machinesworking]

I am one of the (original) authors of LV2 (and also LV1, aka LADSPA). I am also a friend and colleague of all the people who have worked on LV2. It's not my opinion.

This is where I'm a little annoyed really. You're obviously a smart person. IMO U-He make great plug ins, they expressed concerns about LV2 in this thread, and are openly promoting CLAP here. Somehow the LV2 community could not solve conflicts with one of the premier plug in developers who is now promoting a new arguably easier to use universal format, and here you are upset it's happened.

I'm of the impression you're not aware that us end users have never used a LV2 plug in on Mac OS, and probably nobody does on Windows, no DAW I know of does except Mixbus and Reaper because that guy will put any format in his DAW. End users want an ecosystem, IMO LV2 has never established itself as an ecosystem, and it's failing some pretty big tests, if I can't run Diva or Surge in LV2 because the developers are running into roadblocks trying to offer it in that format it really doesn't matter how good it is, it doesn't matter how much you think it's their fault, it's not a useful format at that point for me.

[Urs]

I think for me there was one question only: Why hasn't LV2 taken off in a big way?

I have no objective answer to this, I maybe have some subjective opinion or hunches, nothing tangible. From a business point of view, from stress level from every aspect I'm going through right now, I would have preferred to push for LV2 rather than for CLAP. Yet, I could not.

In my opinion, LV2 had its chance to take off and to me it looks pretty much like it did not. The number of plug-ins is in the tens, not thousands (the list is either shockingly small or the official site is shockingly outdated, whatever it is, it's not good. [Edit: Wikipedia says there's a 1000+ plug-ins]). I have watched videos of developer talks explaining it and explaining its issues, and the issues have not been addressed in several years. Then, there does not seem to be an official body that watches over the standard ("governance"). All that makes LV2 look rather dead than alive.

I have consulted with other developers before I made my decision, and I got the impression that my perception of LV2 was commonly shared. I couldn't risk a massive change in my business by betting on it; I think neither would many others. Hence, our decision.

(talks about the form of body that governs CLAP are just about to start btw. - I guess it should be established some time this year. For those who don't know what this is, think about what the MMA is for MIDI)

[glokraw]

Ages ago, the Suse Linux based Jacklab distribution introduced
wineasio, and a few people started using windows plugins in wine. ZebraCM was very welcome. IK's Amplitube was very welcome. And as the list of working plugins grew, a vocal group of linux diehards to open-source, and diehards to linux-native-only, began segregating the who's who, ridiculing WINE at each opportunity, and to this day, some linux people crow about anything made with 'free' software, and licensed according to their pet religion. But religious crowing doesn't make music, and lv2 devs, among others, rejected anything dangerously mainstream. Eight years ago, something basic like supporting .fxb and .fxp in lv2 might have exposed more people to linux synths, and exposed more linux people to sound design and accessability. But the lv2 ship sailed, and got stuck on a reef of it's own making.

I hope clap narrowly focusses on what's best for musicians and producers. If clap takes a load of man-hours off developers, those hours can then be put to new and improved products for musicians and producers. Let it come around and go around

[dawhead]

Ages ago, the Suse Linux based Jacklab distribution introduced wineasio, and a few people started using windows plugins in wine.

Using Windows plugins on Linux with Wine began with the work of Kjetil Mattheusen, who was the first person to figure out a way to get this to work. Myself and another brilliant developer, Torben Hohn, took Kjetil's work and changed it in many important ways to create something called "fst". This became the basis for (a) a special build of Ardour for Linux that could run Windows plugins (b) almost all technologies used to run Windows VST plugins on Linux, right up to today's generally "best of" tool, yabridge.

This had absolutely nothing to do with Suse, nor Jacklab, nor Wineasio (plugins do not use ASIO on any platform).

ridiculing WINE at each opportunity

For many years, Wine was tremendously in flux, with the result that upgrading it to the current version would frequently end up breaking plugins that used to run (and suddenly enabling those that didn't). It also has an even more fundamental problem, which is that because of extremely deep differences in register usage between a Windows process and a Linux process, it is not possible to use Wine as a "library" to run stuff - you must have an actual Windows executable for Wine to execute, which implies either a "strange" build of the host (something we used to do for Ardour), or you have to run the plugins in a separate process. While the advent of hosts like Bitwig have convinced some people that running plugins in a separate process is a good thing, you cannot offer the same performance guarantees. Consequently, while using Wine in some way to allow Windows VST to be used on Linux is really cool, it really isn't a great solution, and that is why over at ardour.org our official position is that if your workflow is dependent on Windows plugins, you should almost certainly use Windows. We have never ridiculed Wine, and I don't know anybody else who has. It is an absurdly incredible piece of technology.

Eight years ago, something basic like supporting .fxb and .fxp in lv2 might have exposed more people to linux synths

It seems that you don't know what .fxb and .fxp files are and/or don't really know anything about LV2 as a plugin API.

If clap takes a load of man-hours off developers, those hours can then be put to new and improved products for musicians and producers.

I don't know of any plugin developer who would claim that existing plugin APIs are causing them extra man-hours. The complaints about VST3 have almost nothing to do with "more work". Besides, no plugin developer is going to drop support for either AU or VST (and those that suppport AAX (ProTools), probably not that either. So any sort of conception that CLAP is going to save anyone time or be a fix for whatever problems people have with current plugin APIs is really just not correct. CLAP will add the list of plugin formats people will have to consider creating for, nothing more or less. Even if there are excellent CLAP->other-plugin-format wrappers, that won't change the questions.

[dawhead]

Somehow the LV2 community could not solve conflicts with one of the premier plug in developers

Actually, from both my own knowledge (which is incomplete) and also Urs' telling of it, there were no "conflicts". Urs never really reached out to whatever the LV2 community is (which is his right to do, obviously).

I'm of the impression you're not aware that us end users have never used a LV2 plug in on Mac OS, and probably nobody does on Windows, no DAW I know of does except Mixbus and Reaper because that guy will put any format in his DAW.

This seems a little crazy. Are you aware that I'm the author of Ardour (the basis of Mixbus), and that among other things, I wrote the initial AU and VST support within Ardour, a DAW that runs on more platforms than anything else you could name? Of course I am aware that most LV2 plugins are not released in binary format for macOS or Windows (just like most VST plugins are not released in binary format for Linux). And of course I am aware of which other DAWs use LV2.

End users want an ecosystem, IMO LV2 has never established itself as an ecosystem,

1000 plugins looks like an ecosystem to me. JUCE (and D-Plug!) supporting LV2 as a plugin format target looks like an ecosystem to me. Independent hosting implementations looks like an ecosystem to me. What is true is that existing plugin developers for Windows and macOS have, for lots of reasons (some good, some not so good) chosen to not be a part of that ecosystem. The arrival of VST3, which officially "supports" Linux, has taken some of the need for LV2 out of things, because developers (such as U-he) can simply release VST3 for Linux if they choose to do so (and I'm very thankful that they do).

and it's failing some pretty big tests, if I can't run Diva or Surge in LV2 because the developers are running into roadblocks trying to offer it in that format it really doesn't matter how good it is, it doesn't matter how much you think it's their fault, it's not a useful format at that point for me.

I know many proprietary plugin developers who have "run into roadblocks" with VST and AU, and yet I'm fairly sure that these are "useful formats" for you (and millions of other people.

[MirkoVanHauten]

I think it's great that 2 of the major companies come up with something. Those, who have shown for years they're established and have experience in real world issues to solve. I don't think either of them easily trashes other options. And who knows, maybe the LV2 community sees the potential as well (if it's really that similar) and shows support of CLAP to gain traktion. It's a bidirectional situation.

[pdxindy]

1000 plugins looks like an ecosystem to me.

1000 plugins I have never seen and have nothing to do with. Not once in discussions with musicians I know has the term LV2 ever come into the discussion or is used by anyone I know. LV2 is insignificant and invisible. I don't mean that as a condemnation, but just stating what seems obvious. None of the DAW's I use or might use support it. None of the plugin developers I use or might use support it.

I get that you wish the energy and development time that is going into CLAP would go into LV2. It isn't and it wont. Things have a rhythm and timing.

Rather than repeatedly arguing against something that is already happening, just wish it well. Otherwise you make yourself an impediment and drag the energy down. No value in that.