On Mon, Jun 30, 2014 at 5:58 PM, Fons Adriaensen <fons(a)linuxaudio.org> wrote: It is quite related to the future of the audio engineering business as imagined by people inside it. What do do we mean by "the future"? Certainly not "how everyone who does anything that might be called audio engineering works", because demonstrably there will be some number of people whose needs, experiences and preferences lead them to continue to work in ways largely modelled around more traditional approaches and technology. But in the sense that more and more audio engineering is going to a) have the requirement to be an experienced engineer removed b) disconnect itself from the requirement to be physically connected to a powerful DSP engine c) utilize the same kinds of presentation metaphors and technologies as have already impacted dozens of other computer-driven processes then yes, this sort of thing is a part of the future. I don't like it very much, but the reasons why this is happening are not really subject to the veto of those who might oppose it.

Am 01.07.2014 00:27, schrieb David Robillard: One way I could imagine to use this: If you put "this" html site as example on your project site, so "users" could just test it, without download or install. If they like it, they could download it for "real" work.

Am 01.07.2014 07:09, schrieb Stéphane Letz: Regarding FaustLive, do you know that we (guitarix project) working on a analog circuit simulation toolkit, which will generate faust code from gSchem ( http://www.geda-project.org/ ) schematics? Imagine a electronic circuit designer could create a plug schematic and "hear" direct what a change from 1k resistor to 2k resistor will sounds like, if such a toolkit be integrated in FaustLive. Faust rocks!°!

Am 01.07.2014 09:09, schrieb Gordon JC Pearce: It is always good to be sceptically! Only this way we could improve stuff. How close it comes to the real hardware depend surly on the simulation toolkit, and at least, still it will be a simulation. As a example, we have the jcm800pre, a GCB95 and a baxandall tone control simulation as plugin in guitarix, generated with this toolkit.

Am 01.07.2014 12:16, schrieb Fons Adriaensen: Haven't say that, nor suggest that. It's just a example, It is just a other way to go. And, let me tell you, this way is full of fun. Have I mention that the toolkit surly include a interface to plot the results and beside faust code, it could compile ready to run Lv2 and Gxplugins? However, this list here is the wrong place to discus that any further, I just hope that some people here get interested in such a approve (Tim you are more then welcome to join us), if, please join us at guitarix-developer(a)lists.sourceforge.net to discus and eventually help us to get it going. regards hermann

[hermann meyer] Thank you kindly Hermann, I hope I'll be able to join you guys at a later date. In re circuit emulation, I have to concur with Fons: meaningful nonlinear full-circuit emulation (i.e. employing more than just a single isolated idealised nonlinear component) appears to be too demanding for realtime application on current hardware (and I think the problem doesn't lend itself well to parallelisation so the current trends in CPU progress won't help much) and if you stay linear I think you're better off with traditional optimised digital filter designs. Not to say it's not a great tool, but I see its value more in gaining insight than realtime processing. Cheers, Tim

Am 01.07.2014 12:16, schrieb Fons Adriaensen: Well, in this case it gets a Infinite-Impulse-Response (IIR) filter.

On Tue, Jul 01, 2014 at 08:09:22AM +0100, Gordon JC Pearce wrote: Unless you run a full SPICE model, including all the imperfections of all components - non-linearities, parasitic capacitance and inductance, memory and thermal effects, etc. etc. And then 1. it isn't easy anymore, 2. it very probably won't run in real time. And without all that, and idealised analog circuit is just another format for some simple mathematical equations, a simple abstraction. Engineers use it all the time to describe all sorts of systems that are not electronic at all, in the same way they use mechanincal abstractions such as mass/spring/damping systems. Idealised circuits are just another notation for maths, an as such they provide nothing new. They just translate into the standard filters etc. that DSP code authors use all the time. It's an illusion to think that youd could generate any 'magic' DSP code form them that can't be obtained in the normal way. On the contrary, you're likely to get inferior quality results as things such as quantisation and numerical stability are blissfully ignored. Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)

Am 01.07.2014 19:07, schrieb Gordon JC Pearce: Here is the thesis which our work based on: https://dspace.vutbr.cz/bitstream/handle/11012/21132/thesis.pdf?sequence=1 by hermann

On Tue, July 1, 2014 3:09 pm, Stéphane Letz wrote: Or enabling people who do not have the expertise to run a native app to do something that they would otherwise not be able to do. Let them view some advertising and even (god forbid) pay for the service. Not everyone has the privilege of lifetime tenure at a government funded cultural institution or wealthy benefactors with pockets full of cash so they can afford to spend their lives working for free on open source technology. If some random marketer wants to spend their clients money on google ads and happens to have their ad appear on my website I'll take the 30c. It helps cover the cost of paying for the server and domain. Let's call it diversification. Anyone who is serious will quickly understand that a native app is a more powerful solution and will look for those alternatives. If they are also provided with information about the alternatives the web version becomes a marketing tool for native apps. Most of the ad spend profit comes from the ideation, design and development process. The actual cost of the campaign is usually a smaller amount. That gives developers, designers and project managers an income which many of them use to self fund their own personal work with a large number of the developers contributing to open source projects. If we take out the marketing industry then we have to find a replacement for those jobs. With the current state of the world it will be a while before the marketing industry is replaced by an international space exploration social co-operative. It could be argued that the marketing industry has enabled a lot of people to stay employed during the past 6 years of global financial meltdown without having to resort to participating in the military industrial complex. -- Patrick Shirkey Boost Hardware Ltd

Le 1 juil. 2014 à 13:03, Fons Adriaensen <fons(a)linuxaudio.org> a écrit : Missing the fact that Faust can directly compile LLVM IR for easy embeddable dynamic compilation, missing the fact that asm.js can be compiled and directly deployed, missing the fact that VHDL (for instance) is one of the future target. If you better know what Faust can do and will do than us, then… And if the "rather primitive plugin standard" is too primitive then, let's works on improve it. Stéphane

On Mon, Jun 30, 2014 at 06:27:05PM -0400, David Robillard wrote: Couldnt agree more. It's all just creating extra layers on the onion, in the illusory expectation that those will be more 'standard' than the ones below. Commercial forces will make sure that such interfaces will diverge anyway. And more often than not they just exist to ensure some intermediate parasites can profit from them. It sometimes makes me think of they way tomatos are being distributed and sold here in Italy. They will travel up and down the entire country (either physically or just virtually) a number of times before ending up in the shops. Both the producers and the consumers are ripped off by middlemen taking their profit on redundant steps. Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)

On Tue, July 1, 2014 8:52 pm, Stéphane Letz wrote: Apparently Fons believes that everything he does is sPeSHaL and should be exempt from the encroachment of market forces. Apparently everyone else who uses open source code for non altruistic purposes is also a parasite. Apparently if you have a nice cushy job working for a government funded non profit organisation you are a higher being and not subject to the general forces that the rest of the world has to contend with. I believe this is called entitlement. Apparently Fons also exists off air and delicately collected rock filtered natural spring rain water and has no need for the other trappings of modern life that general market forces provide. Fons is truly a saintly enigma destined for greater things that are not of this mortal coil. Oh, Hail to the Fons! -- Patrick Shirkey Boost Hardware Ltd

On Tue, July 1, 2014 9:29 pm, Fons Adriaensen wrote: I'll eat my hat if Fons gives up open source development because he has to commit all his personal resources to the Man. -- Patrick Shirkey Boost Hardware Ltd

On Tue, July 1, 2014 10:06 pm, Etienne Rouge wrote: Your opening rant was pretty emotional and potentially insulting to the guitarix team and the above is not much less. Welcome to Linux Audio. A bunch of emotional artistes/developers ranting about the state of things with occasional forays into useful information and community cohesiveness. The latter often being considered dirty words uttered by uninitiated well meaning idiots. Parasites and pedants, godly beings with entitlement syndrome and the barely literate wholly unwashed. It's a rare brew. These ideals that you speak of don't mean much without the support of a wider community but not everyone can justify contributing their time/energy with zero financial reward. So while some people consider it to be theft if other people find a way to profit from open source other people consider it to be a useful and worthwhile outcome that in the long run benefits the wider community. Fons wouldn't have to work for the man if he was able to generate a decent income from his open source ouvre. Yet he has routinely been one of the most vocal in his abhorrence of anyone profiting directly from his code. Even more so if someone else has figured out a way to do it that he personally doesn't have the motivation or interest to do on his own. Herman doesn't deserve to be painted as the bad guy, neither Stephane. They are both active and genuinely contributive LAD'ers who are making worthwhile progress and providing useful results. -- Patrick Shirkey Boost Hardware Ltd

Am 01.07.2014 14:47, schrieb Etienne Rouge: I've no problem with my conversations in this thread. In opposite, I'm glade that I could fix some issues and bugs in my source. The hints I receive from Fons helps me to do that. That is a lot more then we receive from some other guitarix users, even if they use guitarix regular on stage to generate there income (or a part of it). Also Faust will profit at least. Reducing this thread on some namings, or flames, is the wrong way. Just read between the lines and fetch what is useful. Personally, I didn't think that OpenSource is a thankless ideal, it's a lot of fun to work in Open Groups on the same source base. You just have to keep in mind that it isn't a (children) Rose-garden, this rose-garden here, have spikes.

On 1 July 2014 11:37, Fons Adriaensen <fons(a)linuxaudio.org> wrote: This reminds me of a (robust! :-) ) discussion David and I had a few years back. I'll preface this by saying, which is not really a secret around here, that I'm a Java developer. Praxis LIVE is written in Java for a very particular reason - the ability to rewrite code on the fly, and have it JIT compiled. Development at the moment is very much focused on the idea of a visual patcher where you can rewrite the code of any component *as it runs*. This ability to live manipulate code, either by hand or by meta-programming, would seem to open up a wide range of interesting and creative possibilities - instant feedback as a great learning and development tool. There's lots of interesting stuff happening in this space at the moment - eg. Extempore (https://github.com/digego/extempore) I generally agree with scepticism around JavaScript for this (and to an extent with the drive to make the browser an OS), but that's primarily due to limitations of the language (types, threading, etc.) than the idea of a VM in and of itself. David, in the past you've professed to be pro-JIT and pro-GC but anti-VM - I guess I'm still having difficulty understanding what that position means -surely some level of machine abstraction is a necessary intermediate stage here? Best wishes, Neil -- Neil C Smith Artist : Technologist : Adviser http://neilcsmith.net Praxis LIVE - open-source intermedia development - www.praxislive.org Digital Prisoners - interactive spaces and projections - www.digitalprisoners.co.uk

On Mon, Jun 30, 2014 at 4:58 PM, Fons Adriaensen <fons(a)linuxaudio.org> wrote: It's not very interesting as is. Not useful for audio engineering or performance. However--some people may want to write algorithmic or otherwise computer generated compositions, not streamed but generated live on the client end. It's a platform for performance (see for example, WebPd), in its current form bearing a resemblance to IT porn. Chuck