giving linux a shot

[kensuguro]

just letting you all know that my linux drive's been officially unplugged.

Maybe in a 5 more years I'll try it again. See if the software pool's improved by that time. To not get bad software (with the most academically advanced algorithms), I'm happy to put up with a crappy platform (XP) that does its job.

Here's why linux fails:

1. Linux will be the most awesome OS because it's developed by idealists who are far more advanced than commercial coders working under the clock. We code because we love it.
2. Don't complain about any thing that's broken because it's free. We have, like, reports to write and xbox360 games to play yo. Gotta live a life, yaknow what I'm sayin?
3. If you find a bug, you can fix it. That's what open source is all about. Don't know c++ or c? Then why are you using linux? haha...
4. Don't like it? Just use windows or OSX.

So I went from step 1 to 4 in just a week. The hole in the logic is this. Linux depends on users to develop basic aspects of the operating system. For audio, it expects all audio developers to abide by standards that aren't enforced. The result is, people are forced to fix things. And not DO things. It's been 16 years, and still the most basic things (for audio) are messed up. What will it be like in 20 years? Will we have an operational audio platform by then? How about 40 years? Well, the average of an emperor penguin is about 40 years so... I'll keep my fingers crossed.

I dunno, I know I sound very sour. I guess my expectations were too high. I thought academia could pull it off. A sort of ideal where the smart guy can out do the money monger businessman. It's depressing, and I don't like to have to submit to the notion, but I guess so far, it hasn't happened yet.

On a side note, if scope cards got linux drivers, I'll immediately build a dedicated scope machine with linux. It's great for things like that. unmanned machines that do single tasks. It's the whole VST / sequencer part that's messed up.

[valis]

Bit of a corollary here in reference to some of the web/3d/video aspects I mentioned. In all respects you have one or more commercial companies involved in or primarily responsible for the codebase. In some cases it's an in-house application that gets 'opened up' to the general community (quite frequent in 3d & video), in other cases it's a collection of open source libraries that gets wrapped into a codebase and cleaned up by one (or more) companies that make use of said codebase (hello ibm/sun/etc). I think in most cases you can trace the success of the project back to the fact that there was funding available for quality project management and coding, and the vendors putting the funding in generally had an idea of where the project should go.

Focusing back on the audio world, draw your own conclusions as to why open source audio hasn't fared as well. I know that Ardour was supported by Google's "Summer of Code" and there were press releases of at least 1 audio studio playing a major role in its development (iirc). Yet when I've tried it, it really seemed like an audio-focused application that would have been a lot more competitive 4-5 yrs ago.

[BingoTheClowno]

The result is, people are forced to fix things. And not DO things. It's been 16 years, and still the most basic things (for audio) are messed up. What will it be like in 20 years? Will we have an operational audio platform by then? How about 40 years? Well, the average of an emperor penguin is about 40 years so... I'll keep my fingers crossed.

You are being unfair by failing to mention that PC technology is not stagnant and that PC board manufacturers are failing to share their specifications with the open source community. I have actually run Ubuntu from a CD on my 5 years old laptop without a hitch, with audio, graphics, ethernet detected from the start.

[astroman]

well, I actually am long enough in this business to be able to tell you that PC technology is indeed stagnant for a much longer period than 5 years...
just because things have smaller structures or are bigger in numbers doesn't mean they represent a fundamental 'progress'.

If you compare what 'technology' allows you to do 'as a user' on a current system to a vintage one with barely 1% of the computational resources, there can't be any other verdict but 'crap'.

the original MacOS was introduced by Apple 25 years ago as a conceptual alternative to any ix-ish operating system from the user's point of view, and they pretty much succeeded with that part of the strategy.
Those aspects have been adressed by Apple's own marketing in the famous 'Macintosh Intro' spot by Riddley Scott.

Only the fact that satisfied customers bought significantly less machines because it just wasn't necessary to switch systems once per year made Apple change their policy.
Probably even a Steven Jobs was a bit concerned about his future revenues during retirement...

noone questions that an ix-ish OS cannot do this or that techie job as a server or an embedded piece of code.
but it never was intended to be a user OS

Apple is at least able to fake user friendlyness, but after all they sell us nothing but a resource hog to do the same job as before with only a slight improvement - considering the amount of resources spent

cheers, Tom

[BingoTheClowno]

well, I actually am long enough in this business to be able to tell you that PC technology is indeed stagnant for a much longer period than 5 years...
just because things have smaller structures or are bigger in numbers doesn't mean they represent a fundamental 'progress'.

If SATA interface, Hyperthreading technology, Dual/Quad Core technology, 64 Bit Processors, faster bus speeds are not fundamental changes then yes, the technology is stagnant.

[hubird]

man I'm glad there's Apple

[garyb]

why? it's the same old b$ with apple as with m$.....

[hubird]

just the OS alone already to work with...OS9 and OSX.
then the feeling of sureness about the hardware and everything working as it should...also after little or less little changes, check the problem forum...
but feel free to use XP, Vista or Linux...while gambling with motherboards

[Liquid Len]

just letting you all know that my linux drive's been officially unplugged.

Maybe in a 5 more years I'll try it again. See if the software pool's improved by that time.

Nothing's changed since I tried this in 2004, I see. You got a lot farther than I did! Linux is a curiousity to me, I have it on a partition on one of my computers, but I rarely have any use for it. (Actually what really kills even experimenting with it for me, is I find every version of linux I've tried did not like my multi-computer switchbox, or at least, can't handle being switched away and then back.)

The most use I have for Linux in a music environment is this : there is a version called Knoppix that fits on a CD that you can boot from. I find this useful if the operating system on a computer is corrupted and will no longer boot. When Knoppix boots, you can see all your hard drives in explorer-type windows. It will only write to a FAT32 drive, but a fat32 formatted datastick will work.

[astroman]

...If SATA interface, Hyperthreading technology, Dual/Quad Core technology, 64 Bit Processors, faster bus speeds are not fundamental changes then yes, the technology is stagnant.

as I said, structures were shrunk, clocks increased a little bit, but it's still the same old story.

The signal integrity on serial lines improved by improved controllers - the data and it's protocol is still the same. This is refining, but nothing revolutionary.
Adress spaces ? you might know what the VAX synonyme of Digital Equipments probably best selling machine stands for virtual adress expander. That's early 80s technology.

If you transfer those software algorithms to hardware registers that's not really exiting. It just didn't make sense in 1983, when a gigabyte of Ram had an equivalent price of $ 4 million and 10 mega(!)byte disks sold for $ 5k.
The original Mac could adress (and handle !) far more memory than present in hardware. I've used a machine sharing it's 128 kbyte of Ram between an OS, the Finder and M$ Word with a 50kB document.
Photoshop would have never been written without the sophisticated memory management and a dedicated interupt driven user interface library.

Now go ask the Wikithingy when it was actually written and under which operating system and which developement system they used...
you don't really want to tell me that a bloated Visual-I-dunno-what is even remotely comparable...

Can you even imagine the complete(!) GUI of Scope is handled entirely by the main CPU, simultaneously to your application ?
And this CPU runs at 32 megahertz, not 3.2 giga

well, exactly that type of stuff was possible around 1990 - it had to be, because there were no dedicated GPUs with tons of local Ram .
Today's code simply would stand still under these conditions...

multiple cpu cores ? I just reached behind into my bookshelf Highly Parallel Computing, released 1989 by Benjamin/Cummings covered pretty much of the 'problems'. That a company as M$ will never be able to succeed in this domain is a natural matter of fact.
What's all this hardware for, if the software doesn't use it
Even under full load a current CPU is effectively idle for about 75% of the time in applications like a DAW.

Indeed - it seems fast, if you've never seen any other system, in reality it's slow like a snail if you consider the amount of resources wasted.
but as already mentioned, it's providing much more opportunities to make a living... I'd be a fool to complain.

cheers, Tom

[kensuguro]

Bingo, well you do bring up a valid point. I think I should clarify that the basic aspects of a normal OS, like recognizing audio, video, internet all worked completely fine straight out of install. This was much better than when I tried redhad linux 10 years ago. (couldn't get through install) So, I think I may have already said this, but if it's not for audio, especially time constrained audio work, then linux may have its place.

What was unacceptable to me, as an "audio workstation" was the lack of basic coordination between applications. Most everything was standalone, which in itself isn't a problem, but what IS a problem is that programs had to be run a specific order, get wired properly in JACK, and in worst cases, had to be internally wired within the sequencer app. (in my case MuSE) Not to mention a selection of basic synths, or VST wrappers. I was just shocked that something marked as a "studio" product, such as Ubuntu studio, didn't have these basic things sorted out. There is an app called Ubuntu Studio, that you can set up to run the apps you want to use from a checklist, but this is far from a "total recall" functionality... which should be as simple as "load project" like in every single sequencer app out there. Never mind the mediocre VA synths and other semi-broken apps, it's the lack of total recall that makes it very unusable for me. Perhaps it's relying on the success of linux as a video editing studio machine, which I hear they've succeeded in a formidable scale. (because Ubuntu studio is marketed as a general "creative studio" machine, audio being a part of it)

Again, I'll re-emphasize that I don't hate linux for no reason. I actually like linux, and want it to succeed. It's just painful to see it fail so miserably at what it wants to be. I would have thought it would be a "so-so audio workstation with so-so apps", and not "can't really be used for serious work at all".

The other thing, though... "XX company won't opensource their stuff" is heard all over the open source community.. I don't believe many hardware companies will open source their stuff. It's unfair to their business. They spend time and money to R&D a product, and I think they deserve to sell it, and keep it to themselves.

That said, I don't think hardware support is a problem for linux audio, it's all the crappy apps and bad platform structure. Linux can support JUST sound blaster, but the software desperately needs to improve. Or else everyone will just set up a linux box and try to run VSTis on it, and upon failing, just return to using VSTi on Windows. The verdict? People go where the instruments are. At least for me. And then they can worry about getting a proper sequencer. Dunno, maybe I should go through rosegarden in detail, but for now, I've lost my appetite.

[valis]

Another 5 years and I bet linux audio will be more user friendly and coordinated, as competing codebases die off slowly, stagnate or wind up finally coming to their collective senses and combining. Unfortunately commercial windows/mac software will also be 5 years ahead at that point as well which brings us essentially back to where we are now.

[bilbobugginz]

hi guys. I see you do try linux.

When you try next time, please don't be tempted to use the default ALSA sound architecture, give a try to the OSS. Many professional devices are better off supported by the LATEST OSS drivers (it's another type of drivers). So, even if you use multimedia optimized distribution of linux/GNU try the OSS drivers, and see the results (latency etc.), I mean OSS drivers from the OSS manufacturers, www.4Front.com
But this isn't what I'm intending to talk about.

It is expected that a system optimized for developers and IT people is not yet ready: most pro hardware doesn't work.
IF it worked, the "UbuntuStudio" would find way to make it a 1-click installation.
It's a chicken-egg situation: the pro users won't use system that doesn't support their hardware. new users won't go for it too.
and system distributors won't optimize for audio installation for 0 amount of users.
This balance must be broken, because it's ABSURD now.

And for the real question I want to ask:

I agree that current distributions set is not usable for audio professionals NOW. Why ?
I think, because most audio devices and software (incl. Creamware) consider the market profitability WRONGLY.
The line of thought that drives the manufacturers from linux is the mantra:
"there are far more windows users in the world (5% vs. 90%)"
I must agree: This is true. As true as that 90% of the body is water.
But this has NO USE to the subject.

For example:
Why do companies develop the applications in English first ?
There are FAR more arabic and chinese language speakers around the world.
How come you allow yourself to ship applications to China with a delay ?
The truth is that MOST chinese do not have a computer. and most of arabic speakers do neither, although that most people on earth are chinese/arabic speakers. And if some of the above DO buy a computer, they most certainly don't buy software for it , same stays for Russia.
And this is why the market that gets more attention is English speaking.
Because YOU - BUY.

The similar situation is with audio professionals.
Yes, there are more windows users around the world.
But how many of them are audio professionals ?
And how many out of audio professionals/musicians use windows?
I think the ratio win/mac is very different than total windows/mac usage ratio.

The interesting thing is that in particular SCOPE/PULSAR or UAD users are usually more technology embracing: look at you, you like to try new things.
This means the beans the market people should count are not from ALL the global users, but from audio professionals.

I think if somebody surveys this market correctly, they discover that MOST Pulsar users would go for linux, if the system would allow so. (Hmm.. considered the linux/gnu based operating system and most of related tools are free, it would not be such a difficult decision)

You may ask then: "Why the heck is there no linux or Mac support for SCOPE ?"

I can only guess 3 REAL reasons:
1. creamware management doesn't want to increase their workers base: more salaries to pay, more managers, more accounting complexities.
(IF this is the case... I don't want to buy their products )
2. creamware management simply bought the traditional mantra: "there are more windows users"
(IF this is the case... managers, wake up: count how many systems you have sold, and compare it to how may windows users are there around the world)
3. Maybe (very unlikely) Creamware has signed some kind of pact with Microsoft. Maybe. But, unlikely.

I doubt there are other reasons.
I mean "linux applications must be open source" - is a false statement, because it IS easy to make linux drivers without disclosing the proprietary details, via binary only OSS drivers. They can sign NDA and provide the drivers. Look at nVidia. they support linux, and don't disclose the code.
Same stays for many others.

And about "there are more windows users", true, but off topic, c'mon guys!

I want Scope, but I don't want Windows.

And I'm sure many would go for linux, the moment the hardware is supported.

Also note that GNU/Linux systems run both on macs and PCs, and other hardwares.

So... what's going on ?

Creamware can even develop a "Creamware Linux" that would run on any audio hardware

[hubird]

they won't. they need survival and resurrection, and Linux won't bring them any further at the moment.

[astroman]

...You may ask then: "Why the heck is there no linux or Mac support for SCOPE ?"

I can only guess 3 REAL reasons:
1. creamware management doesn't want to increase their workers base: more salaries to pay, more managers, more accounting complexities.
(IF this is the case... I don't want to buy their products )
2. creamware management simply bought the traditional mantra: "there are more windows users"
(IF this is the case... managers, wake up: count how many systems you have sold, and compare it to how may windows users are there around the world)
3. Maybe (very unlikely) Creamware has signed some kind of pact with Microsoft. Maybe. But, unlikely.

I doubt there are other reasons.
... it IS easy to make linux drivers without disclosing the proprietary details, via binary only OSS drivers. They can sign NDA and provide the drivers. Look at nVidia. they support linux, and don't disclose the code.
Same stays for many others.
...
And I'm sure many would go for linux, the moment the hardware is supported.
...

yes indeed, even me as (probably) the most sarcastic ranter would switch immediately to some kind of 'embedded-like' system version.

But that would be about it - I don't need all the fake stuff, I'd rather pay for original work.
I fail to see the latter entirely - ok, possibly I didn't dig deep enough...
My major point against Open Source is this 'but we can do it for free' attitude - Open Office is a typical example.

I mean what's up with these guys 'n gals ?
they can do their stuff beyond 'corporate' or business pressure, there are no deadlines, investors, shareholder-value-concerns and all they end up with is a copy of something boring that existed for years ???

you may have overlooked that the announcement for Linux support by the original Creamware is as old as the Pulsar One itself.
They had a full scale developement department back then and Linux was supposed to be a 'next big thing' so it's support was announced by many.
It simply didn't live up to expectations, and so other projects got a higher priority.

Regarding the protection of intellectual property your assumptions are just wrong. A (proper) description of a driver interface is perfectly enough to break into whatever.

After all it's a matter of resources versus revenus, and those are (currently) ruled by Vista support (for sheer marketing) and Mac support (if there ever would be a PCIe or external box Scope).

cheers and welcome, Tom
(nice post btw)

[garyb]

are we more about music production or cool computer stuff?

[bilbobugginz]

Regarding the protection of intellectual property your assumptions are just wrong. A (proper) description of a driver interface is perfectly enough to break into whatever.

Hi, Tom!
I didn't get the point.
The driver can be written in binary only form.
A special application (closed source) can use additional special driver, for using "not only audio interface" capabilities of the card.
And, if there are other functionalities the developers want to expose to programmers of 3rd party applications it's possible via special "scope-special-super-dooper-functions-library", closed source.

If you are saying the previous linux effort was unsuccessful for some reason, I would love to hear about this.
For all my attempts to ask "do you work on linux support" I got at a time the answer "we may do this in future, but we don't and currently do not intend to"

So, I don't know when you've heard about Creamware's announced linux support, I don't say it never occured, but ... I don't know when it occured.

And, let's say that I DO know how software is developed, and even am working on development sometimes.

They had a full scale developement department back then and Linux was supposed to be a 'next big thing' so it's support was announced by many.

I would love to find out why this happened, I mean never came to life.
Today, linux is far cheaper and easier to get working with than mac.

And I am sure whatever the reasons for the closing up of that department were, it was not due to the fact it was impossible to implement or impossible to license. Both these issues wouldn't simply allow these departments to be OPEN initially.

are we more about music production or cool computer stuff?

The thing is, GaryB: cool computer stuff makes eventually the music production MORE productive, more stable, with more capabilities and not less importantly - MORE FUN.
IF you are in the approach "hey guys, we're not into fun here, we've got work to do".

In terms of numbers,
if you could record much more recording channels per 1 recording it saves time.
If you can record with more effects and instruments, it allows you do either more complicated effects OR more effects.

And isn't the whole idea of scope/pulsar/luna - to squeeze the whole studio into a computer ?

I mean, multi track hard disk recording was once considered a complicated task, when operating systems didn't support time sharing properly, when hard disk had slower bandwidth,
and it was thanks to the "cool computer stuff" that software instruments were born.

Anyway, I didn't get the point of your remark, garyb.

[hubird]

it never came to life because Creamware got into difficulties 4 years ago, plus Linux didn't make the expectations true.
About the difference between mac and Linux system: just read this thread and you'll know that Linux isn't ready for the average music maker.
Mac works 'out of the box', believe it or not

[astroman]

that stuff is about 10 years old, so why would you even want to know ?
if you enter 'creamware linux' in Google you'll get this e-mail quote (which was already posted by Piddi a couple of pages back) and a few more.

They even searched developers via this forum, but in the end it was all trashed apparently due to intellectual property concerns.
Matter of fact is that Frank Hund obviously withdraw an NDA offer which was supposed to be signed by members of a Linux developement team outside of CWA. You probably have heard about Frank's enterprises since then

well, indeed you really don't seem to get the point... and it's not that easy to write about things that I actually prefer to stay out of public attention
{ censored by self }
If you're smart, you'll figure out the original content in brackets and why there is a reason to keep things closed.
It is a pure question of logic, you don't need even the slightest amount of technical information.
... it's all in the file right in front of you, agent Starling...

cheers, Tom

[bilbobugginz]

it never came to life because Creamware got into difficulties 4 years ago, plus Linux didn't make the expectations true.
About the difference between mac and Linux system: just read this thread and you'll know that Linux isn't ready for the average music maker.
Mac works 'out of the box', believe it or not

Hi
I am not saying that Mac is less ready than GNU/Linux. no. (FYI, I am writing this post from a Mac )

But I am saying that the fact Crameware got into difficulties doesn't change the profitability of the market. And it doesn't change the fact there are currently only 2 alternatives.

And you as user should be demanding the right to compare.

How can you know if Creamware is better than some Kyma system ?
only by comparison.

This is the idea of free market: competition.

And since Mac runs only on Mac computer, and you cannot legally put it on regular PC today, it leaves you simply no alternative.

Linux has the following advantages:
1. makes a more efficient use of computer hardware, because it is more configurable.
2. runs on all personal computer and server hardware I may need for audio
3. scales well for clustering, it means rendering complex effects can be done in realtime, even if you don't use SCOPE, you can stack up PCs, hook them up with a 1Gb LAN and there you go.

but it has the following disadvantages:
1. there are no optimized and easy to use audio distributions now. the existing ones are maintained by TOO techie people
2. the linux frightens commercial manufacturers (though it shouldn't), because they think you will have to expose your Intelectual Properties. This one is ... an imaginary problem. (Add to this, some linux developers are religious in keeping the source open, even if it makes no logical sense)

3. The usage needs to be supported at 1st by a professional, this one is real. Guidance etc. are a must in the beginning. and in music it is annoying.

Anyway, I hope I am only bringing up a HEALTHY discussion, not a flame war
Regards!