Muse Research RECEPTOR - Here's an idea for Creamware

symbiote · Post by **symbiote** » Sun Aug 29, 2004 7:55 pm

corporate code isn't really exactly clean all the time either. the only difference with open source, is that you can't see it, nor fix it. otherwise it's just humans pushing bits around.

innovation wise, i guess you could look at freebsd/openbsd. even more than that, i'd venture that without bsd and the bsd licence, windows wouldn't be what it is today. sadly, with sp2 released, i think we've seen the best of windows. gonna go all downhill from now

blazesboylan · Post by **blazesboylan** » Mon Aug 30, 2004 12:39 am

Wow. I should really not post messages when I'm tired.

A few stray thoughts:

Dehuszar, my opinion on M$'s audio "standards" is that they are far less important than those provided by Steinberg (VST, ASIO). M$ is on just about every standards committee in the universe (W3C, OMG, ...) but they go "unilateral" when it comes to audio (Direct-Crap).

Linux is increasingly difficult to keep clean. I have been very disappointed with its direction over the past 5 years -- toward a "bigger" more "inclusive" OS. Having a webserver and audio engine built into the kernel is not small or clean.

For what it's worth, Mach 3 was (along with QNX) another very small clean OS.

On all your other points about Linux being a very useful OS for DSP programming, I agree 100%.

Now to the followup on my unfortunate half-awake post...

You missed my point about Amazon and EBay, Dehuszar. (Easy to do since my post wasn't exactly lucid in the first place.)

The code that maintains the data is irrelevant. Tim O'Reilly's point was that the data is inaccessible. Therefore, if you want to be able to access it in a way that is different from the interfaces given you -- you're screwed.

If you've ever heard Stallman give a speech, this is exactly what motivated him to produce Free software -- free as in speech, not as in beer, etc -- he wanted to be able to access printers in whatever way he found useful and bug-free. Not in the way that HP told him he should -- often using a driver which was buggy and would not be fixed for months or years.

The idea with data is similar. If you want to access EBay or Amazon with your own interface, tailored to your own needs and automating the things you find tedious, you can't do it.

O'Reilly didn't suggest that this problem is widespread at the moment. But he was predicting a future in which basically all of the data we need to use is, in fact, painful to use in daily life. I don't necessarily agree with him, but I definitely see the potential. Hmmmm, I was just thinking about being able to use Yahoo! Mail without the bloody ads...

As to these questions:

I'm not sure how bash/perl/python/etc scripting and general package management aren't automatable interoperating modules... And do you feel that Windows is well automated and modular?

Which "general package manager"?

Scripting is more than 30 years old. Perl ("Pathologically Eclectic Rubbish Lister") is hardly a useable programming language. Python and Tcl are great. So are awk, sed, emacs, find, grep, ...

But what software has been developed in the Open Source community recently?

Mozilla
Open Office
Gnome

That's right, GNOME SUCKS. If I posted this comment in a Linux newsgroup I guarantee I would be ripped to shreds. Everyone loves Gnome but it sucks. It sucks it sucks it sucks.

I have 2 pieces of software on my Windows machine (besides Tcl, mind you) that are automatable:

Sony Vegas 5.0 (Javascript scripting)
SFP 4.0 (MIDI-automated)

Windoze sucks every bit as much as Gnome. I might even go so far as to say that Gnome is worse! If you agree, great. If you disagree I would be happy to debate it at length via PM or email (jtienhaara AT yahoo DOT com).

Astro -- the idea that Apple had the best developers in the industry in the 80s doesn't make sense to me. Companies like Silicon Graphics, Inc., Sun Microsystems, and various universities, were lightyears ahead. Apple perhaps had the best "application" developers. But even Commodore, with the Amiga, beat them in the best "home computer environment" category IMO.

Nevertheless you will not see anyone argue that Microsoft has ever had the best programmers for anything. And I would certainly agree that Apple had a lot of great programmers, whether or not they were the most innovative!

As for the Linux port: CreamWare should hurry up. It already looks like 2005 will be a hot year for Linux devices. It's time that CreamWare had an option out there for developers to put into rackmount boxes for effects, recording, synths, ... An option which can be configured however a 3rd party company needs, in order to sell to its own market.

Small and modular and interoperable. The way to go IMO.

(Huh. Looks like I'm just as bad at writing when I'm awake... Oh well, $0.01...)

Johann

astroman · Post by **astroman** » Mon Aug 30, 2004 6:05 am

you already guessed it, Johann. My statement was indeed mainly about application programming, though there were also some great achievements on the not so obvious (for the user) OS and interface parts.

The point wasn't how good they were (it's past anyway) but the massive change in direction - together with a clear decrease in quality of their own developements.
They still manage to camouflage this fact under brilliant marketing and 'fake' innovation

The quality of M$'s programming is beyond any discussion - afaik they never ever developed ANYTHING on their own.
Starting with Bill's first deal to buy it's infamous Basic interpreter from a college friend and ending with the takeover of Connectix for their VirtualPC app

But if you have 'JAVA' in mind when mentioning SUN, then I suggest a peek at an old Apple app called HyperCard, which contains a programming language that can easily be identified as the JAVA blueprint.
Just more simple and straight forward...
True, it's outdated, but still worth mentioning when it comes to basic concepts of programming languages.

Interestingly Apple released it as a free tool for everyone to program, delivered with the OS that accompagnied the machine.
So it wasn't taken serious by professionals and treated as such...
It's a brilliant piece of code including an object oriented language that auto-declares it's structures and types in just a few 100k of machine code. It outperforms JAVA by at least a magnitude of 10

If they would have released it as a professional SDK the history of programming languages would have been different...
On the other hand - who would need new hardware if the tools work great on a current system

cheers, Tom

blazesboylan · Post by **blazesboylan** » Mon Aug 30, 2004 11:30 am

Hi Tom.

Actually I wasn't thinking of Java (it was crap when it first came out -- they've finally cleaned up the libraries with Java 1.4). But then I wasn't thinking about HyperCard, either. Interesting stuff!

Even though Apple / Mac has become a huge marketing force, they still have some relatively recent innovations under their belt. Particularly FireWire. I'm glad they "gave" it to IEEE -- otherwise I wonder if Z-Link would exist now?!?

Let's hope that another company close to both our hearts puts out its freely available SDK sometime soon...

Cheers,

Johann

dehuszar · Post by **dehuszar** » Mon Aug 30, 2004 2:22 pm

Well, to an extent, I agree... gnome is somewhat irritating, but it's (in my opinion) much more usable than KDE which is beautiful but becoming corpulant. However, the real importance is that you can use whatever you want, like fluxbox, windowmaker, etc. And all you need is the minimal library packages of whatever window manager you don't want to use for general machine tasks, but which are useful for programs designed for use in a particular manager.

Even more, I think many people are drawing the same conclusions insofar as bloat and over-featuredness are concerned. Look at how XFree86 has been dumped for trying to force the community in a particular direction and now X.org is in the process of breaking their current monolithic release structure into a more modular inter-connected series of parts in response to peoples complaints regarding huge mess releases of both X-systems.

Or even concerning Mozilla, how Firefox, Thunderbird, and nVu are now replacing the rather tubby Mozilla (which even in spite of it's paunch renders faster and safer than IE).

In response to a few of your statements:

" If you want to access EBay or Amazon with your own interface, tailored to your own needs and automating the things you find tedious, you can't do it. "

--sure you can. All you need is some mastery of SOAP and HTML/XML and you can create your own interfaces to mine information out of their databases (that they make available anyway ...I'm a whitehat kind of techie). You can take those results and use them in a number of ways and automate that data in whatever way is useful. Now you may not be able to turn around and sell such a tool, but the technical capability is there.

If fact one of my personal projects/coding goal-posts is to create a personal shopper which mines various shopping engines (like PriceGrabber, eBay, etc.) for product keywords, creates a product profile (shopping list) and can send notifications when certain items get below certain prices. There's endless possibilities if you understand the interface you're using.

Think of the Mozilla mycroft search engines. Really simple examples of using tools outside of their original context.

You may not have author/editor access to anything, but if it's web based (or at least port accessible) and non-encrypted, you can generally create whatever tool you want to access a system and parse out what ever results you want in whatever way is useful to you. In such cases, the code is indeed very important because you can look up object and script libraries, converse with communities, and share ideas freely.

Anyway, this is getting very far away from the discussion of music, but I'll rest my case with one final thought. The point is not where things currently are and what's good or not about them. The issue is where things are likely to be in 5 years and what influence do you have in shaping the direction of those tools, ideas, and resources that matter to you. Whether it's code contribution, vocal or monatary input, any contribution, so long as it's measured in it's delivery and reasonable in its expectation can make a difference. I don't see such contributions having the same effect in the commercial world except in a few marginalized ways.

Most large companies will only respond to overwhelming backlash, and the backlash usually only comes when a new release appears. Smaller companies are usually required to have their ear to the ground if they intent to stay fed. These generalizations are just that, not always the case, but on the average, about how things play out. While certain requested features often arrive, they can sometimes arrive in such an awful manifestation that the original purpose of the program gets buried under a landfill of poor ideas. Such ideas are often born from the want to "expand their market" in some fashion. I'm thinking particularly of the awful beast that EasyCD Creator has become, and only for the sole purpose of making it useable by idiots instead of creating a medium between ease of use and customer educational materials.

From where I stand, CubaseSX is an amazing compositional tool. Like Microsoft word (or OpenOffice.org too, I suppose) I'm part of the 80% which only use 20% of the program. I usually am flexing the synth/instrument end and not straying too far beyond the piano roll, basic wave editing, and some time-stretching as I am hashing out the feel of a new song. Midi automation is crucial, but just about every major player (including open-source sequencers) have some form of automation.

I could quite reasonably be using anything to do these tasks. But the problem comes in when CubaseSX sits on top of, and develops it's engines and APIs to interact with a fundamentally flawed foundation (from a realtime-multimedia perspective anyway).

And I don't think that Steinberg or TwelveTone (Sonar) will stay out of the Linux world if enough people are using it. Remember that they had a BEOS version which was more than halfway complete before BEOS had a parking cone shoved up its arse (otherwise known as the M$ special). Sonar especially may be able to gain some marketshare as WineX has been doing an amazing job porting DirectX and DirectAudio APIs to Linux. If nothing else, it's a stellar head start.

There. I've said it. Shoot me.

(actually don't, I'm just being dramatically fatuous)

Sam

<font size=-1>[ This Message was edited by: dehuszar on 2004-08-30 15:48 ]</font>

astroman · Post by **astroman** » Mon Aug 30, 2004 3:55 pm

On 2004-08-30 12:30, blazesboylan wrote:
...But then I wasn't thinking about HyperCard, either. Interesting stuff! ...

re-thinking it, I've been focused a bit on the programming side of that app, but it has been even more relevant for future developements than one might think.

Actually Hypercard is the paradigm of the typical website layout we've all become so used to.

there's the file called 'stack' - your site
the 'stack' is built of 'cards' - the pages
you have scripts linked to screen (and data) items
and finally, (hence the name) there are hyperlinks, 'sensitive' text that drives the navigation through both the current 'stack' and external ones, just like you switch websites by clicking links.

I'm not shure but they probably released that stuff around 1987...

sorry for the OT, but I found that part of history worth mentioning

cheers, Tom