Blackout of 2003

C
Hi Donald

Simply make all your vendors (and the others who wish to sell you something) aware that you would prefer Linux solutions. Eventually they'll get the idea. Especially those whose products you don't buy.

Regards

cww
 
M

Michael Griffin

I belive Mr. Pittendrigh was referring to setting up new systems with Windows XP. How do you handle the problem of product activation? If you copy an image of an activated system to multiple computers, Windows will notice that it
isn't on the original computer any more and require re-activation. If you contact Microsoft with multiple re-activations from one copy, are they going to decide that you are a pirate?

Large customers solve this by using a corporate (site) license which they can image (there is no copy protection). However, it is contrary to the software
license to use your own corporate license to prepare a computer for a customer.

--

************************
Michael Griffin
London, Ont. Canada
************************
 
D

Donald Pittendrigh

Hi Curt

You know how I feel about this, it just damn difficult to get the Simatic manager and Step 5 programming package running on Linux.

Cheers
Donald P
 
D

Donald Pittendrigh

Hi All

Mr. Pittendrigh (who incidentally prefers to be called Donald) has had a look at this thread and can't find the context which has caused Michael Griffin to make the statement in the first sentence of the attached post.

For the record, Mr Pittendrigh's best moment last week, was when his teenage Son decided to take WinXP off of his PC (which also happens to be on Mr. Pittendrigh's home/office network), and replace it with Win2000. Mr Pittendrigh hates WinXP with a passion and finds Win2Kpro and Win2K server to be the only acceptable muckrosoft operating systems for use in industry at present.

Jokes aside, I have my web server running on Win 2003 server, I think I may become a fan of this product in the long run, but at present I have problems with this server which I never had with Win2000, or even for that matter with NT. The problems I had with the other Win2K2 and NT, I was able to resolve quickly easily and without reading tons of MCSE books, (well at least 95% of the time) At present I am considering re-installing 2003 for the 3rd time, to get active directory working properly.

I have absolutely no time for XP, I find it insults my intelligence everytine I need to do something more complex than open a powerpoint presentation, about a month ago it took me about an hour to figure out how to get files and folders in C:\Program Files, to be declared read write and authorized for me to make manual changes to some INI file or another.

There is in existence, a feature on Server, that allows one to set up a network boot and operating system install controlled from a central server. It is a fine solution for industrial machines to ensure accurate and consistent installation of the OPSYS, and is something I could even use in my business to cut the setup time of a new PC I am selling and reduce the competence level required to get an operating system installed and working. There would be some constraints which would have to be dealt with in order to use this feature to full advantage, but the issue which at present thwarts any incentive to utilize this feature, is that every time it was being used to set up a new machine, it would be necessary to determine if the operating system patches in the install partition were up to date or not. In fact one would spend so much time servicing this install partition, that it is more efficient to do a manual install for each new PC.

There are ways around, (and I don't mean illegal ways of circumventing), the WinXP licencing issue, the mechanism is described on the OEM support disk supplied with OEM distributions of the MS operating systems, the trick is in the setting up of "run once" software that prompts for the licensing information and other details on first time start up of the new PC. It also provides neat ways of customizing the operating system such as putting up a company logo on the windows desktop, and including supplier name and telephone details in the "about windows" general tag of the system properties panel.

If you have ever started up a new Siemens programmer, you will see the techniques employed to good effect.

Cheers Donald P
 
C

Curt Wuollet

If we make enough requests, sooner or later one of the vendors will see the light. And I really don't like using DOS or Windows software on Linux either. It's kind of a kludge and often no better than running it on those platforms. I want a port that runs native on Linux. And at times it's damn difficult to get the Simatic manager and Step 5 programming package to run on Windows :^)

Regards

cww
 
R

Ranjan Acharya

They just laugh when I suggest Linux. A few of them, like me, piddle around with it at home - to get all the peripherals working is quite a challenge - my shiny new laptop at home is stuck on Windows XP because the Wireless
Ethernet has never heard of Linux. The general expression there is that it is still not ready for the big time. I'm sorry, but I don't think that re-compiling my kernel just because I have an off-the-beaten path sound card is acceptable (in an unscientific anecdotal survey, I have not found a hard-core Linux user yet who has not re-compiled their kernel). Also each tool looks as though it was written by a different person (it probably was,
but they should be using some sort of guideline for consistency and help files should be written in reasonably grammatically correct English for the English "distro").

I'll be there when:

- It installs consistently in graphics mode and recognises all my peripherals
- The release set is stable
- Each tool in the system has a consistent user interface
- The big OEMs start to support it
- Re-compiling the kernel is not a requirement for day-to-day stuff
- SCO are bought-out by IBM and become a footnote to history

Until then:

- It is on an old Pentium II box at home for LegOS fun stuff
- All my customers remain with Windows or HP-UX

RA
 
J

Joe Jansen/TECH/HQ/KEMET/US

Ranjan,

I would like to comment on a few items in your message. Also, lets make sure that we hold Windows to the same standard that you want to hold Linux too. Only fair, right?

> is acceptable (in an unscientific anecdotal survey, I have not found a
> hard-core Linux user yet who has not re-compiled their kernel). <

This is because _they can_! Re-compiling the kernel to optimize it to the machine it is on is a HUGE performance benefit, which is why the "hard-core" Linux users do so. How do you get rid of the built in support for the stuff you don't need for windows?

> Also each
> tool looks as though it was written by a different person (it probably
was,
> but they should be using some sort of guideline for consistency <

Of course, windows software always uses consistent operator interfaces, right? Is that why I keep accidentally trashing stuff in Lotus notes when I hit the wrong key combo trying to send a message?

Windows has Common Controls DLL, Linux has QT or GTK. Beyond that, both have the same problem of people using there own set of design criteria.

> I'll be there when:
>
> - It installs consistently in graphics mode and recognises all my
> peripherals <

Agree, although for me, the only stuff it doesn't recognize is stuff that windows won't either.

> - The release set is stable <

Defined as? By stable do you mean that they will not release another version? Or do you mean that it won't need to be patched? If that is the case, windows has a long way to go.......

> - Each tool in the system has a consistent user interface <

Both have this problem.

> - The big OEMs start to support it <

Like who? IBM? Microsoft won't, of course. I think IBM counts as "big". Cisco? I would like to see your list of big OEMs, if IBM doesn't make the list.....

> - Re-compiling the kernel is not a requirement for day-to-day stuff <

Already done. Although re-compiling will give you a major performance boost, you are obviously free to not do so. I have systems that have never recompiled, and they run just fine.

> - SCO are bought-out by IBM and become a footnote to history <

Doubtful... I think IBM will wage this fight long enough to kill the company.

> Until then:
>
> - It is on an old Pentium II box at home for LegOS fun stuff <

This is where the Linux doubtfuls always get me. They put Linux on a crappy old machine, and then compare it to XP or Win2K running on a 2GHz P4. Then they complain about how slow linux is. Give me a break! here is an experiment: I will put win2K on my 600MHz P2, and carefully craft a RedHat installation onto the 2GHz P4. Do I get to complain about how worthless windows is then?

> - All my customers remain with Windows or HP-UX <

Good for them. Did they evaluate the options? If so, and they made an informed decision, then they should be happy with it. If they did not bother to compare the two tho, or worse, you did not bother to give them the information and take the time to offer a *fair* comparison, then they
have been short-sighted or short-changed.

--Joe Jansen
 
C

Curt Wuollet

Hi Ranjan

This attitude puts us squarely in the "chicken or egg" problem. Many of those things you demand simply can't be done without vendor cooperation. The drivers for the thousands of different peripherals are written for MS but seldom for Linux. This is changing as Linux becomes an important market, but making it important to have Linux drivers can only be accomplished by the community. You are opting out of that community. And, no doubt buying products that support only Microsoft. The Wireless Ethernet should know about Linux because I believe this was working on Linux before it was supported on Windows. Buying a product that supports Linux would solve your problem and move us forward at the same time. There is a great deal of synergy there.

And if your "off the beaten path" sound card vendor considers everything top secret and won't release the information to write drivers, you're asking the impossible from the community. Again, making Linux support a criteria in your selection will help much more than you might think. And I doubt that you would have to sacrifice anything to do it as Linux support tends to be available for two classes first, the cheapest, and the best. That said, I haven't had a sound card problem for quite a while except for notebooks and extremely stingy integrated MBs where they only give you half a soundcard and burden your processor with the rest. Sort of like WinModems and other Winjunk. I'm about as hardcore Linux as you get and I'm running three stock kernels across a multitude of different PCs The interface consistancy problem is being addressed with KDE and to a lesser extent GNOME. For my own part, I prefer the UI to fit the program, not some framework copied from a word processor, but that's personal preferance.

Oh, and all the big OEMs do support it, including HP and especially IBM. There is only one holdout that doesn't support it and probably never will. Unfortunately, they have most of the market by the short hairs for the moment and can thus exert life or death pressure on all except the biggest companies. But their power is waning and slowly, bit by bit, the monopoly is crumbling.

So, by your actions, you can make your criteria occur, or not. I, and a lot of other folks, think it's very worthwhile to enable competition once more. You may think your little bit doesn't matter. But it does. Progress tends to be almost linear towards an inflection or tipping point, after which it becomes exponential. We have almost enough of a community to push it over. Just a few more people that care will make it happen. And even if you never move to Linux, the competition will be of great benefit to all. I wouldn't want to see a Linux "monopoly" either. Monopoly and a monoculture have been demonstrated to be bad for consumers and security and the market in general. Diversity will get things moving again and prevent the strangulation of new and better ideas. Then we can judge which are better by comparison rather than speculation. Like I say, I'd be happy with RSLogix for Linux 1.00 even if it doesn't run any better than the Windows version.

Regards

cww
 
R

Ranjan Acharya

FYI, I prefer Linux over Windows, but from the point of view of Factory Automation I am stuck with Windows.

R=RA
J=JJ

On October 6, 2003, Ranjan Acharya wrote:
R>>> is acceptable (in an unscientific anecdotal survey, I have not found a
hard-core Linux user yet who has not re-compiled their kernel). <

On October 8, 2003, Joe Jansen/TECH/HQ/KEMET/US wrote:
J>This is because _they can_! Re-compiling the kernel to optimize it to the
machine it is on is a HUGE performance benefit, which is why the "hard-core"
Linux users do so. How do you get rid of the built in support for the stuff
you don't need for windows? <

Yes, but that makes things too complicated. Out of the box installation is what users want. My customers don't want to deal with re-compiling the kernel. When you are a "power user", you forget what it is like for many people who either cannot be bothered (i.e., they could figure it out, but why should they) or just cannot (i.e., their expertise is in other areas).

R>>> Also each tool looks as though it was written by a different person (it
probably was, but they should be using some sort of guideline for
consistency <

J>Of course, windows software always uses consistent operator interfaces,
right? Is that why I keep accidentally trashing stuff in Lotus notes when I
hit the wrong key combo trying to send a message? Windows has Common
Controls DLL, Linux has QT or GTK. Beyond that, both have the same problem
of people using there own set of design criteria. <

True. Common User Access and so on died a long time ago. But there is still a consistency I "feel" when I use Windows that I just do not have with Linux. The reason for this complaint is from when I was setting up some
networking tools with Linux. I had software from Sweden and North America - both were wildly different and it was a royal pain to get it all working.

R>>> I'll be there when:
R>>> - It installs consistently in graphics mode and recognises all my
peripherals <

J>Agree, although for me, the only stuff it doesn't recognize is stuff that
windows won't either. <

That was not my case; I had to re-compile my Kernel just for sound. Then the next time I got a Kernel patch ... and so on. I found it annoying and ended up throwing in the towel and buying a new sound card. Lately, I have
found Windows quite acceptable when recognising my peripherals. Over the last few years, with "normal" peripherals I have not had a case of device in-fighting (thanks to Ethernet, USB and FireWire, the old IRQ battle has finally gone away - but that would hold true for Linux too). I don't have hundreds of peripherals beyond the USB tower for Lego, a camera, keyboard and mouse. I don't even have a printer at home.

R>>> - The release set is stable <

J>Defined as? By stable do you mean that they will not release another
version? Or do you mean that it won't need to be patched? If that is the
case, windows has a long way to go....... <

True, also Linux patches are much easier to apply. But then I prefer Linux. The only down side is the perception of release du jour. Also, stability means that these Linux OEMs have to get their collective act together or go
the way of Unix (Solaris, AIX, HP-UX ....).

R>>> - Each tool in the system has a consistent user interface <

J>Both have this problem. <

Agree

R>>> - The big OEMs start to support it <

J>Like who? IBM? Microsoft won't, of course. I think IBM counts as "big".
Cisco? I would like to see your list of big OEMs, if IBM doesn't make the
list..... <

You forget, we are only talking about Industrial Automation. This is Control.com after all! IBM indeed does NOT make the list of automation OEMs.

-Rockwell Automation
-Schneider Automation
-Siemens
-CiTect
-GE-Intellution
-Honeywell
-Invensys-Wonderware-Foxboro
-Omron
-Toshiba
...


R>>> - Re-compiling the kernel is not a requirement for day-to-day stuff <

J>Already done. Although re-compiling will give you a major performance
boost, you are obviously free to not do so. I have systems that have
J>never recompiled, and they run just fine. <

Not in my experience or that of all the Linux users who work for the same employer as me. I'll wait and see.

R>>> - SCO are bought-out by IBM and become a footnote to history <

J>Doubtful... I think IBM will wage this fight long enough to kill the
company. <

Good. Until then, another AMD - Intel brawl.

R>>> Until then:
R>>> - It is on an old Pentium II box at home for LegOS fun stuff <

J>This is where the Linux doubtfuls always get me. They put Linux on a
crappy old machine, and then compare it to XP or Win2K running on a 2GHz
>P4. Then they complain about how slow linux is. Give me a break! here is
an experiment: I will put win2K on my 600MHz P2, and carefully craft a
RedHat installation onto the 2GHz P4. Do I get to complain about how
worthless windows is then? <

Whoa! Where did that come from? I never once complained about Linux being slow or mentioned P4. Give ME a break! The reason it is on an old machine is because at home my wife has the shiny new P4 laptop that runs Windows XP - not out of choice, but necessity (the WAP hardware from Netgear does not have Linux drivers, she was using Linux before that). I bring my own P4 laptop home from work if I want to use a performance tool - Linux or Windows. Read what I originally wrote! The LegOS is for F-U-N on an old P2 box (I have an old P1 box that came with Windows 98 on it for my eldest daughter, she is four and only cares if the Internet is up for on-line fun, the 486DX2 is up on a shelf and still has DOS on it). The P2 runs on Linux
because the old box would never run a newer version of Windows, but runs Linux acceptably well. Also, LegOS requires Linux.

R>>> - All my customers remain with Windows or HP-UX <

J>Good for them. Did they evaluate the options? If so, and they made an
informed decision, then they should be happy with it. If they did not
>bother to compare the two tho, or worse, you did not bother to give them
the information and take the time to offer a *fair* comparison, then they
have been short-sighted or short-changed. <

Options? What options? They had no choice. None of my customers are silly, they choose to work with large automation OEMs with years of experience, engineering staff, good product selection and so on. They come to me with a standard platform - Siemens, Allen-Bradley, General Electric and so on. What is this "you did not bother to give them the information"? Pointing fingers and making accusations like that is a little bit rude and very daft - what information? - you know perfectly well that the big OEMs are in bed with Microsoft. Is there a release of RSView/32 for Linux from Rockwell? Did Siemens announce an S7 programming tool for Linux? As pointed out again and again and again by many postings to those very keen
fans of Linux who think we can switch right now; we use the OS we have to use not the OS we might want to use. Let me know when CiTect have their Linux version out eh? Until then, let me suffer in peace.

At the end of the day, Windows gets the job done. Linux would too. More choice would be nice. I integrate solutions based on tools available in the market place and based on directives from my customers (sometimes given with advice from me, sometimes not).

RA
 
R

Ranjan Acharya

Curt,

I was wondering if you were reading this thread, as I said to Joe, I'm stuck with Windows at work and a little bit at home (one machine has Linux, one machine has Windows 98 and the other has Windows XP) for various reasons BEYOND MY CONTROL (I am a fatalist - I know I can't fix everything - Linux can't find those pesky WMDs in Iraq can it?). I just don't have the time to re-write CiTect (for example - fill in any large industrial OEM here) to work with Linux (I don't have the source code either), plus I have not written in C or C++ for several years.

I don't care about those "big OEMs" like HP and IBM supporting Linux. I care about big OEMs in the industrial control field supporting Linux. Until then, I use Windows at work.

RA
 
G

George \(Jim\) Hebbard

Where the action is:

Sixnetio.com 40% Sales Growth Uses Linux

China. Going Linux with a bang!

See Article, below
___________________

Blackout of 2003 - a large part of the perceived problem is time-based data analysis and state-communication. With recent TC/IP security breaches, I bet a lot of the new SCADA WAN infrastructure will go Apache (Linux) whether the above Western Vendors like it or not.

Article: Per Joe Feeley - Publisher, Chemical Processing magazine, September 2003

"When there's any opportunity to stop for a minute, catch our collective breathes, and take a hard look at our industry, It's clear that there are many, many subjects we should talk about inside the covers of this magazine."

"Job #1 is providing information that helps you do your job better. But, from time to time, there are important issues that involve something more than that. "

"Like what, Joe? Well, of late, I'm hearing more and more about the loss of US-based technical jobs to countries such as China and India, in a fashion similar to what's happened to the IT world."

<snip>

"The design work for that new sulfone polymers process in Georgia or that more energy-efficient stem cracker operation in Louisiana is being out-sourced to countries where process engineers earn one-third of what their US counterparts do. The trend is no longer a trend, it's an
expanding fact of business."

_____________________

Me: What part of the re-vamping of our transmission infastructure is subject to outsourcing to Siemens. Note the reference to Siemens PLC and Linux TRU's in
http://www.ats.nl/press_releases/sixnet-versatrak-11-2002-gb.pdf (Sixnet again.)

Two years ago I was stopped by a sweaty pump operator - non-control room type- in a plant in India. He asked me a technical question and I asked him where he got his engineering knowledge. He told me he was a graduate
chemical engineer, and that India graduates much more technical talent than they can absorb domestically. He said lots of his friends had to settle for laborers jobs, and that he was lucky. Six years ago in China I was beginning to see the same trend - except China CAN absorb much of their talent.

Hey! maybe the Indian laborer/engineer would be willing to do controls design work for $10/ hour. Much more than he's making now...

The writing is ALL OVER THE WALLS. We got a hurt coming unless we do new things in new ways....

=>GMH<=
 
A
Curt,

As someone who's writing Linux applications right now for a Linux HMI, I'll tell you one major reason why no one is using Linux -- there are basically _no_ drivers for any industrial automation hardware. In the Windows world, there are numerous companies that sell OPC and COM and
.NET and other acronym-compliant drivers to talk serial, Ethernet, or whatever to PLCs and the like.

In comparison, I look at http://www.linuxincontrol.org/, and I find none of those projects dealing with the constant question... "How do I talk to my PLC from my Linux box". Nothing. This isn't something done on
Windows by AB or Siemens or what have you. This stuff is all done by small companies with tens of employee, and I can't find anyone, open source or not, doing this for Linux.

You have a GPLed Modbus driver in the MAT project. I know of Ron Gage's AB Ethernet library also, and there are probably a couple more out there, but they're all wildly different projects. Omron Hostlink for Linux? GE SNPX for Linux? Nope. Even then, customers on Windows platforms can call up their vendor and go "THIS STUFF DOESN'T WORK, FIX IT", while the Linux answer for the free software is "fix it yourself".

This is a problem. I am writing most all my own drivers for all the devices I have to talk to, but I'm a programmer and that's my job. Integrators won't do that, they want to get the job done, and as long as your answer is "write some code to talk to a PLC", then most people
won't be using Linux to replace Windows, because they don't have the time or skills.

Alex Pavloff - [email protected]
ESA Technology ---- www.esatechnology.com
------- Linux-based industrial HMI ------
-------- www.esatechnology.com/5k -------
 
C
Hi Ranjan

Reading you loud and clear here.

I'm not suggesting that you rewrite Citect. I'm suggesting that, if you would like to see a Linux version, it's more helpful to demand a Linux version from Citect (or name your favorite vendor here) than to simply wait for it to magically happen. If everyone just waits, it will never
happen. But, Linux users, acting together are making some rather remarkable things happen. Almost all of the recent events were held as impossible by those who simply wait. And the solution to the problems you mention is the same. If the community makes it important to vendors, it will happen. They will release Linux drivers for your hardware, they will port their products to Linux and it will no longer be impossible.

Regards

cww
 
R

Ranjan Acharya

Yes again, all wonderful news, China tell Microsoft where to go and so on. Some users implement Linux.

Again, it all simply puts the ball back in the court of Allen-Bradley / Rockwell, Invensys-Wonderware-Foxboro, Siemens, General Electric-Intellution, Schneider Automation-Modicon-Telemecanique and so on. This is where I integrate.

I suppose they will use Linux when their commercial model allows it.

Until then I wait.

RA
 
D

Donald Pittendrigh

Hi Ranjan and Curt

Should you happen to re-write Citect, please let me know where I can download your source code............. for free.

Regards Donald Pittendrigh
 
C
Hi Alex

Lets say that I had a secret stash of drivers using Open Protocols to avoid being stomped into dust by the monopoly beyond the law. Everything you could imagine, with a core commonality so translation would just be a matter of loading a DeviousNet and a PreppyBus module. And I could talk to every PLC beause they had support on their end for my Open protocols. That would be Nirvana. Arguably possible and certainly well within the technical capability of the OSS community.

Now let's get back to reality. I can't even get Modicon, who obviously wants Modbus to be considered an Open protocol, to give me a clear go ahead to popularize their protocol by making it available free of charge or commercial interest. I can only imagine they would like to retain the right to do to us what SCO is trying to do to Linux. If I won the lottery, I could perhaps license a few, but even there it's highly unlikely that I could release code under the GPL. Which is pretty much the point of our existance.

These little shops and widget makers have what they need to build and sell interfaces of various types. They have paid and, no doubt continue paying for those rights. And since they maintain secrecy and sell only on the terms in their licenses, their continued existance is tolerated. We cannot, within the limits of our charter and scope, operate that way. We need to find another way that involves cooperation from dozens of for profit corporations. This has not been forthcoming. So, do we reverse engineer their IP? That would give us part of the puzzle and very likely dozens of lawsuits. Can we force them to provide register access and the like? No, we probably can't. Can we then simply emulate and use OPC, .NET and the other peepholes provided by agreement between monopolists? Well, yes we could, with even greater legal exposure.

It will require a change of paradigms, and a change of heart for us to be able to provide that Nirvana. I feel the OSS community is the right vehicle if that Nirvana is to ever come to pass. We've seen that private consortia and even international standards bodies fail miserably in even the first faltering steps toward cooperation. The OSS community, more any other entity, owns the cooperation paradigm. World Wide. And even big corporations play by the rules as it is in their best interest to do so.

It will come to pass, but not until the rest of the world is eating our lunch, having rejected Microsoft and moved ahead on common Open ground. The current market leaders may well be in the position of the guy who makes the best yardsticks in the world, but who can't understand why they just don't sell anywhere else but the US. They just don't get it. :^)

Regards

cww
 
G
On October 13, 2003, Ranjan Acharya wrote:
> Yes again, all wonderful news, China tell Microsoft where to go and so on. Some users implement Linux.
>
> Again, it all simply puts the ball back in the court of Allen-Bradley / Rockwell, Invensys-Wonderware-Foxboro, Siemens, General Electric-Intellution, Schneider Automation-Modicon-Telemecanique and so on. This is where I integrate.
>
> I suppose they will use Linux when their commercial model allows it. <

Exactly. For most vendors, support for any given feature, protocol, or operating platform is an business decision driven by the market. What Curt's been championing is a deliberate effort by the market to influence those business decisions in a particular direction.

> Until then I wait. <

As you should. For you, the tools are simply a means to a very concrete end, and the tools you have adequately meet your needs. As far as I can tell, there is absolutely no incentive for you to move to Linux until the choice is not only painless, but painless with clear benefits.

For most users, the choice of what product to use is also an economic decision, in the broad sense of maximizing return within the user's own utility function. (Well, stated that broadly, every choice everybody makes is an economic one. :^) I phrased it that way because everyone choosing from among the same products isn't using the same utility function.)

For many or most automation integrators, the utility function consists largely of maximizing profit, which usually implies using readily available components with which they and their customers are already familiar and comfortable. Today, that means Microsoft-based software for a very large percentage of the market.

Some integrators - and their clients - are willing to absorb an additonal cost (expertise, effort, local rather than outsourced support) in order to create more elegant, stable, scalable and maintainable solutions. Of those, some are using Linux to realize their goals. That's true for me and most of my clients.

If we are successful (and I believe that we are being successful), the market will, over time, perceive a reduction in the cost/risk of using Linux. More people with high enough risk tolerance - or high enough pain level with their current situation - will implement automation systems on Linux, the market for Linux-based automation components will grow, and eventually the big automation vendors will see a large enough potential user base to make it worth their while to provide their tools on Linux.

Even then, I don't expect to see a wholesale abandonment of Microsoft. The learning curve for using and administering Linux will still be a cost, inertia on the part of integrators and their clients will still be a barrier, and Microsoft's marketing ability and market clout will provide other real and apparent benefits to using their products.

We've seen exactly this course of development in other areas of computing, particularly the realms of network infrastructure and internet services, which (not coincidently) more closely resemble automation integration than does anything that happens in the office on a desktop. I don't see any reason why this won't play out in industrial automation, too.

Regards,

Greg Goodman
Chiron Consulting
 
D

Donald Pittendrigh

I would just like to correct the statement made below, Siemens does support Linux, and several flavors thereof, and also two flavors of unix, for which the softnet driver source code can be purchased for Ethernet or profibus.

No I don't give out part numbers, and yes I do sell the packages, or should I say I would love to, at the price.

Regards
Donald Pittendrigh
 
C
If I were to rewrite it, or actually reimplement it's functionality with OSS tools and help from the community, you can bet I would let you know where to find it. GUI programming not being my cup of tea, I would suggest you watch the folks who are interested in OSS similar to Citect. I have played with the idea of something similar to Panel Builder or Quick Panel for use on inexpensive Linux based Panel PCs. Of course, Since you can't easily use a Panel View with GE or Quick Panel with AB, getting them to open up to an OSS project would be problematic.
As the public becomes more educated and proprietary and closed become dirty words, this will change. It would, at present, be a worthwhile addition to the MAT project, so I think about it off and on. My passion lately has been to make an Open, fully documented, PC compatible, industrial quality PLC for OSS (and perhaps other stuff) to run on.
I've had health and financial setbacks to overcome, but I should be back on track this winter.

Regards

cww
 
Top