Open Control

M

Michael Griffin

I don't know exactly what Mr. Torvalds does at Transmeta either (I've never inquired) and I don't particularly want to get in on this particular argument. However, I thought you may want to know that Transmeta's product probably involves more system level software than it does actual hardware. The idea is to run a series of very sophisticated virtual machines on very
simple RISC hardware.

If you are designing a new large scale microprocessor, you need to have system software people either on staff, or under contract (or some other arrangement). In the old days, you just came out with the hardware, and perhaps a simple assembler. These days you need to port compilers and at least one popular operating system to your hardware. To cut the time required
to get the product ready for market, you would have the software and hardware teams working in parallel.
In addition, most RISC designs are optimised by having the compiler design help steer the hardware design. There used to be a saying that "RISC" stood for "Relegate the Impossible Stuff to the Compiler".

These companies need to have some people working on software to make their product (microprocessor) saleable, but the software isn't
really their main line of business. It's more or less a loss leader to get their hardware designed into products.


************************
Michael Griffin
London, Ont. Canada
************************
 
Michael Griffin:
> > Would you (or anyone else) have a rough idea of how many man hours
> > is involved in writing a usable soft logic system? Assume the
...

Alex Pavloff:
> You know what's funny? I HAVE this system. There was a company
> called Controlware (bought by Allen Bradley and then killed) that did
> a soft logic engine that could run ladder under DOS and fit in a nice
> small system. Their documentation was alright, but they were too busy
> to support me (working on the Next Big Thing), and as a result, I've
> got bugs that I can't find or fix.

Yeah, that's one advantage of Open Source - if the original authors lose interest, someone else can always take over. If worse comes to worst,
you can take over yourself, or at least fix bugs that are in your way.

> So this Controlware system that I have now -- a small, DOS based
> ladder editor and soft logic engine, took about 10 man years to
> develop with full-time developers.

My friend, if you've been charged 10 man years for that, you've been royally shafted.

Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
A
Curt Wuollet wrote:
> If you look back a few years they were also doing it when they weren't
> paid by anybody. The argument that you must be paid to code a major
> project is absurd. What is your basis? If I want to sit down and write
> code right now, do i need to find someone to pay me? If I decide to give 4
> hours a day to a project, is there anything preventing me from doing so?
> I'm not sure what you're driving at. If I can do that, logically, everyone else can.
> Would the work done evaporate?

I think we have a fundamental disconnect in what we're thinking the end result of a project is going to be. I'm coming from a position where I want something that I can put out in the field fairly soon to get things done. Free would be nice, but I'm more concerned with getting the job done.

You appear to be coming from a position where it really doesn't matter when things get done, as long as they're free when they get there.

Your constant rhetoric against other companies shows that you do not like the current situation in automation, yet so far, you have not produced
anything that would cause a drastic change in that situation besides multi-megabytes of sometimes valid, something overblown screeds against most every large industrial and computer company.

You've been doing this for over two years now. You may not be the best programmer in the world, but if you had spent 4 hours a day working on your
project, I think you'd have accomplished a lot more. (As well as getting better at programming.)

Since your project hasn't had any design wins in the real world and hasn't registered as a blip on anyone's radar screen, I'm forced to sit here and
figure out why not. I compare your project against sucessful projects (e.g. GCC, Linux Kernel, XFree86, Gnome, and KDE), and note that they have developers that can concentrate full time on their projects. (GCC people at RedHat/Cygnus, XFree86 people at HP Labs, KDE people at SuSE, etc).

I look at some of the GPL'd libraries that I'm using and have contributed to, like CommonC++ ( "http://www.gnu.org/software/commonc++/":http://www.gnu.org/software/commonc++/ ) and see that the main developer works for a company. "http://www.ostel.com/":http://www.ostel.com/ They're doing free software and don't appear to have gone under yet. They're small, but their main developer can spend most all his time on Bayonne and CommonC++. Then I can contribute stuff that I need, and we're all happy.

You also say that "Linux wasn't done with full time people". You're wrong.

>From an interview with Linus Torvalds.

"http://www.hio.hen.nl/~eniac/Commissies/CommIT/95_96/it4/09_linus.html":http://www.hio.hen.nl/~eniac/Commissies/CommIT/95_96/it4/09_linus.html

"4. What do you consider a normal day (e.g. what takes your time, normally: school, Linux, spare time)

Linux takes up my "working hours" - even just reading email takes a minimum of two hours a day, and to actually react to that email and do development fills up the rest of the day quite nicely indeed, thank you very much.. However, I do take time off for hobbies etc, and I can essentially do linux at my work at the university (they know I do Linux development, and they allow for that fact)."

Unfortunately for everyone, you seem to be wedded strongly to the idea that you can't take any money at all for doing your work. You also seem to think that you can't DO what you want to do full time.

This is sad, because there are companies that can pull it off. For example:
"http://www.linuxdevices.com/articles/AT9754390998.html":http://www.linuxdevices.com/articles/AT9754390998.html . Have you tried to get some investment for a company? You don't need much, after all. Just enough so that you don't have to try and find an IT or consulting gig so you
can buy food.

If you want to change the world, do it. Your constant rhetoric doesn't do anything besides make pigeonhole you.

Alex Pavloff - [email protected]
Eason Technology -- www.eason.com
 
A
> On September 17, 2002 05:23 pm, Alex Pavloff wrote:
> <clip>
> > So this Controlware system that I have now -- a small, DOS based
> > ladder editor and soft logic engine, took about 10 man years to
> > develop with full-time developers.
> <clip>
>
> I've never done one of these myself, but does that 10
> man years include market research, promotion, overhead, project
management,
> etc? You may
> have seen another posting I made in which I speculated that slightly
> over 1 man-year for the programmers and software designers only (no
> overhead or other costs).

Yup. This system, however, was written mostly in x86 assembly , ran on DOS in ~2 MB of RAM for the editor and runtime system and gave good performance on low-end 386 hardware.

It also did have a pile of drivers for various pieces of hardware, and supported networking, as well as provide a way to spawn user-written
processes and access the rst of the system via an API.

> I could be easily persuaded that my estimate was off
> by a factor of 2 or
> even 3, but by a factor of 10? Perhaps this was 10 developers working
> for 1 year, which would probably make inefficient use of manhours. I
> suspect though that a lot of indirect man-hours is in that
> 10 man-years
> estimate.
> When I think of some of the companies which have
> developed their own, I am pretty sure they didn't have that kind of man
power
> available to them.

I'd also have to believe that they were no where as feature complete as this system. This system was a full softplc with ladder editing and the whole she-bang. If you could prune features you didn't want, I'm sure the same system wouldn't take 10 man years to do, but this one system or one like it could probably replace all the softlogic systems that those companies are doing.

Alex Pavloff - [email protected]
Eason Technology -- www.eason.com
 
> On September 17, 2002 05:23 pm, Alex Pavloff wrote:
> <clip>
> > So this Controlware system that I have now -- a small, DOS based
> > ladder editor and soft logic engine, took about 10 man years to
> > develop with full-time developers.

> You may
> have seen another posting I made in which I speculated that slightly
> over 1 man-year for the programmers and software designers only (no
> overhead or other costs).

Somehow, I cannot see both of above to win any tender in the current environment.

Perhaps more detailed specification of the task would help explain suggested man-hours...

Petr

--
<[email protected]>

Petr Baum
P.O.Box 2364, Rowville 3178
fax +61-3-97643342
 
C
Yes, the transmeta idea is mostly software, it caches instruction sequences that occur frequently, so it gets faster on a given workload as the scope of the application is exercised. A large part of the rumors of lackluster performance are because a benchmark that isn't iterative never receives the benefit of the idea. Without the algorithms it's not much of a processor. But, if they stay afloat long enough to get it fully optimized, the potential is awesome. With x86 compatibility becoming a bottleneck, I expect it will be a novel approach like this that will eventually take us forward with backward compatibility. Of course this won't be as big an issue as the world moves to OSS platforms which need only source level compatibility.

Regards

cww
 
C
Hi Alex

List Manager wrote:
> ------------ Forwarded Message ------------
> From: Alex Pavloff
>
> Curt Wuollet wrote:
>
>
>>If you look back a few years they were also doing it when they weren't
>>paid by anybody. The argument that you must be paid to code a major
>>project is absurd. What is your basis? If I want to sit down and write
>>code right now, do i need to find someone to pay me? If I decide to
>>give 4 hours a day to a project, is there anything preventing me from
>>doing so? I'm not sure what you're driving at. If I can do that,
>>logically, everyone else can.
>>Would the work done evaporate?
>
>
> I think we have a fundamental disconnect in what we're thinking the end
> result of a project is going to be. I'm coming from a position where I
> want something that I can put out in the field fairly soon to get things
> done. Free would be nice, but I'm more concerned with getting the job
> done.
>
> You appear to be coming from a position where it really doesn't matter
> when things get done, as long as they're free when they get there.
>
> Your constant rhetoric against other companies shows that you do not
> like the current situation in automation, yet so far, you have not
> produced anything that would cause a drastic change in that situation
> besides multi-megabytes of sometimes valid, something overblown screeds
> against most every large industrial and computer company.
>
> You've been doing this for over two years now. You may not be the best
> programmer in the world, but if you had spent 4 hours a day working on
> your project, I think you'd have accomplished a lot more. (As well as
> getting better at programming.)

We're all doing what we can to change things, for some it's writing code, for some it's challanging minds. You should not discount the importance of the dialog. Even if we produce the best system in the world, we need to provide a good reason for trying it rather than simply going with the flow. Our "product" is as much the very real benefits that Openness and cooperation would provide for automation as it is the software and hardware to instantiate same. To point out those benefits and make them real is my job with the project at the
moment. I don't recall ever lying or misrepresenting the situation as it is, there is no need, as the situation is much less than optimal. And, I'm not above publicly apologizing if I get something wrong. But I don't think you can seriously suggest I'm not doing what I can to
bring about change. My personal goal, is to bring about change, and more specifically, change things for the better. Why that should bother anyone is a subject that could stand illumination.

> Since your project hasn't had any design wins in the real world and
> hasn't registered as a blip on anyone's radar screen, I'm forced to sit
> here and figure out why not. I compare your project against sucessful
> projects (e.g. GCC, Linux Kernel, XFree86, Gnome, and KDE), and note
> that they have developers that can concentrate full time on their
> projects. (GCC people at RedHat/Cygnus, XFree86 people at HP Labs, KDE
> people at SuSE, etc).

Again, the cart before the horse.

> I look at some of the GPL'd libraries that I'm using and have
> contributed to, like CommonC++ (http://www.gnu.org/software/commonc++/)
> and see that the main developer works for a company.
> http://www.ostel.com/ They're doing free software and don't appear to
> have gone under yet. They're small, but their main developer can spend
> most all his time on Bayonne and CommonC++. Then I can contribute stuff
> that I need, and we're all happy.

And I have no problem with that. Only with the concept that you must have a company or commercial undertaking to produce anything of value. I'm not saying that companies cannot focus resources and accellerate progress, but the equality of all contributors, paid or unpaid, rich or poor, is a fundimental concept in the success of OSS and a major differentiator in marketspeak. MS, for example, owes much of it's
success to empowering users, a popular concept. OSS is very much about empowering users and users _as_ developers, so the distinction is important. It's a step further towards software that doesn't suck.

> You also say that "Linux wasn't done with full time people". You're
> wrong.
>
>>From an interview with Linus Torvalds.
>
> http://www.hio.hen.nl/~eniac/Commissies/CommIT/95_96/it4/09_linus.html
>
> "4. What do you consider a normal day (e.g. what takes your time,
> normally: school, Linux, spare time)
>
> Linux takes up my "working hours" - even just reading email takes a
> minimum of two hours a day, and to actually react to that email and do
> development fills up the rest of the day quite nicely indeed, thank you
> very much.. However, I do take time off for hobbies etc, and I can
> essentially do linux at my work at the university (they know I do Linux
> development, and they allow for that fact)."
>
> Unfortunately for everyone, you seem to be wedded strongly to the idea
> that you can't take any money at all for doing your work. You also seem
> to think that you can't DO what you want to do full time.

On the contrary, I like to get paid for what I do. I have a Linux consultancy, one of only two computer consultancies that I know of in the county. I'll bet you can guess what the other one does. I would love to take that to being a vibrant, prosperous concern and hire some like minded people. The reality is that in Mille Lacs county, MN at this time, I can't. And travel is most disruptive and unproductive for me. I'm just a little ahead of the wave. It's inevitable that I will be able to do automation with Linux full time. The timeframe is the only question. In the
meantime, I'm holding out for simply working with Linux and trying to avoid patronizing the monopoly. If things get really bad, I might have
to take a job booting and reloading Windows. If you don't think there's a monopoly, scan the want ads some time. I thought briefly of buying a
MCSE certification, but in line with your comment, it's not what I want to do and knowing what I know, I wouldn't feel right doing it. If I
can't work with Linux, I'll probably do hardware. Or maybe cars or HVAC. It's honest work.

> This is sad, because there are companies that can pull it off. For
> example: http://www.linuxdevices.com/articles/AT9754390998.html. Have
> you tried to get some investment for a company? You don't need much,
> after all. Just enough so that you don't have to try and find an IT or
> consulting gig so you can buy food.
>
> If you want to change the world, do it. Your constant rhetoric doesn't
> do anything besides make pigeonhole you.

I like to think I am. Kinda short on horsepower at the moment, but it's a comfortable pigeonhole in keeping with my principles. I'm just thankful I can still afford my principles.

Regards

cww

>
> Alex Pavloff - [email protected]
> Eason Technology -- www.eason.com
 
Curt Wuollet:
> > If you look back a few years they were also doing it when they
> > weren't paid by anybody. The argument that you must be paid to code
> > a major project is absurd. What is your basis? If I want to sit down
> > and write code right now, do i need to find someone to pay me? If I
> > decide to give 4 hours a day to a project, is there anything
> > preventing me from doing so? I'm not sure what you're driving at.
> > If I can do that, logically, everyone else can.
> > Would the work done evaporate?

Alex Pavloff:
> I think we have a fundamental disconnect in what we're thinking the
> end result of a project is going to be. I'm coming from a position
> where I want something that I can put out in the field fairly soon to
> get things done. Free would be nice, but I'm more concerned with
> getting the job done.

> You appear to be coming from a position where it really doesn't matter
> when things get done, as long as they're free when they get there.

Actually, the practical difference between the outcomes of the two approaches is not that great; the latter approach (no deadlines) may result in earlier delivery and higher quality, but that's minor.

If you want to join in, download the code and start. If you want to sit on the side-lines and wait for it to be finished, then you'll have to
wait for it to be finished, just like any other product. The only difference is that we don't release vapourware press releases with unrealistic release dates (or, for that matter, any release dates).

> Your constant rhetoric against other companies shows that you do not
> like the current situation in automation, yet so far, you have not
> produced anything that would cause a drastic change in that situation
...

Actually, the idea of an OSS automation suite itself is a significant achievement, as is Curt's stewardship of one of the projects.

> Since your project hasn't had any design wins in the real world and
> hasn't registered as a blip on anyone's radar screen, I'm forced to
> sit here and figure out why not.

Who's forcing you?

In any case, the project is progressing. Perhaps not as fast as we'd like sometimes, perhaps not as fast as you'd like (but then why not join and help speed it up?), but it is progressing. At some stage in the future it'll be ready for the real world. Perhaps for some purposes it is already.


Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
M

Michael Griffin

The criteria were left deliberately vague, as I didn't expect anyone to produce an answer to a detailed spec. My question was sparked by curiosity rather than genuine need, and I wanted an opinion rather, not free software
engineering.
The criteria were laid out in a previous e-mails. I won't reproduce that here, but the general objective would be a soft logic system equivalent in features to a small simple conventional PLC. The editor would be ladder only.
The developer had their choice of development software and operating system. All drivers would be purchased. The program editor would be ladder only.

The system described by Mr. Pavloff was considerably more complex and difficult than what I had in mind. The system was written in assembly
language, drivers had to be written, and there were numerous advanced features.

************************
Michael Griffin
London, Ont. Canada
************************
 
A
> > Curt Wuollet wrote:
>
> We're all doing what we can to change things, for some it's writing
> code, for some it's challanging minds. You should not discount the
> importance of the dialog.

Curt,

I suggest you look at the Linux advocacy mini-HOWTO.
"http://www.datasync.com/~rogerspl/Advocacy-HOWTO-6.html":http://www.datasync.com/~rogerspl/Advocacy-HOWTO-6.html

* Focus on what Linux has to offer. There is no need to bash the competition. Linux is a good, solid product that stands on its own.

* Respect the use of other operating systems. While Linux is a wonderful platform, it does not meet everyone's needs.

* Don't insist that Linux is the only answer for a particular application. Just as the Linux community cherishes the freedom that Linux provides them, Linux only solutions would deprive others of their freedom.

* There will be cases where Linux is not the answer. Be the first to recognize this and offer another solution.

Quite frankly, your advocacy sucks. It might work on Slashdot, but you aren't going to get karma points here for calling Bill Gates more evil that Osama Bin Laden.

You're preaching to the choir (yes, I am developing a Linux based product for industrial automation), and the choir is looking at the door wondering when they can leave because they're tired of being covered by spittle.

You haven't used Microsoft software in years, and you haven't done a automation project in years using the tools that most everyone on this list
uses. You are unqualified to tell people that their current system "is awful". People every day get things done with these systems, and they work. You come along, say "your stuff DOESN'T WORK, and the companies that sold you the system are evil".

If you've got a better system, bring it on. You don't. Until then, you're just spewing FUD.

> And I have no problem with that. Only with the concept that you must
> have a company or commercial undertaking to produce anything of value.

Curt, I said you need to have FULL TIME people working on a project to make it a success. You can do that either with commercial companies (like Red Hat), working for universities (like Linus when he was in Finland), or having independently wealthy folks (like Linus-the-millionaire now).

Alex Pavloff - [email protected]
Eason Technology -- www.eason.com
 
C

Curt Wuollet

Well, to go to the other end of the scale, I have in a production setting, a simple "PLC" I wrote in about 18 hours. It does boolean logic and you could certainly add integer and floating point math by simply calling the appropriate math functions from the rich set supported with gcc. I wrote A/D and D/A functions as well, but didn't
need them. I've got it throttled back to about a 10msec scan time as that was fast enough to keep up with the robots and other hardware. The mechanics of emulating a PLC aren't that difficult. Generalizing it and providing tools for
non-programmers is by far the largest part of the picture. There is a graphical ladder editor that I could modify to output C logic statements and have about 90% of a simple system. No telling how long that would take, but it wouldn't be a major project.

It's a case of the 90/10 rule. The first 90% of the project takes only 10% of the time with the remaining 10% consuming 90% of the time. Indeed, such a project could go on forever as you keep adding bells and whistles and fancy graphics, etc.

The current state of development does everything I need to do for most applications where I would otherwise need a PLC and a PC. Once you have a PC in the mix, taking over the PLC function is straightforward. Making the thing usable by anyone is a much bigger deal. But, since major parts are available ready made, Open Source, dramatic time savings can be had. In a rational
world where you could use TCP/IP to connect everything, you would be home free. The Tower of Babel adds another lifetime. The MAT/lplc is a general class solution and much more of an
undertaking. But it will be much more than a soft PLC.

If the question was how long does it take to produce a basic programmable logic controller, there's your answer. It arguably does a big chunk of what's needed and took less time than learning
someone else's tools. It's been running for about two years as long as there is power. 48 digital points. If you don't believe me, load up a Linux box and I'll mail it to you, or there's a copy on the MAT site, I believe. It's OSS so feel free.

Regards

cww
 
C

Curt Wuollet

Hi Alex

List Manager wrote:
> ------------ Forwarded Message ------------
> From: Alex Pavloff
>
>>>Curt Wuollet wrote:
>>
>>We're all doing what we can to change things, for some it's writing
>>code, for some it's challanging minds. You should not discount the
>>importance of the dialog.
>
> Curt,
>
> I suggest you look at the Linux advocacy mini-HOWTO.
> http://www.datasync.com/~rogerspl/Advocacy-HOWTO-6.html

Read it. Practice it where appropriate.

> * Focus on what Linux has to offer. There is no need to bash the
> competition. Linux is a good, solid product that stands on its own. *
> Respect the use of other operating systems. While Linux is a wonderful
> platform, it does not meet everyone's needs.

I Agree. When people have the option, which is absent in this market segment, I have no doubt they can select the most appropriate choice.

> * Don't insist that Linux is the only answer for a particular
> application. Just as the Linux community cherishes the freedom that
> Linux provides them, Linux only solutions would deprive others of their
> freedom.

I also agree with this. However this situation is such that you have the opposite going on. With Microsoft being the only option proposed for the
solution of any problem. And for the wrong reasons. And merely suggesting that a particular task could be accomplished more easily with OSS and Linux draws pretty much the same criticism.

> * There will be cases where Linux is not the answer. Be the first to
> recognize this and offer another solution.

Again this is based on there being a choice. And if you have noticed, I do suggest alternatives for those few cases where Linux is
inappropriate. In fact, I offer hardware solutions where software is inappropriate. To the best of my knowledge, nothing I have proposed
is questionable from a technical or engineering viewpoint and I seldom offer anything that I don't know from first hand experience is an easier
or better solution.

> Quite frankly, your advocacy sucks. It might work on Slashdot, but you
> aren't going to get karma points here for calling Bill Gates more evil
> that Osama Bin Laden.

I didn't quite go that far. I merely said I would prefer one over the other hacking my machine. And I don't frequent Slashdot. Too much empty
unreasoned argument there. I greatly prefer specific, factual, substantive discussions. They are certainly fantastic enough with actions beyond belief portrayed as being in the interest of the public. I certainly don't go in for misstatements of that magnitude. In fact, that's part of the problem. MS requires that they should be able to counter your security and invade your privacy at their convenience and that is just accepted with a yawn. To suggest that is a bad idea and totally unacceptable is somehow irrational and hyperbolic and spreading FUD. Hmmm.... I'm perfectly willing to yield to objectivity there.

> You're preaching to the choir (yes, I am developing a Linux based
> product for industrial automation), and the choir is looking at the door
> wondering when they can leave because they're tired of being covered by
> spittle.

I do apologize for that, It's nothing personal.

> You haven't used Microsoft software in years, and you haven't done a
> automation project in years using the tools that most everyone on this
> list uses. You are unqualified to tell people that their current system
> "is awful". People every day get things done with these systems, and
> they work. You come along, say "your stuff DOESN'T WORK, and the
> companies that sold you the system are evil".

I have never said that the status quo doesn't work. I am of the opinion, and it's somewhat qualified opinion, that the market is beset with gross inefficiencies and deliberate artificial.obstacles. I didn't invent this. I doubt that anyone seriously believes that we couldn't have common protocols, reusable knowledge, connectivity, interoperability, and the other goals I have mentioned. And it's painfully obvious that it would make life much easier and greatly improve productivity. The situation is perpetuated by the exclusive use of closed products. This mostly benefits the commercial entities at the expense of the
practitioners and ultimately the customers. We can ill afford this. It won't be changed as long as folks are convinced it has to be that way. I seek for the way to get people to act in their own interest. I'm not at all sure why this makes people angry.

> If you've got a better system, bring it on. You don't. Until then,
> you're just spewing FUD.

We're working on it. And yes, there is Fear. But, no Uncertainty or Doubt that nothing will change without an effort. You have offered sage advice on Linux advocacy. You have not offered up the way to change things in automation. I'm all ears. For now, I am of the opinion that it will only change when people can see how it could be and get tired of working a lot harder than neccessary to support the status quo.

>
>>And I have no problem with that. Only with the concept that you must
>>have a company or commercial undertaking to produce anything of value.
>
> Curt, I said you need to have FULL TIME people working on a project to
> make it a success. You can do that either with commercial companies
> (like Red Hat), working for universities (like Linus when he was in
> Finland), or having independently wealthy folks (like
> Linus-the-millionaire now).

Close enough, the other points are really more important.

Regards

cww

> Alex Pavloff - [email protected]
> Eason Technology -- www.eason.com
 
Top