On July 23, 2003, Ralph Mackiewicz wrote:
> > Your assertions about the failure of commercial enterprises
> > directly contradict the experience of most people. It will be hard
> > to gain traction that way. <

On July 25, 2003, Curt Wuollet wrote:
> Once again, perception. The way things are done is always adequate
> until something better comes along. And I've not opined on the
> failure of commercial enterprises, although there does seem to be
> some shake out going on. It's wonderful for them, high margins,
> zero liability, an upgrade money machine that people can't opt out
> of, and very low expectations from their customer base. Far from
> failing, the vendors have all the advantages. <

I wish the world of commercial software was really like that. My life would be a whole lot easier. Pass some of that over this way! I'm not sure what commercial software company has a business model like this, but I seriously doubt that you would find many (if any) commercial software vendors, or their customers, that would agree that this accurately reflects their relationship.

> If you couldn't use a better way of doing things about now, by all
> means, dance with them that brung ya. If you suspect you could do
> better, try doing things differently. 99% of the companies who had
> a good enough way of doing things are now forgotten. <

Different <> better. Your references to the failure of commercial enterprises brings up something about non-commercial software that I am curious about: success/failure.

One of the reasons, IMHO, why commercial markets have worked so well, is that commercial markets have objective criteria for success/failure. Either there are enough people who see value in what you do, such that you are able to generate a profit on your sales, or there aren't. Period. If your customers decide that you can't deliver value to them, they leave. You end up being sold and dismembered, or bankrupt. There are numerous commercial software companies that failed to live up to customer expectation and failed to deliver value and no longer exist as a result. This is a good thing (getting rid of the non-performers). And, it can happen to any company that doesn't deliver value. Size doesn't matter. All the big bad IA companies are subject to this without mercy. Some of them look to be getting their heads lopped off soon. No external organizational entity has to decide if a company delivers sufficient value, or whether it has a good/bad business model. The decision gets made automatically by the invisible hand of the market by the only people that matter in the
decision: customers and investors.

How is success/failure measured for OSS? Is it an advantage or disadvantage that a theoretically "bad" piece of OSS (one in which very few people see any value) can stay around because it can avoid the invisible (and ruthless) hand of the market while the same from a commercial company would find its way to the chopping block sooner, rather than later?

Regards,
Ralph Mackiewicz
SISCO, Inc.

 

Curt Wuollet (on commercial enterprises):
> > It's wonderful for them, high margins, zero liability, an upgrade
> > money machine that people can't opt out of, and very low
> > expectations from their customer base.
...

Ralph Mackiewicz:
> I'm not sure what commercial software company has a business model
> like this, but I seriously doubt that you would find many (if any)
> commercial software vendors, or their customers, that would agree that
> this accurately reflects their relationship. <

Well, I can guess *one* that CW had in mind, we probably all can

> Your references to the failure of commercial enterprises brings up
> something about non-commercial software that I am curious about:
> success/failure.
...
> How is success/failure measured for OSS? <

Basically, it's measured in terms of how popular it is and how much
development effort it gets. The two are related - a piece of software
that has more users will be more likely to get improvements etc.

> Is it an advantage or disadvantage that a theoretically "bad" piece of
> OSS (one in which very few people see any value) can stay around
> because it can avoid the invisible (and ruthless) hand of the market
> while the same from a commercial company would find its way to the
> chopping block sooner, rather than later? <

On the balance, probably an advantage: there's not much cost to having
it around - a few MB on an FTP server somewhere and a couple of webpages
that nobody visits don't cost much. On the other hand if there are in
fact a few people that do value it, they can keep it going for and among
themselves - various special interest groups, like the blind or speakers
of particular (natural) languages, have been given the short shrift by
commercial companies, on the grounds that there aren't enough of them to
matter commercially.

Indeed, many on this list will be able to name occasions when they
personally have been inconvenienced by a commercial company's
termination of a software product (perhaps after a takeover).

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

 

Hi Ralph

Some interesting points.

On July 28, 2003, Ralph Mackiewicz wrote:
> On July 23, 2003, Ralph Mackiewicz wrote:
>>>Your assertions about the failure of commercial enterprises
>>>directly contradict the experience of most people. It will be hard
>>>to gain traction that way. <
>
> On July 25, 2003, Curt Wuollet wrote:
>>Once again, perception. The way things are done is always adequate
>>until something better comes along. And I've not opined on the
>>failure of commercial enterprises, although there does seem to be
>>some shake out going on. It's wonderful for them, high margins,
>>zero liability, an upgrade money machine that people can't opt out
>>of, and very low expectations from their customer base. Far from
>>failing, the vendors have all the advantages. <
>
>
> I wish the world of commercial software was really like that. My life
> would be a whole lot easier. Pass some of that over this way! I'm not
> sure what commercial software company has a business model like this,
> but I seriously doubt that you would find many (if any) commercial
> software vendors, or their customers, that would agree that this
> accurately reflects their relationship. <

Perhaps, But that is certainly colored by the various perspectives. For example, with the pricing and volumes of some of the commercial software, if they don't have great margins, their cost of doing business is out of control. And the liability aspect comes directly from the license language. And the upgrade issue isn't illusionary as these pages attest. And the cost of swapping a GE solution for an AB solution _is_ mostly prohibitive and forced upgrades or abandomnement are something I have personally experienced as have many others.

These are the factual issues, I'm quite sure the vendors prefer not to see things that way. I'm less sure about the customers, these closely parallel what we see in these pages every day. Why people put up with it gets to the crux of the matter. It's the lock in factor. Once you get in, you have a lot invested and on top of that it's very expensive to do anything but go along. Many would say that's simply smart business on the part of the vendors, but that's exactly what I am saying, The balance is managed to the benefit of the companies and the detriment of the customers. It fundimentally alters the free market system when you have deliberately high barriers to switching vendors.

>>If you couldn't use a better way of doing things about now, by all
>>means, dance with them that brung ya. If you suspect you could do
>>better, try doing things differently. 99% of the companies who had
>>a good enough way of doing things are now forgotten. <
>
>
> Different <> better. Your references to the failure of commercial
> enterprises brings up something about non-commercial software that I
> am curious about: success/failure. <

Conversely, 50 functionally identical parallel solutions, with similar business models, offering pretty much the same deal isn't a vibrant market.

The same formula multiplied <> better.

Diverse = better.
Interoperable = better.
Choice = better.
Customer biased = better.

> One of the reasons, IMHO, why commercial markets have worked so well,
> is that commercial markets have objective criteria for
> success/failure. Either there are enough people who see value in what
> you do, such that you are able to generate a profit on your sales, or
> there aren't. Period. <

It's not quite that simple, many ubiquitous things fail those criteria. Considering only profitability is an extremely subjective measure applicable only to the commercial model.

The Internet is worth billions out of all proportion to how much revenue it has generated directly. And very few have found a magic formula to capitalize on it's vast potential. And much of the infrastructure operates at below break even. Yet, I dare say that it's successful and vital to the economy. It's value is user to user.

Similarly the measure for automation should be if more things get automated and benefit users rather than if a few multinational corporations make a killing doing it. And that the people who provide the solutions get to keep more of the pie is hardly a sign of failure. Here too, the value is user to user. Whether a platform is free or hidiously expensive, if it gets the job done the real value is delivered. Every one benefits if the cost is lowered, except the huge corporation.

> If your customers decide that you can't deliver
> value to them, they leave. You end up being sold and dismembered, or
> bankrupt. There are numerous commercial software companies that
> failed to live up to customer expectation and failed to deliver value
> and no longer exist as a result. This is a good thing (getting rid of
> the non-performers). And, it can happen to any company that doesn't
> deliver value. Size doesn't matter. All the big bad IA companies are
> subject to this without mercy. Some of them look to be getting their
> heads lopped off soon. No external organizational entity has to
> decide if a company delivers sufficient value, or whether it has a
> good/bad business model. The decision gets made automatically by the
> invisible hand of the market by the only people that matter in the
> decision: customers and investors. <

Lock in and the long lifetime of the systems puts a real damper on the normal market pressures. Monopolization and the artificial restriction of choice can actually inhibit the market from working as well as raise the barriers to entry such that there is no real competition. I'm a firm believer in the free market, but the ability of monopolies and partners of monopolies to restrict trade and limit choices means that the vernacular of free trade doesn't apply in a market with colluding dominant players. If the railroads had been able to successfully prevent the building of highways, they would still own transportation. The conditions for competition must be there before there can be competition. Right now this is hardly a free market. Sure, you have your choice of railroads, but trucks can't run on the privately owned rails. No matter how much more efficient they are.

> How is success/failure measured for OSS? Is it an advantage or
> disadvantage that a theoretically "bad" piece of OSS (one in which
> very few people see any value) can stay around because it can avoid
> the invisible (and ruthless) hand of the market while the same from a
> commercial company would find its way to the chopping block sooner,
> rather than later? <

Exactly the same forces work on OSS, it's just the metric is different. OSS survives if people use it and contribute. Revenue is irrelevent. A lot of people want to continue to use dead commercial software that has been declared unprofitable, it's usefulness is irrelevent. In many cases good but unprofitable commercial software is replaced by bad but profitable software. And even excellent OSS will never make a profit in financial terms.

Which system is the better measure for the user?

Regards

cww

 

On July 28, 2003, Ralph Madkiewicz wrote:
> > I wish the world of commercial software was really like that. My
> > life would be a whole lot easier. Pass some of that over this way!
> > I'm not sure what commercial software company has a business model
> > like this, but I seriously doubt that you would find many (if any)
> > commercial software vendors, or their customers, that would agree
> > that this accurately reflects their relationship. <

On July 29, 2003, Curt Wuollet wrote:
> Perhaps, But that is certainly colored by the various perspectives.
> For example, with the pricing and volumes of some of the commercial
> software, if they don't have great margins, their cost of doing
> business is out of control. And the liability aspect comes directly
> from the license language. And the upgrade issue isn't illusionary as
> these pages attest. And the cost of swapping a GE solution for an AB
> solution _is_ mostly prohibitive and forced upgrades or abandomnement
> are something I have personally experienced as have many others. <

Yes, but there are large numbers of AB and GE customers that consider the prices they pay and the products and services they get in return to be reasonable. In fact, AB, GE, and most of the other majors, have a large number of satisified customers. That is why they are generally considered successful and are able to generate profits. The characterization of commercial relationships that I was responding to is not an accurate characterization of the large majority of customers of commercial software. Therefore, your portrayal of that relationship won't get many of these people to look at an OSS alternative based on that argument.

> These are the factual issues, I'm quite sure the vendors prefer not to
> see things that way. I'm less sure about the customers, these closely
> parallel what we see in these pages every day. Why people put up with
> it gets to the crux of the matter. It's the lock in factor. Once you
> get in, you have a lot invested and on top of that it's very expensive
> to do anything but go along. Many would say that's simply smart
> business on the part of the vendors, but that's exactly what I am
> saying, The balance is managed to the benefit of the companies and the
> detriment of the customers. It fundimentally alters the free market
> system when you have deliberately high barriers to switching vendors. <

I would disagree that a free market system precludes there there not be barriers to changing suppliers. Any customer of an OSS would also face barriers to switch. If you want to state this in a way that illustrates the benefits of OSS you would claim that the cost barrier to abandon an open source solution if it doesn't meet your needs in the future is lower than a commercial solution that fails to meet your needs in the future. Again, I'm not saying that there are no benefits to OSS. I am saying that if the reasons for using OSS are always stated on some moralistic terms that requires somebody to, for instance, accept that commercial software is a failure, then anyone who doesn't accept that premise (as I think is true of most of the potential users) won't be very interested in an OSS.

> > Different <> better. Your references to the failure of commercial
> > enterprises brings up something about non-commercial software that I
> > am curious about: success/failure. <
>
> Conversely, 50 functionally identical parallel solutions, with similar
> business models, offering pretty much the same deal isn't a vibrant
> market.
>
> The same formula multiplied <> better.
>
> Diverse = better.
> Interoperable = better.
> Choice = better. <

Everyone of these would apply to commercial software as well.

> Customer biased = better. <

Actually, I think that OSS is developer biased. That is why most people using OSS today are developers. There just aren't enough developers for OSS in IA to gain traction against well funded profitable commercial implementations.

> > One of the reasons, IMHO, why commercial markets have worked so
> > well, is that commercial markets have objective criteria for
> > success/failure. Either there are enough people who see value in
> > what you do, such that you are able to generate a profit on your
> > sales, or there aren't. Period. <
>
> It's not quite that simple, many ubiquitous things fail those
> criteria. Considering only profitability is an extremely subjective
> measure applicable only to the commercial model. <

Although I don't agree that profitability is a subjective measure, this is the whole point of my question: what is the criteria for success if it's not measured by a result of the invisible hand of the market? I suspect that all other measures of success are purely subjective because market acceptance is not necessarily needed for an OSS to survive. Is all it really take to be successful with OSS one person using it with an FTP server?

> Similarly the measure for automation should be if more things get
> automated and benefit users rather than if a few multinational
> corporations make a killing doing it. And that the people who provide
> the solutions get to keep more of the pie is hardly a sign of failure.
> Here too, the value is user to user. Whether a platform is free or
> hidiously expensive, if it gets the job done the real value is
> delivered. Every one benefits if the cost is lowered, except the huge
> corporation. <

1. There are very few IA companies making a "killing" in the market.
2. Keeping more of an economic pie that equals zero is still zero.
3. Commercial software is, generally, getting the job done and delivering real value which is benefiting both users and some huge corporations.

Here we go again. The implication is that as long as a huge corporation is benefiting, then users are not. I don't buy that argument. And, it seems to be the only consistent message given with respect to an OSS IA product in this forum. I think most people that are buying the products of the large IA vendors consider the money well spent. I'm sure that they would like to be spending less. But, if they didn't think it was worthwhile, they wouldn't be buying the products and a Linux based OSS PLC would have a dominant market share.

> Lock in and the long lifetime of the systems puts a real damper on the
> normal market pressures. Monopolization and the artificial restriction
> of choice can actually inhibit the market from working as well as
> raise the barriers to entry such that there is no real competition.
> I'm a firm believer in the free market, but the ability of monopolies
> and partners of monopolies to restrict trade and limit choices means
> that the vernacular of free trade doesn't apply in a market with
> colluding dominant players. If the railroads had been able to
> successfully prevent the building of highways, they would still own
> transportation. The conditions for competition must be there before
> there can be competition. Right now this is hardly a free market.
> Sure, you have your choice of railroads, but trucks can't run on the
> privately owned rails. No matter how much more efficient they are. <

I disagree. It is a free market. Some people happen to disagree with the results of the free market and want to believe that there are reasons more complicated than the fact that the market delivers only and exactly what the most people want. AB, Siemens, GE, etc. came to their dominant positions because people bought their products.

> Exactly the same forces work on OSS, it's just the metric is
> different. OSS survives if people use it and contribute. Revenue is
> irrelevent. A lot of people want to continue to use dead commercial
> software that has been declared unprofitable, it's usefulness is
> irrelevent. In many cases good but unprofitable commercial software is
> replaced by bad but profitable software. And even excellent OSS will
> never make a profit in financial terms. <

If that is true, then perhaps OSS will never be successful. If people cannot make a profit delivering an OSS in IA then no one will ever deliver an OSS for IA. Someone somewhere, other than the user, has to make a profit or it doesn't make any sense to produce or deliver a solution. Perhaps, what you mean to say is that you think it is better that the profits be generated by the system builder rather than the software developer. If true, is it any wonder then why the IA OSS solutions can't seem to get a lot of developers involved?

Regards,

Ralph M.

 

On July 28, 2003, Ralph Mackiewicz wrote:
> > How is success/failure measured for OSS? <

On July 29, 2003, Jiri Baum wrote:
> Basically, it's measured in terms of how popular it is and how much
> development effort it gets. The two are related - a piece of software
> that has more users will be more likely to get improvements etc.

...snip...snip...

> there's not much cost to having it around - a few MB on an FTP
> server somewhere and a couple of webpages that nobody visits don't
> cost much. <

But how do we objectively judge that it is successful and therefore, worthy of spending your effort on improving? If it only takes one person with an FTP server to keep an OSS alive, is that considered successful? BTW, I use freeware from Mike Lin that fits this criteria and I like using it so it is successful for me. The meager donations that Mike Lin attracts couldn't pay the rent but its just a sideline to his main career. But can we build a real business in IA around this kind of solution?

> On the other hand if there are in fact a few people that do value
> it, they can keep it going for and among themselves - various
> special interest groups, like the blind or speakers of particular
> (natural) languages, have been given the short shrift by commercial
> companies, on the grounds that there aren't enough of them to
> matter commercially. <

I understand how niche markets can benefit with OSS due to the very low cost of entry. If the market is not being served, then any solution is better. But what about a mature market with numerous solutions, like IA, where OSS doesn't seem to be attracting the level of development needed to make it competitive with commercial offerings? Here is my concern: maybe it doesn't really matter if OSS is competitive. Maybe only those who find the whole idea of commercial software to be objectionable on a moral basis will be able to justify spending their time developing an OSS product for IA. I don't think that this is necessarily true. I am seeking some answers. My interest is in working with solutions that enable some profit to be made. If there are no profitable business models involving OSS in our markets, I have charities I consider more worthy that I will spend my charitable efforts on instead. The well being and interests of my company's customers and employees are at stake. I need profits to serve these interests (and mine as well).

Ralph M.

 

Ralph wrote:

> Actually, I think that OSS is developer biased. That is why most
> people using OSS today are developers.

a) Not true for most of OSS by very long margin
b) Possibly true for automation software at present - but the same apply to
any new piece of software regardles of licence. So what?

Strikingly false statement in otherwise sensible response.

OSS is unlikely to win over commercial software in process control on basis
of software price advantage, it will have to provide in long term better
choice of features. The same applies in other areas - that is why Munich
decided for more costly Linux option and not for cheaper Windows one
recently.

http://www.usatoday.com/money/industries/technology/2003-07-13-microsoft-linux-munich_x.htm

Petr

--
<[email protected]>

Petr Baum, P.O.Box 2364, Rowville 3178
fax +61-3-97643342

 

On July 28, 2003, Ralph Mackiewicz wrote:
> > > How is success/failure measured for OSS? <

On July 29, 2003, Jiri Baum wrote:
> > Basically, it's measured in terms of how popular it is and how much
> > development effort it gets. The two are related - a piece of
> > software that has more users will be more likely to get improvements
> > etc.

> ...snip...snip...
> > there's not much cost to having it around - a few MB on an FTP
> > server somewhere and a couple of webpages that nobody visits don't
> > cost much. <

On July 31, 2003, Ralph Mackiewicz wrote:
> But how do we objectively judge that it is successful and therefore,
> worthy of spending your effort on improving? <

That's the important question, I guess; fortunately, it doesn't have to be all that objective:

- from a commercial point of view, the question is whether the system is close enough to what you need that using it and developing whatever's missing (if anything) is faster/better/cheaper than with all the alternatives;

- from a hobby point of view, the question is whether you'll have fun.

Neither of these really requires any particularly objective global knowledge...

> If it only takes one person with an FTP server to keep an OSS alive,
...
> But can we build a real business in IA around this kind of solution? <

The real business will probably be in custom solutions, which people will pay for just as they do now. Basic foundation software accounts for only a small percentage of the work, anyway.

> > On the other hand if there are in fact a few people that do value
> > it, they can keep it going for and among themselves
...
> I understand how niche markets can benefit with OSS due to the very
> low cost of entry. If the market is not being served, then any
> solution is better. But what about a mature market with numerous
> solutions, like IA, where OSS doesn't seem to be attracting the level
> of development needed to make it competitive with commercial
> offerings? <

There's always a start-up hurdle (with any software, OSS or closed); and as far as I can tell it's not actually that far off.

> My interest is in working with solutions that enable some profit to be
> made. <

Like I write above, the trick is to find a system which is close enough to what you need that using it and developing whatever's missing (if anything) is faster/better/cheaper than with all the alternatives. You might use this internally in your company, or externally as an integrator or turn-key system supplier.

Thus, the only real recommendation is to stay informed

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