Re: Open Source vs. Commercial

Captain Jake wrote:

I suspect it is less likely that an open source effort will produce quality software than a commercial effort. 

I suspect you can't base your suspicions on any hard facts, because the
necessary facts are extremely hard to collect.

In order to prove or disprove any of the theories about how many Open
Source projects produce quality software vs. Commercial, we need to
count all "quality" projects of each kind and divide by the number of
total projects of each kind, respectively. This would give us the
precentage of quality software out of the total for each category.

The probem is that none of this data is available, and when data is
available it's not trivial to evaluate.

We don't know how many Commercial projects started and ended as crap
that never saw the day of light, because it happens behind closed doors.

Information about Open Source projects may seem like it's more
available, but in fact it's even harder to count how many projects there
 were or still are. Should we count every kid who opened an account on
SourceForge and forgot about it a week later? There is no barrier to
declaring an Open Source project, so the number is huge. Of courlse a
significant precentage won't become a real project. So again it's hard
to count.

It's also nearly impossible to measure quality, because this is very
much a subjective matter. So there goes the other half of the data we need.

This leaves us with nothing more than "I suspect" which reflects more
emotion than knowledge.


really bothers me is when some luminary suggests that if a commercial
product were open sourced that the quality and timeliness of the releases would go up. There is absolutely no evidence that suggests this is even remotely likely. Find me examples where this has happened. 

I will not defend what other people claim, especially when I don't know
the context. I can only offer my view in this regard.

When a company loses interest in a product, to the point of hardly any
progress, that software will be better off as Open Source. If nothing
else then each user who wishes to fix the bugs that bothers them will be
able to do so.

In some cases this turns into a "live" Open Source project, with timely
releases and constantly improved quality. Some other times it doesn't
happen.


I can find plenty of examples of products that died as open source projects. Turbopower, Interbase, etc. SourceForge is full of dormant
projects that represent the graves of unsuccessful commercial ventures turned open source. 

This is a great example of why it's hard to measure.

You consider TurboPower as a failed conversion from Commercial to Open
Source. I see it as success, *compared to the alternative*.

TurboPower (the company) was shut down by its owners. By the way, not
because of fierce competition from Open Source, but becasue the parent
company is more interested in gambling software or something like that.
So the choice was to bury TurboPower products, or release them as Open
Source.

If they had been buried - there would be exactly ZERO releases and
improvement. Since they were released, there was some activity and even
a few minor bug fix releases. As an Open Source project it may not be a
poster child, but at least it's better than having nothing at all.

Personally I do benefit from open-sourcing TurboPower products, because
I have a large code base which I can (and do) use, from which I can
learn and borrow ideas.

You other example - Interbase - I don't really understand. The OS
incarnation of Interbase, aka Firebird, is doing great.


To the extent that programmers are working toward making open source software timely and excellent, they are undermining professional programmers' career possibilities, especially their ability to escape
the 9-5 rat-race and make a living selling their own software. 

Hmm... what about all those professional programmers who get paid rather
nicely in order to produce Open Source software?


It is not quite that simple. You are not differentiating between the
labor market and the final market. A software developer wanting to enter the market for software has a significant barrier if there is good software available for free in that market. The only way he can
benefit from that market is to enter the labor market for that market. 

There's a significant barrier also when there is good Commercial
software available. It doens't really matter who paid to develop the
exiting software - if it's good it will make it harder to enter the market.


If this seems obtuse, consider the fact that FOSS made by paid programmers is no different than "dumping", where a company produces
a product that it sells below cost for the express purpose of
driving out existing and would-be competitors. 

The vast majority of Open Source projects, maybe even all of them, were
not started as a way to dump prices. Linus Torvalds started Linux
because he couldn't afford to purchase a copy of Unix. IBM invests in
Linux because it needs a good operating system with a large base of
applications to run on its hardware. The fact that Linux takes business
from Microsoft is a side effect, not a goal.

Suppose is a scarce resource that's carved out of the earth in only a
handful of small mines. The company MicroPlastics, who owns most of the
mines makes billions selling plastic at any price they want. Then some
luminary discovers that plastic can be had at almost zero cost in the
distillation process of crude oil. Is this dumping? Should this
discovery be shushed up so MicroPlastics can continue its reign over the
market?

So people discovered that producing software isn't all that difficult as
it once was. Today's kids can develop software, sometimes good software,
although we older programmers struggled for many years to acquire the
knowledge and decent tools to do our job. So what?


If they don't get paid, they only make software that interests them
(and some argue that is of low quality). So who makes the "uninteresting" software for businesses? People who write software
for money, so again there's a market and the argument is void. 


The argument is far from void. You in fact just argued that FOSS will
 lead to software development careers that involve only the creation
 of uninteresting software.


Uninteresting to *hobbyists*. I develop accounting/financial software
and I enjoy it very much, although I doubt such kind of software is the
first priority for a hobbyists who wants to write code.


"big iron". The development cost a lot of money, which made the deal very expensive for the customers. Naturally with a proprietary
and expensive environment there wasn't much of a market around IBM's environment (or any of the other proprietary systems). In the
end the whole setup became too expensive and uncompetitive. 

IBM uncompetitive when it developed its own operating systems? Do you
 have any idea how silly that sounds to anyone that was around during
 that time?


Read again:

*IN THE END* the whole setup became too expensive and uncompetetive.

Yes, in the end there were cheaper and better alternative such as PCs,
Unices, etc.


Today IBM shares the cost of developing an operating systems (Linux) with many other companies. Linux isn't really free - IBM customers pay for it when they buy IBM hardware. However, as IBM shares the cost of development with many others, IBM's cost per-customer is next to nothing.

Their per-customer operating system cost when they wrote their own was relatively small as well. But obviously, their costs are going to
be lower if they can use something that was for the most part created by an army of volunteers. 

Linux may have been created by an army of volunteers, but it wasn't
created to run on mainframes and certainly not IBM mainframes. So IBM
contributed to Linux a lot of code to make it compatible with their
hardware, and as a side effect make Linux work better on big iron in
general.


That is not the issue here. The issue is whether or not FOSS has
resulted in a larger wage fund for software developers, not just
whether it has helped a large multinational firm increase it's
profits.

And helped the many users of IBM computers get access to a huge base of applications. Let's not forget the bottom line.


If it has cut IBM's labor costs then it has reduced the 
amount of money going into the pockets of software developers, pure
and simple. Here is a direct example of how FOSS is undercutting
software developers even in cases where firms are paying software
developers to create it.


Not at all!

Suppose all those volunteers (who produced a great free product), were like - professional paid programmers. Would that make any difference? No, because they increased the supply of programming labor, and drive the wages down anyway.


The most widely used (but unfortunately so) example is that software can be given free and money comes from support.

Great, this turns software developers into pure cost centers, and makes the support personnel the economic stars of the firm. Again, abysmal for software development careers.

Again wrong, because with bad software the cost of support sky-rockets, so programmers are still an important piece in the puzzle.


"reduce cost" is synonymous with "less money paid to workers". Again,
 these schemes are thereby hurting software developers in general.


The cost is shared with others, the cost doesn't disappear.


The money simply won't be spent on that Commercial
software, but on something else. Whatever that something else happens to be - food, clothing, entertainment, hardware - doesn't

That is a loss to society, equal to the difference in productivity between software development and the next best alternative for those
individuals best suited to software development.

You assume that most programmers do programming because this is the thing they do best. Sorry, but there's no evidence to support this claim.

Maybe most programmers would be better off, for their own sake or for society's sake, growing vegetables or fixing cars? There's no way to prove one way or the other.

And once again you neglect to account for the fact that without FOSS un-paid developers will become paid developers, and the bottom line for programmers wages will be exactly the same.


In addition to this direct loss due to productivity differences, there is also the loss that would be caused by large-scale sectoral shifts as software developers lose their jobs and take time finding
jobs in other sectors.

So the current state of things should be locked forever? But how did we get into the current state - wasn't it through radical changes that were caused by advances in computers and communications?

In a free market, when the market chooses a path it's because that path is the best compromise between all the pressures that drive the market. If FOSS is good for many people in many ways, then the cost of the change doesn't matter or matters very little.


Everyone, including those who object to FOSS, are happy when prices
 of food or gas drop.

Not everyone. Farmers are not very enthralled with drops in the price
of food, nor are the myriad industries that sell their goods to farmers. Likewise for workers in the oil industry with regard to drops in the price of gas.

Sure farmers and oil workers aren't happy, but my point is that *you* are happy to pay less. Therefore you have no moral standing to criticize others who are happy to pay less for software.

The fact of the matter is that society as
a whole is not better off with lower prices (or higher prices). Society as a whole is benefited the most by prices that accurately reflect the total social cost.

Great. And what is the total social cost? When oil is extracted from the earth, turned into gas and other products, and ends up polluting the environment and contributing to climate changes that cause floods in India - what is the social cost?

The social cost can't be calculated, and even its definition can't be agreed upon - it doesn't exist.


If this still seems odd to you, consider how society would fare if cars were totally free. Do you really think society would be better off? Who would produce cars? The only ones that would produce cars would be those that need them for their own purposes AND have the resources to create them. You and I certainly would not be able to find cars for our own use. Who would go through all the bother of creating a car just to give it to us? Nobody. Ditto for most people on this planet. 

Great wrong example.

Oil companies may have an interest to produce cars and give them away for free, because without cars they won't sell oil. Of course, the price of the cars would have to be reflected somehow in the prices of gas, but the cars would be just as free as FOSS.


And won't you be really happy to get free gas or free groceries?

No, because I know that the prices of other things will be higher as

When you can get something you want, for free or for a price lower than normal, do you actually refuse the deal? Or you just say so for the sake of this discussion? I suspect the latter.


Suppose air isn't free, so each and every person has to pay for the
air they breath. This will create a huge market for air, won't it?
How wonderful! we created a huge market with lots of money flowing
around, and everyone must be happy because of all this new business. Really? Where all the money will come from? That money will be taken from other markets - food, clothing, entertainment, hardware and... software!

You seem to be fixated on the constancy of the money supply as if it

Not money supply but money *flow*. I think I've made it very clear in the original message that I'm talking about the flow of money, not the quantity of money.


What IS of consequence for human welfare is how efficiently resources
are allocated, in other words how close the allocation of resources
comes to creating exactly those goods and services most sought after
by human beings. This allocation will come the closest when people can express their relative preferences in some manner. The pricing system works very well for this purpose, when prices are allowed to be driven by supply and demand, and money can act as a medium of exchange.

Great. So when the cost of software goes down because some people do it just for fun, it's reflected in the price of software. Resources are still allocated very efficiently, and the society as a whole benefits despite the facts that some programmers now need to find another job.


We express our relative preferences every time we buy or decide not to buy something. Via a non-zero price, our value system gets a say in the allocation of resources. This mechanism is completely gutted by a zero price, and a lot of valuable allocation information gets lost as a result. This leads to an undersupply and over-demand for the zero-priced good. Society is worse off because of
the resulting misallocation.


Suppose someone finds a way to make diamonds by placing a piece of charcoal in the microwave oven for a few minutes. Will this hurt the economy? Will this make society less efficient? Of course not.

It may be bad for some people in the diamond supply business, but may create new opportunities for people who cut and polish diamonds, make jewlery, etc.

Similarly, we're now discovring new mines of programming code, at zero or almost zero cost (I don't really agree with this, as many Open Source projects are maintianed by paid programmers - but let's assume all FOSS programmers are volunteers).

What's the problems here to society? There isn't any problem. The most efficient resource allocation may be to convert all paid programmers to do other work. Or it may be so that the "mines of code" only produce certain kind of code, and other kinds still require paid work.


Exchange? If you give away something you have not exchanged anything.
You have made a transfer outside of the economy. This has nothing to
do with a free or unfree economy. It certainly is not the cornerstone of a free economy. Or capitalism. 

You really don't understand FOSS.

It is an exchange. I give you some code, you give me some code. Together our code is a program. Whether other people find it useful or not doesn't matter. The exchange is for our benefit, and the rest is just a side effect.

People start Open Source projects because they want or need software that doesn't exist or that they can't afford, and they want to exchange their ideas and code with others, to create software. Sure other people can download this software and use it without making any contribution, but the initiator(s) of the project don't care - they got the software they wanted.

By the way, even free riders contribute QA work in the form of bug reports, and ideas in the form of feature requests. As it happens, people who work on FOSS projects interact with users of their software better than Commercial software vendors. But that's another issue.


The effect of FOSS on Commercial software, is that it changes the economics of software. No value is lost. Money is redirected from some software markets to other software markets or to non-software
markets.

As I pointed out above, money is not value, so this argument is dead
on arrival.

Money is not value, therefore the fact that money flows differently doesn't make any difference to the market as a whole.


Furthermore it embodies the very serious logical error commonly called the Fallacy of Composition, in that it confuses the whole with a singular part. The argument that FOSS hurts software developers is not disproved by stating that society as a whole is not
hurt by it ( a claim I also found specious ).

FOSS doesn't hurt all those developers that get paid to produce FOSS, and it doesn't hurt all those programmers who have other jobs but very much enjoy programming, so they write software for fun.

FOSS hurts (so far only in theory and not in practice) those programmers who produce software that costs more than many people can afford, to the point that those people decide to do it on their own.


Any attempt to prevent this market-driven change is anti-free-economy and anti-capitalist.

Capitalism is the allowance of voluntary exchange of private property. I fail to see what this has to do with the argument that FOSS undermines software developers' wages and careers.

I fail to see how FOSS hurts developers wages and careers. If anything it makes programming more accessible and lowers entry barriers - good for many programmers and would-be programmers.


So why do they argue against FOSS? Because they're afraid of change.

Not all change, just change that is destructive of the careers of software developers.

I'm a developer, I make my living from Commercial software, and FOSS only helps me do it - I can get a lot of free stuff that helps me build and sell my product. I couldn't do this 10 years ago, so FOSS is a very welcome and very *constructive* change for me.

I also contribute to some FOSS projects, and I hope that my modest efforts help other people make a profit or just get something they want and couldn't have before.

I think the world, and me in it, are better off this way.

Eyal.
.