Perens Discredits Mundie's Attack On GPL

SaxMan101 writes "CNET has an editorial from Bruce Perens that quite handily dismantles Mundies attack on the GPL and the Liberty Alliance. He takes the time to make YA strong argument for free software which he backs up with real numbers. Well said, worth the read."

Comments

Strong argument?

[platos_beard]

I think there's a strong case to be made for free software, but this ain't it. Bruce Perens touts the money saved by not buying MS software, but completely ignores the much more significant expenditures on people to administer all this software. Does it cost more to administer sendmail than Exchange? Apache vs. IIS? Is in-house development with VB cheaper to get the same results as Java on Linux?

I'm not sure how the numbers balance out, but these concerns far outweigh the price of buying the software. If Mr. Perens is going to dip his toe in TCO waters, he'd be better be sure he can jump all the way in and not get himself drowned.

Re:Strong argument?

[JordanH]

Without doing any real research, I couldn't say what the TCO issues really are. It's been my observation that sendmail is cheaper to maintain that Exchange, Apache cheaper than IIS, etc., but I don't really know, and I doubt that you do either.

I just notice that they're always doing maintenance on the Exchange server, but I rarely hear about problems with the sendmail gateways here. Same goes for Apache vs. IIS.

But, this is somewhat a distraction from Bruce's point. Actually, Bruce Perens in the article actually tries to avoid the economic issues and instead focuses on the control issue.

It was Bruce's thesis that the control issues, through people benefitting from competition in those to support and extend the products they use, will lead to lower prices.

I agree that the TCO issues are complex. In fact, they are too complex to really address naively. For example,

• Is in-house development with VB cheaper to get the same results as Java on Linux?

Please tell me... How do you get the "same results" with VB/MS as Java on Linux when the Java solution can be deployed across platforms, giving you potentially huge advantages in deployment flexibility?

If, for example, you were able to deploy to near-zero administration Terminals based on Java/Linux and you needed to deploy tens of thousands of seats, who wins then?

Sure, .net may do similar things someday, but what if MS starts ratcheting up the licensing fees? Any guarantees against it? With Open Source you always have the option of competing support groups or self-maintenance if a product requires extension or maintenance. This is dicey with Closed Source products where you are often forced to upgrade or have to live with the problems if the vendor has decided to take the product in another direction.

You see, static analysis of what TCO is today is a secondary concern to the control you gain with using Open Source. I think that would be what Bruce might say, at least.

Re:Strong argument?

[GSloop]

You ARE right...but it's truely ironic.

MS is THE premiere player in trying to sell cheap/free until they get marketshare, then raise the price.

Here in Portland OR, MS convinced First Interstate Bank to install NT 3.11 instead of Netware in their new 3000+ employee loan center - and almost purely because it was cheaper than upgrading from Netware 3.X

The bank didn't have any serious tools (Sniffer etc) in the old environment, and the hardware was ancient. But the OS was going to save them like 20K+.

What most of the outside world didn't know, was that the network went down almost daily for months. The result was thousands of people sitting idle (a double drain - their getting paid, and NOT making money).

Finally, after coming hours from chucking the whole thing, the MS engineers finally called the ONE guy who wrote the TCP stack. After a short conversation, the MS programmer suggested an undocumented TCP stack option. All of a sudden, the SNA session timeouts just stopped.

The point? MS SOFTWARE was like 20K cheaper, but the whole experiement cost the bank like HUNDREDS OF THOUSANDS (possibly even millions) of dollars. Even without the NT problems, the costs were VERY substantial to switch vs. staying on Netware.

MS has used the "IT'S FREE" or "IT'S LOTS CHEAPER" approach more than I can believe - Office bundled with the OS, (Office 4.3/95) IE/OE, NT (early on), MS Windows Plus, Windows 2.X-3.X (Bundled with Windows).

BP may not have dealt with the entire problem, but frankly, the PHB's aren't looking at TCO. If they were, we'd have run screaming from IE/OE a LONG time ago. We'd have set ourselves on fire when we see the rising cost of Office (Now it's MUCH more expensive than before - un upgrade used to cost like $200, now it's like $400), and the moving platform of MS's site licensing (I forget what MS called it - I think it used to be License+, now Select something? Doubled and Tripled in less than 5 years - loss of concurrent licensing)

PHB's only see the INITIAL costs. If they are concerned about TCO, they will look at the HUGE problems with viruses, crashing boxes (Re-Image anyone?!) and lots of features that really waste time and aggrevate users. (Clippie Anyone? How about how Word decides how you REALLY NEED that numbered list done etc!)

Sure, it's difficult to learn a totally new platform. But I do think that the Linux platform isn't any more difficult to administer. Ever tried to figure out Active Directory - it's got me confused! [Grin] How about when Exchange just stops sending mail in or out, but everything LOOKS fine - but a reboot fixes it? What about when IIS gets remote rooted and you get to rebuild your entire server?

I don't think you were defending the MS status quo, but even if you were, I think that defending MS will be a loosing battle in TCO. Bugs and security problems seriously compromise the TCO calculations on ANY MS software.

Finally, TCO numbers are SO perfect for manipulation. Everyone can make TCO numbers say anything they want. It's like the 10 year USA Gvmt budget. You can CLAIM you know where things are going to be, but frankly, you don't have a clue. TCO is usually just a massive marketing ploy.

Cheers!

An interesting perspective

[EschewObfuscation]

If you drill down a bit you find this letter from a programmer that complains about Open Source. While I found it both sad and funny, it does shed light on how Microsoft and other commercial software vendors view the movement.

To summarize: OSS is a bad thing because if free software is available no one will want to pay for software, which will drive programmers out of work. OSS is good in that it establishes competition for Microsoft, but that competition is better done through litigation or other commercial software.

Applying this point of view to Microsoft is humorous, of course, considering what they did with IE.

I actually don't think the developer has a point, though. Open source software has created far more jobs than it took. Linux, Apache, and other free platforms and development tools have meant, in my experience, that corporations are financially able to deploy systems that would otherwise have been prohibitive. The spread of such tools has also increased the number of people who are exposed to them - how many people would be running personal Unix systems if they had to have commercial systems? These people are able to get jobs in IT they would otherwise not be qualified for, or perhaps even know about.

In any case, Perens' response likening software development and protective measures against open source competition to buggy whips (actually ice, in his analogy) is only half the story.

Re:An interesting perspective

[tshoppa]

If you drill down a bit you find this letter from a programmer that complains about Open Source.

The most paradoxical comment within this letter is this:

The RIAA wants its intellectual property (music) to be protected. Authors want their books protected. I want my industry's intellectual property to be likewise protected. Is this too much to ask?

In other words, he think that the way to protect his intellectual property is to ask that it be illegal for others to give away their intellectual property. And this isn't too much to ask. Scary thought.

He also seems to give the RIAA implicit control over all music, but that's another flame war.

Open Source arguments to justify Free Software

[renehollan]

Perens' article was a good rebuttal to Mundie's FUD slinging, but it left me wanting: it was an open source justification for a free software license.

While all the points he makes are true, and the economic beneifits of free software are obvious, that is not the primary moral justification for software being free. Repeat after me, "When software is free, the world is a better place."

Now, it stands to reason, that part of the world being a better place is certainly the economic benefit that free software provides to reduce operating costs. In fact, one could argue that if there were no such effect, free software wouldn't be too great a thing -- who'd want it if it had no value (rather like some excuses for programs I've seen)? And they'd be right. These are open source arguments, though, and miss the fact that freeing software not only results in lowered operating costs for businesses that use it, but it changes the every environment in which they operate.

There are two primary schools of economic thought: planned economies and free markets. Politically, you have the statists on one side and the libertarians and anarchists on the other. Proponents from both sides argue that "their" way serves to distribute scarce resources in the most effective way, and that's what we want, no? -- effective distribution of scarce resources.

Well, yeah, but that doesn't make the scarcity go away, does it? Oh sure, the technological advancements that lead to efficiency improvements do eventually trickle down to everyone so that certain scarcities are less visible, but that's just a kludge. Think water. Most cities have methods for distributing drinking water to the point that, although the amount of water available may remain the same, it hardly seems locally scarce, even though it may have come from far away.

Free software serves to reduce the scarcity of good code out there. It provides value without relying on scarcity as the source of that value. It is a threat only to those who seek to leverage their possession of a scarce resource for maximum value. Now, if that resource is naturally scarce, fine: once sold, it is gone. But if the resource is artificially scarce, you can manufacture more of it for no cost, and charge whatever the market will bear, for pure profit (until you saturate the market, that is, but time-limited use licenses take care of that "problem" -- Microsoft's latest licensing strategy). It gives the owner incredible power over society as a whole (until society revolts).

But it costs money to produce code! People can't afford to give it away!! Well, if they depend on making it scarce for their livelyhood, no, but that is a bootstrapping problem: you make something artificially scarce in order to deal with real scarcities in your life. You'd have to do this less if there were less scarcities to worry about (imagine if we had solar-powered food-generation machines). And indeed, some have managed to give code away. RMS has done this exclusively, though by living a rather austere lifestyle. His choice. Others give code away when they can afford to. Each such contribution changes our environment for the better. For hackers who breathe code, this is, of course, a godsend (RMS, an atheist, might not like that choice of wording -- "GPLsend" then). Perhaps that's why we like the GPL so much, even those of use that produce restrictively-licensed code for a living.

So, you don't need economic arguments to defend the GPL. It is as good and wonderful for the world as are the lack of patents on fire, wheels, and language. The only people who will criticize it are those that profit from the misery that scarcity brings.

Re:The GPL is bad...to WHO ?????

[cperciva]

Who is the GPL bad for ?

I'd add at least the following to your list:

3. Those who would like to use code, are entirely willing to give credit where credit is due, but haven't decided yet if they want to (or, legally, are allowed to) release their own code.

4. Anyone who wants to see open standards. It was only the existance of free-for-any-use code which lead to the global use of TCP/IP -- back when every company had their own proprietary network protocols, the only reason they added TCP/IP support in was because they could do so (almost) for free.

5. Anyone who wants commercial software companies to release their source code. Companies which operate by selling software are never going to GPL their code; they might, on the other hand, release it under a less restrictive license which would allow them to incorporate improvements back into their own codebase.

Mundie needs an economics lesson

[Wateshay]

What Mundie doesn't understand (or chooses to ignore) is how wealth is created. Simply passing wealth back and forth between companies doesn't create wealth. Paying taxes doesn't create wealth. Government spending doesn't create wealth.

Wealth is created by increasing efficiency. If I pay you $10/hour to build widgets worth $3 a piece, and you can build 4 widgets per hour, then I make $2/hour profit. If you figure out how to increase you efficiency and make 6 widgets per hour instead of 4, my profit has now increased to $8/hour. This can then be reflected in increased wages for you, fewer work hours, or a cheaper product. Regardless, net wealth of the economy has been increased, since more output is produced from the same input.

Where does the GPL work into this? Because one GPL application has effectively infinite supply, it drastically reduces input costs of production and therefore leads to a very high net increase in the entire economy's wealth. Commercial software necessarily leads to less wealth increase, because it has an artificial cost added to increase the producer's personal wealth at the cost of the whole economy's net wealth.

Mundie's argument is that the artificial cost is necessary for software to get produced, because there will otherwise be no incentive for the producers to produce software. The thing he ignores, though, is that obviously the software does get produced. If OS software gets produced, then it is out there. It has increased the net wealth of the economy. That increase will never go away (unlike the commercial company, which could go out of business). If OS is not enough incentive for the software to get produced, or OS doesn't lead to a solution that is sufficient, then the demand for a commercial version will be high enough that commercial development will be supportable. There is room for both.

Microsoft, of course, is just beginning to realize that the software they make is quite compatible with OS development, and there is no way they can compete with the efficiency of an OS product. Therefore, Mundie is arguing that we will all be better off if the economy's net wealth is held down in favor of MS's personal wealth gain. I just don't buy it :-)

On the other hand, he's absolutely right that there may not be as high a demand for software developers in the future. So what. So, a few programmers may have to change careers. They're smart people (and yes, I am one), and shouldn't have too much of a problem finding work. Yes, it sucks for a few, but where would we be today if we always held back progress in favor of old, established industries. Not to be cliche, but I'm sure the development of the automobile sucked for the buggy whip manufacturers, too. Personally, I'll risk my long term personal stability for the chance of great wealth increases for both myself and the economy as a whole.

Re:The GPL is bad...to WHO ?????

[CDWert]

I write code, I can do whatever I want with that code,

If I choose I can sell it, (Its value is lessened if any tom, dick, or harry can, and legally, get it elsewhere) I can do so.

I can open source it, in doing so I am granting others the right to use it, Under whatever liscence I deem appropriate(remeber its my code)

Or I can let it sit on my hard drive and rot. not much use there.

But If I write the code, its my choice ho I make it available, If I am gracious enough to give it to the world, why should they dictate the terms under which they would like to use it ? That greed, and being ungrateful.

I have MANY time seen things I needed similar solutions too and said damm be nice to use that in my project, but I couldnt, mine was proprietary theirs was GPL, so what, I wrote my own. No bitching no moaning, just an understanding its not my code who am I to tell someone else how to make the code their blood and sweat into available to me under my terms, I wouldnt do it and I dont expect anyone would do it to me.

I am a capatilist, simple period. If I can make more money using open source I will, but if it interests dont meet mine, I will write my own code to fill that need.

Guilt unnecessary

[Bruce Perens]

It's not necessary to feel guilty! We want users. After all, writing software that nobody else uses would just be playing with ourselves. So consider that you save us from much embarassment :-)

Have you considered being a technical writer or something? There are many ways that anyone can help.

Bruce

Fundamental difference in Linux vs. Windows Admins

[Bruce Perens]

I was at a CTO roundtable the other day. A New York City investment banking CTO was talking about the difference between Linux/Unix admins and Windows admins (yes, Linux is widely employed in investment banking these days). Linux admins script a fix and don't touch it again, they just re-run the script. Windows admins don't script, for the most part. They push the same buttons on each system. This might be a big factor in increasing the Windows TCO. The bank claims they have many more Windows admins per system than Linux/Unix admins. I'll ask the CTO to write an article.

Bruce

Re:Microsoft has blinders on

[Bruce Perens]

You're infuriated! Are you sure that coffee you've been drinking isn't too strong :-)

Whether Perens and crew acknowledge it or not, what they are in actuality saying is that software development is an exceptional sector of our economy where regular rules needn't apply

Yes, I've made this explicit many times. It takes a pound of flour to "copy" a loaf of bread. In contrast, once you have amortized the cost of creating a piece of software, there is essentially no marginal cost associated with creating another copy. The result of this is that the current proprietary model drastically overvalues software. You complain of IBM and HP computers being overvalued with respect to the raw material cost. As we drive the market toward commodotized software, it becomes more competitive for hardware manufacturers. If they have high margins, isn't it because of anti-competitive factors like customer lock-in?

Can we amortise the creation cost of software without a direct revenue capture per unit sold? The answer seems to be yes for a lot of people. And why would this be important? Because decoupling the money from the process makes the mechanics of collaboration a lot simpler. Collaboration works to distribute cost, making it tolerable, and improves efficiency by avoiding redundant development. That redundancy happens all too much for "in house" software, and businesses have recently realized that they can collaborate with their competitors on non-differentiating software. This is not to discount the entire "freedom" agenda, I simply need not argue in those terms this time.

Perens and crew are basically selling out software development as a profession

This smacks of the old guild system which operated to support costs rather than allow the free market to set them. It seems anticompetitive. But yes, if you want to consider me as selling out the software development profession, I'm doing it for the customer. People seem to forget that capitalism is supposed to operate for the ultimate interest of the customer, by keeping the costs that the customer pays as low as possible.

Regarding your argument about software developers providing and everyone else consuming, most people are able to participate in a free exchange of information. In this same topic we've been carrying out a thread about how an illustrator can help.

Bruce