Fog Creek Software
Discussion Board




Eric Raymond and Richard Stallman

Somebody discussed about the open source movement leaders such as Eric Raymond (ESR) and Richard Stallman (RMS), and asked how they pay their rent.

These are guys who live on grants...

I remember that RMS (or ESR - I don't remember) said that he doesn't own a car, because he thinks the kind of lifestyle that requires a car is too expensive for him, and that he lives in campus dormitories.

Well.. that may be fine for him, but it's certainly NOT the lifestyle I am willing to accept!

John
Monday, December 08, 2003

http://linuxtoday.com/news_story.php3?ltsn=1999-12-10-001-05-NW-LF

That (notorious) story, for those who don't follow the link, is about how ESR basically got handed 150,000 shares of VA Linux when it went public in 1999 (just for being a swell guy), and how, for at least a while, he was worth close to $40,000,000. Those 150,000 shares at today's LNUX prices would be worth around $540,000 -- still nothing to shirk at, but it's a world different from the original amount. Of course LNUX has probably been continually dumping shares in his lap while the price submarined, so ESR may still be quite wealthy.

So what did ESR do when all this money (enough to pay for 450 person years of a pretty high developer's salary) seemed to fall in his lap? Did he refuse it, all to enhance shareholder equity and avoid stock repurchases that would lead to price increases to the end consumer (per his recent moronic comments)? Did he talk about how he would set up a Linux deveopment centre to enhance and expand the technically capabilities of Linux?

Of course not.

Instead he went on a tirade basically proclaiming that yeah, it's great that he got lots of money, but don't expect anything for free -- he still expects his laundry list of financial demands to be met to speak at user groups, etc. Extremely unsavoury little rant.

Dennis Forbes
Monday, December 08, 2003

Open source makes the software development world look more like the rock star world:

- a handful of guys (RMS, ESR, Linus, Miguel) get a lot of money and recognition

- the other 99% of the guys barely make a living

Of course, the situation is not yet the above, but it may be, if the open source march continues.

Marinchip
Monday, December 08, 2003

Eh, so, ESR got some money?

Who cares? It's his personal life, if he got money from it and lives on it, great, good for him.

Who the hell are you to say if that's a good or bad thing? Really, who do you think you are?

Please tell me how much you make and how you spend it so I can rant at you about it.

fw
Monday, December 08, 2003

ESR got a lot of money, while at the same time preaching and pushing more and more programmers to work for free.

Isn't this a bit hypocritical?

John
Monday, December 08, 2003

I don't care. You are in no position to judge him, are you now?

You didn't tell me about where you work, your name, and how you spend your money yet.

fw
Monday, December 08, 2003

Normally, I would say that I'm in no position to judge ESR.

But ESR, and other vocal members of the OSS community feel that they are in position to judge other programmers, and especially programmers who don't want to give their work away for free.

So - if ESR and the OSS programmers are judging us, the commercial programmers, then we have the right to judge them, too.

And also, I have the right to judge ESR because he is a public figure, like a politician.

If a politician says something like "everybody should fight against pollution", but owns a factory which pollutes the environment - then yes, I have the right to judge him.

John
Monday, December 08, 2003

Firstly, fw, myself and John are two different people.

"It's his personal life"

Uh, no, it isn't his personal life. He sat on the board of a _public_ company whose whole existence was on the backs of the many people busy developing in their basements.  LNUX handed ESR a theoretical $40 million that was rewards for the work of thousands of other programmers doing their work for free (why for free? because people like ESR and RSM tell them that that's the way it should be).

"Who the hell are you to say if that's a good or bad thing? Really, who do you think you are?"

Wait, isn't this the ESR who just proclaimed that it's a good thing that software developers get their salaries chopped in half? Yeah, who am I to critique Mr. Raymond.

Dennis Forbes
Monday, December 08, 2003

I can see your point, really, however I'd have to say that if you insult his code, or writing, in some way, that's fine, but I think it's going over the line when you attack his character based on money.

I think the guy has a serious ego problem, along the lines of deraadt and djb, but I don't think you can attack the guy becuase he was donated money. A large mount of OSS developers have a paypal donate, or amazon buy-me links on their page, granted it's not on the same scale, but ESR did well out of this, but do you think he'd still be doing it if he hadn't got this particular donation? I think he would be.

fw
Monday, December 08, 2003

Dear John,

I don't know if you realise that actually you don't know anything about this guys, for example you call RMS a leader in the Open Source movement which is a sure way to make him angry at you, and you question ESR right to judge closed source programmers as evil when he does no such thing.

Bothering to know what you are talking about is no fun, I know, but still it is a good thing.

If you want to write closed source software, go ahead, do it, who cares? and why should you care about what other people do with their time?

Andres
Monday, December 08, 2003

Andres, I know that RMS wanted to call it "free software" and was upset about the "open source" name.

But now most of the people call it "open source", and admit that the terms "open source" and "free software" are almost equivalent.

So, while it may be upsetting to RMS, the expression I used ("open source") is used correctly, according to most people.

John
Monday, December 08, 2003

"you call RMS a leader in the Open Source movement which is a sure way to make him angry at you"

RMS heads up the GNU foundation. The GNU foundation is entirely about "Free Software", with the "Free" referring to _source_code_availability and unencumbered rights. If this doesn't make him a leader in the open source movement, then what in the world does? Or is this one of those "hacker/cracker" ridiculous pendantic arguments, and it's error to call him anything other than a "Free Software" promoter?

"If you want to write closed source software, go ahead, do it, who cares?"

ESR heads up the Open Source Initiative (or at least he did. Maybe with all of that LNUX money he took up different pursuits). RMS heads up GNU. Both are propaganda, err evangelist, organizations for the sole purpose of replacing closed source software with open source. THEY ENTIRELY CARE about closed source software, and try to do whatever they can to eliminate it.

"and why should you care about what other people do with their time?"

Per the above -- As software developers whose field is absolutely being impacted by the efforts of RMS and ESR, we have every right to care and comment upon what they did with their time. If someone started up a "Closed Source Initiative" organization, I would fully expect (and demand) that ESR and RSM, and all of the open source fanatics out there, would take them to task for it. Don't proclaim that we can't do the same about them.

Dennis Forbes
Monday, December 08, 2003

I'd like to add one more thing: other people may do whatever they think is good for them. I don't mind.

But, some times, some people may think that something is good for them, while in fact, in the long term, it is bad for them.


I belive that this may be the case with open source:

Enthusiastic people write open source software, and they feel good about the technical challenges, the peer recognition, and other things.

They think that it's good for them, and for the world, to write open source.

But, in a few years, they may discover that their job disappears because it's been taken away by a free piece of code, written by another open source programmer.

Also, they may discover that programmer's salaries drop, and their salary is cut in half because managers see that open source is developed for free, and starts to belive that software development is a kind of work that can be done for very little money.

They may find that they have to put up with low income or give up and find a field of work other than programming.

John
Monday, December 08, 2003

Yes - Open Source vs Free Software is one of those very pedantic quasi-religious arguments.  It seems to have settled down a bit now but at one stage it was real blood on the floor stuff - a lot of which was more about the personalities involved than material differences in outlook. 

The potted version of the two phosophies are that Free Software is about the freedom of the user and open source is about a peer-reviewed development method.  Both arguments have some merit as well as some holes.  Both also originate in an academic's view of software. 

In practice it doesn't matter that much but it is possible to be in favour of open source and not free software (i.e. to use licenses which are "Open Source" but not "Free" ) or to distribute "Free" software without an open development  method.

A cynic writes
Monday, December 08, 2003

I belive that this distinction doesn't matter for the current discussion.

John
Monday, December 08, 2003

Probably not John - but it is the sort of thing that gets RMS frothing at the mouth. 

Possibly the key thing you're missing is that  FOSS (there that keeps everyone happy) comes from academia - where initially (for say the first 20 years) programs were "free".  In essence the free software movement is a reactionary not revolutionary one.

What I find more interesting is the number of firms who accept *some* of the arguments and publish their source code but without an open licence.  Such as our hosts I believe.

A cynic writes
Monday, December 08, 2003

I read the "Surprised By Wealth" essay and didn't think it was too bad.  Like all of ESR's writings, it's bombastic, and in this case perhaps a little too fake humble ("all this money, and shucks, I'm just a plain ole hacker").

What I'd like to know is how many people would be OK with writing an article on your Internet stock *at all*, but offended by this one? In other words, how many people feel that he should never have written the article at all, regardless of what it said?

Portabella
Monday, December 08, 2003

It seems that you don't see the problem. We are not offended by the fact that ESR made lots of money. That's his business. We don't care.

However, if he preaches a lot that all software should be free or open source, and preaches that programmers should all work for free, and then makes a lot of money as a result of this (that is, on the backs of lots of programmers working for free because he told them to work for free), then ... I don't know, this sounds very bad to me.

John
Monday, December 08, 2003

> if he preaches a lot that all software should be free or open source, and preaches that programmers should all work for free

As far as I know, Raymond does not advocate either of these.  I do agree that it is a bit ... much ... for someone who is probably at least a millionaire to tell other programmers to be glad that their salaries are dropping.

Even Stallman, who is far more radical, knows that programmers have to support themselves, and has standard agreements for folks with day jobs who wish to contribute to the GNU project.

Portabella
Monday, December 08, 2003

errmm...no he doesn't preach that programmers should work for free.  That *may* be the outcome - although that is debateable - but it isn't the point. 

The big idea of "free software"  (RMS) is that you (as a user) should have the right to use, modify and redistribute your modifications to a program whereas the big idea of "open source" (ESR) is that you should publish your source code so that any coder that looks at it can contribute.  Neither explicitly is about working for free.

Now as a thought experiment - at work we use a database costs us  ~$1500/user/year.  If it was "Free" software - what difference would it make ? and why? 

BTW I meant to post a pithy quote from our host here - but I missed and stuck it on the "open source" thread.  It applies to both.

A cynic writes
Monday, December 08, 2003

No one says authors of FOSS can't (or don't) charge for their work, they simply have a hard time charging per copy.

Devil's Advocate
Monday, December 08, 2003

"If someone started up a "Closed Source Initiative" organization, I would fully expect (and demand) that ESR and RSM, and all of the open source fanatics out there, would take them to task for it."

It's called Microsoft.  (And yes, the open source fanatics do.)

Jim Rankin
Monday, December 08, 2003

John, nobody preaches that programmers should *work* for free (no money). They only argue about not selling a binary file (useless for modifying/extending).

Let's make a (usually inappropriate) analogy with the car industry. When you buy a car, you can change it any way you like, pay someone (different from the original manufacturer) to modify it for you, etc. With closed source software, you can't do the same thing. Please observe that if you don't know how to modify the car, you have to *pay* someone to do the job. You also must *pay* for the original car. Same with software.

About programmers being made redundant because there is a piece of FOSS that already does what the programmer was supposed to do: Why do you think you have the right to reinvent the wheel over and over again and be paid for it?

ice
Monday, December 08, 2003

John, its not like you can stop people from writing free software, so why worry?

Very few programmers work shrink wrap stuff anyway. Everyone I know do custom solutions of one kind or another, and they benefit from open source software.

Im more worried about stuff like Dreamweaver MX that lets you put together a webapp with having a clue about whats going on. Ok, its not quite there yet, but soon it will be.
Perhaps there will soon be tools that do solid datamodels behind the scenes and lets any bozo put together the kind of stuff I make my living on.

But you cant fight evolution.

Eric DeBois
Monday, December 08, 2003

"Wait, isn't this the ESR who just proclaimed that it's a good thing that software developers get their salaries chopped in half?"

Did ESR really say this?  Do you have a link?

If so, I'm very surprised.  His backers are mostly coders, many of whom get paid to code in some capacity, so saying such a thing should be career suicide for him.

Jim Rankin
Monday, December 08, 2003

John, I don't understand why you keep arguing a position that boils down to trying to keep a house of cards intact.

If you really can be outprogrammed by some college students in Europe, why should anyone pay for your product? You seem to want to institute some sort of union work rules where every programmer has to be employed rewriting software for problems that were solved years ago, just to keep them employed.

FOSS software is the salvation of the software industry, precisely because it will free us to solve new problems instead of trying to re-solve the old ones. If your talents end at rewriting basic application servers, I suggest you're in the wrong field.

jason

JasonB
Monday, December 08, 2003

Newsflash: "People developing software on grants say software development should be grant funded through  taxes".

Duh!

Just me (Sir to you)
Monday, December 08, 2003

> Did ESR really say this?

Yes, he did, see http://esr.ibiblio.org.

The quote is taken out of context: he's arguing that the Free Market is a good thing, even when it means that our salaries fall.

Portabella
Monday, December 08, 2003

As others have said, ESR and RMS don't advocate that programmers should work for free.

They generally say that software development should be sold as a service, not that software should be sold as a product.

Coincidentally, that actually matches the way the software industry works; most people who are employed to write software are employed to write software for use, not for sale as a product.

Chris Hanson
Monday, December 08, 2003

"Let's make a (usually inappropriate) analogy with the car industry. When you buy a car, you can change it any way you like, pay someone (different from the original manufacturer) to modify it for you, etc. With closed source software, you can't do the same thing."

Wrong. Closed source software can be modified in the same way cars can - through some set of (typically small) published interfaces, either provided by the car (application) or by the parts vendor (operating system).

No car of which I'm aware provides full disclosure of every component and interaction within. It seems to me that customization after-market is more like reverse engineering than anything else.

MD
Monday, December 08, 2003

If you provide the source to your product along with the right to freely redistribute and coopt said code, as the FSF promotes, then the market value of that code is very close to $0.00.  Given this, arguing that open source/free software and getting paid for software development are not mutually exclusive seems to be a living-in-denial propaganda (so often open-source crusaders try to put a sheeps outfit on their wolf). Whether you can make money in other aspects of the economic chain (support, or for many of the major distributions "hassle avoidance" [i.e. putting the distros on a Commodore 64 with a tin can internet connection to piss people off enough that they just give up and order a copy]) is absolutely irrelevant to the topic of getting paid for coding.

As to the other claim that open source is to avoid reinventing the wheel - I'm not sure what world you live in, but generally when I deploy a project it's on a commercial, closed source operating system, using a closed source development tool, using loads of closed source components, talking to a closed source database, and I've integrated all of the functionality together with my own value-added glue - open source doesn't have a lock on the idea of standing on the shoulders of others.

Dennis Forbes
Monday, December 08, 2003

esr is a nut. has anyone (tried to) read his art of unix programming book? also, pre VA linux, he survived because his wife is a lawyer.

_
Monday, December 08, 2003

Not so much living in denial as this thought:

It is possible to be paid for doing something by a company that the company does not directly profit from.

Not if you think about it too shocking.  If memory serves the operating systems of early main frames were - to all intents and purposes - public domain.  I bet the blokes who programmed them were paid.  If on the other hand your argument is that it is difficult for a company to be paid directly for free software - I agree - but that's not entirely the same thing.

A cynic writes
Monday, December 08, 2003

Fair enough, though this really is like saying that IE is free -- the cost is being buried in another product (i.e. in the IE case you're paying for IE when you buy a copy fo Windows, and for your mainframe example they can pretend that the software is free when you're paying a huge markup on the hardware).  Generally such submerged costs are disingeuous and bordering on fraudulent, and humorously they exist when there is limited choice for the consumer (such as cheap inkjets when you have to then refill with the grossly overpriced cartridges from the same manufacturer).

The other prospect of getting paid for free software is a variation of international trade "dumping" -- IBM, Novell, Sun, etc are all paying developers to work on Linux basically to undermine Microsoft. Again this isn't legitimate commerce.

Dennis Forbes
Monday, December 08, 2003

As I understand it the FSF's point is that it is unethical to distribute programs without source code as this prevents the users of the programs from making the changes they would like to make. Closed source allows the programmers to exploit the users by effectively holding them to ransom, free software (freedom, not beer) prevents this, if the user doesn't like the service provided by programmer A they can take the source to programmer B. John's view seems to be that it's fine to exploit the users, in fact, screw them for as much money as you can cause it's the programmers who are important, and FOSS undercuts this by empowering the users and lowering prices.  The FSF thinks this is wrong, and that the users of a program are not just there to provide money to the writers of the program, but should have some rights.  Wanting to keep the wages of programmers high by exploiting the users of programs seems a bit unfair.

If all cars had to go back to the manufacturer for any service, at whatever price they decided, and someone produced a car you could service yourself, or contract with someone else to service, which would you prefer?

Alex
Monday, December 08, 2003

> Again this isn't legitimate commerce.

It's both legal and good business practice for them to do so, so in what sense it is not "legitimate"?

Portabella
Monday, December 08, 2003

Then it was legal and good business practice for Microsoft to distribute IE for free specifically to undermine Netscape.

Actually I'm saying this a bit disingenuously because I personally saw no problem with Microsoft's actions in regards to the whole IE/Netscape fiasco, but I guess my point is more of a "see it for what it really is, and not what the ideologically-tinted visors make it appear".

Dennis Forbes
Monday, December 08, 2003

> Then it was legal and good business practice for Microsoft to distribute IE for free specifically to undermine Netscape.

1/2 right.

It was (unarguably, I think) good business practice for them to do so, but illegal under current U.S. law.

Portabella
Monday, December 08, 2003

"Then it was legal and good business practice for Microsoft to distribute IE for free specifically to undermine Netscape."

When you already have a monolopy in some item A you are not allowed to use that to produce a monolopy in some other item B. This was all decided at the beginning to the last century because, iirc, of a little company called Standard Oil. So, basically, that's why it may have been "good business practice" (all hail the capitalist gods) but it certainly wasn't legal.

Alex
Monday, December 08, 2003

"""Then it was legal and good business practice for Microsoft to distribute IE for free specifically to undermine Netscape."""

It would have been, if Microsoft weren't a monopoly. Rules for monopolies are different in most of the western world, plain and simple.

Democracies and Free Markets are great, but they are unstable in the sense that they don't stay that way for long without some non-democratic limitations and various monopoly and disclosure regulations respectively.

The whole "price dumping" idea is fuzzy when applied to Software (and information in general), for which the marginal cost of producing an additional item is close to zero.

Ori Berger
Monday, December 08, 2003

The only reason IE won the browser war was because Netscape started to suck royally.  All I used IE 1, 2, or 3 for was to get NN.  Once Netscape did that whole Communicator crap I abandoned them. 

It had nothing to do with IE being distributed with Windows, in my experience.

nathan
Monday, December 08, 2003

> The only reason IE won the browser war was because Netscape started to suck royally.

I agree that Netscape began to suck royally , and I personally switched to IE 4 for that reason.

However, I also think it's obvious that any client application that Microsoft bundles has a huge advantage; WinAmp must be clearly superior to Windows Media Player in order to justify downloading and installing it.

The only two applications I can think of that are ubiquitous but *not* bundled are WinZip and Adobe Acrobat ... and Microsoft has an Acrobat replacement in the works.

Portabella
Monday, December 08, 2003

Definitely in agreement about the decline of Netscape.

Regarding Microsoft and bundling, note that Windows XP also includes full ZIP management as well, so Winzip is no longer necessary.

Dennis Forbes
Monday, December 08, 2003

"software development should be sold as a service"

Great! This is what Microsoft does -- you pay for a license to use teh softwarre for a term as a service only, not for the software itself. Excellent we are in agreement.

"Software should be modifyable and extensible"

Great! This is what Microsoft does - hence the scripiing and programming facilities that come free with each copy of Word and Excel, which are far more powerful and easy to use than distributing source code in C would be.

Any one who believes as I do that software should be sold as a service and the software must be extensible and modifyable and the extensions made by the end user should be redistributable, definitely must agree that Microsoft software is the best as it matches this description and has been more successful than any other software at exactly these points.

Millions of people worldwide enjoy the freedom to write extensions to Word and Excel.

Tony Chang
Monday, December 08, 2003

"open source/free software and getting paid for software development are not mutually exclusive"

This is true. Likewise, Chrysler should give its cars and SUVs away for free. This could be a profitable enterprise because it does not preclude them from also offering the same models for sale as well as giving them away. Som customers will want to pay for the exact same cars that are being given away for free at the same dealership. If there are any concerns about the profitablility of this model, consider how much income Chrysler will make from charging for oil and tire changes -- when cars are free, more people will have them. that much is obvious. These new people in the market will require oil changes and custom paint jobs, the income from which can fund the development of new models of cars and new features.

Tony Chang
Monday, December 08, 2003

Tony - "which are far more powerful and easy to use than distributing source code in C would be".

Ehh... run that by me again? I get the easier part, but how does a published VBA interface let you do more than the source code would? Perhaps its easier and more useful, simpler to develop effective scripts quickly, whatever. But there is no way you can possibly argue that the extension interface is more powerful than the source code. I cannot begin to formulate an argument against your assertion, as you are so bafflingly wrong.

And furthurmore, your statement that Microsoft software is the most modifiable, extensible, etc., I have to say you're wrong. Having a good API you can use to implement extensions in is not the same as having the application's code. Now maybe your favorite OSS project doesn't have hooks for scripting, but you could put them there.

My point is that there is nothing more extensible than the code, and to argue counter to that is downright foolish.

Your statements are of the same vein as the statements made by OSS fanatics that are criticized here. I find such fantastic arguments annoying irregardless of the side that makes them.

I continue to be boggled. The boggledness is not dissapating. D'oh!

Mike Swieton
Monday, December 08, 2003

I'd like to add that just because you don't WANT to spend the time adding those features to XYZ OSS project doesn't mean that you don't need to accept the ability to do so in your comparison. Your comparing based on what can be done, not what you want to do. Choosing Microsoft over Open Source because you'd rather write a VBA script is valid. Choosing Microsoft over OSS because you can make more changes and customizations is so utterly off that I cannot begin to contest it.

Mike Swieton
Monday, December 08, 2003

I think Tony was being a bit hyperbolic (ph 3.7) as a sort of inverse of many of the open source claims.

Having said that, it really isn't that outrageous to make the proclamation that for an average development situation, with a developer trying to solve a problem within realistic confines like deadlines and domain knowledge, fixed APIs with closed source products are often _much_ more poweful than throwing the source code over the wall. For example if you made a compression utility with a fixed, well thought out set of entry point APIs that properly black-boxes the functionality of the module, I'll probably get it going in my app no problem, empowering my app with compression with minimal effort, even though compression isn't my domain. That is power. To make matters even better as you upgrade your component I haven't done anything more than used the fixed contract API, so I can use your new component without problems.

If, on the other hand, you have a "you have the source -- figure it out!" attitude, it's often a complex mix of interdependencies and ordered calls, making integrating simple functionality a nightmare (basically requiring domain knowledge over the subject matter, and the whole codebase).  It doesn't _matter_ if I have all of the code because practically it is useless to me.

Of course what I'm really talking about is proper compentization, which is equally possible with either closed or open source, but the point is that proclaiming that having the source is more powerful than a cleanly defined API on a black-boxed component is fundamentally flawed, at least from a practical perspective.

Dennis Forbes
Monday, December 08, 2003

"but the point is that proclaiming that having the source is more powerful than a cleanly defined API on a black-boxed component is fundamentally flawed,"

well yes, but you are comparing apples and oranges theres.

If you have a cleanly defined API in the opensource, you _can_ treat that as a black-boxed component and therefore the two are equivalent, closed source or open source makes no difference if its a quality product.

Where the advantages arise is if the black-boxed component is _not_ cleanly defined and/or is buggy to all hell.  (<g> we've all used third party components before, right?  whether opensourced or not problems tend to crop up)

In the closed source version, the problems have to either be solved by ourselves by endless testing and/or exploration of the API;  or solved by the third party which is likely to, well, one way or another thats going to be at their schedule not ours...if we are lucky they wiill be _good_ and responsive, if not they wont be.

In the opensource version the above two options are still available, but we also have others.  We can work through the source code to work out how we are supposed to be using the API (duh! I should have been setting foo via SetFooValue() before I setup the bar notifier).
Or we can work through the damn code itself and find out where the bug is (stupid developer stupidly forgot to call the stupid constructor stupid opensource devepopers..)

So what it comes down to is that if your third party components are non-buggy, or _extremely_ responsive to support requests then its exactly equivalent to an opensource component that is of the same high quality..there are no advantages either way.

OTOH if the closed source component is buggy, non-responsive or goes broke 1/2 way through your development cycle then you are going to have a lot more problems than you would if you were using opensource components.

or, to put it more succinctly, when things are going _well_ closed source components are fine...when things go to shit through you are better off with open source components.

FullNameRequired
Monday, December 08, 2003

Dennis - you hit the nail on the head on that one. Exactly what I mean. Both sides are throwing excessive arguments here. People seem to get very extreme when it comes to open vs closed software. Here people flame open source as a bunch of hippy fanatics, and on slashdot the closed source guys look like a bunch of money-grubbing suits. Niether is anywhere near reality, of course, and it's disturbing how much exageration I see here. I expect it on slashdot, but I expected better here.

Mike Swieton
Monday, December 08, 2003

Actually, IIRC, Joel used similar logic to justify his providing access to the code for certain of his products.
In his case it was for registered users only of course, but by doing that he is taking advantage of the same set of facts.
When things are going well, not having access to the source is fine, but if things start turning to custard then its hard to overestimate the advantage that having access to the source code will give you.

FullNameRequired
Monday, December 08, 2003

"I remember that RMS (or ESR - I don't remember) said that he doesn't own a car, because he thinks the kind of lifestyle that requires a car is too expensive for him, and that he lives in campus dormitories."

It was RMS. Its a misquote but he does live spartanly.

"Well.. that may be fine for him, but it's certainly NOT the lifestyle I am willing to accept!"

He live that lifestyle by choice, so I guess by definition it is fine with him. Note that he's had plenty of money come his way. The half million from the MacArthur grant plus an undisclosed amount he received in stock grants from the various linux IPOs. But in each of these cases he put his money where his mouth is and given the money to the FSF.

John Eikenberry
Monday, December 08, 2003

Raymond is on the boards of a few companies, including a bank or two. He lives well encouraging programmers to work for free.

anon
Monday, December 08, 2003

In case anyone is interested, ESR's 'Surprised By Wealth' article is here:

http://linuxtoday.com/news_story.php3?ltsn=1999-12-10-001-05-NW-LF

"Anyone who bugs me for a handout, no matter how noble the cause and how much I agree with it, will go on my permanent shit list. If I want to give or lend or invest money, *I'll* call *you*. (Sigh...)"

I love it.  The sorry thing is the whole article reminds me of '99 in such a powerful way.  Reminds me of the ego, the smugness, the greed, and the 'forging new territory' kind of vibe.

Konrad
Monday, December 08, 2003

"there is no way you can possibly argue that the extension interface is more powerful than the source code"

I am making exactly that argument.

VB for applications is specifically designed so the user can easily extend the core program. The source code to open source projects is not designed with this in mind. People do far more with VB than they ever would have the patience for in C. C is great for implementing applications but is not the best or most powerful choice for end users to modify the programs.

Your argument is like saying that machine language is more powerful than C. It's technically true, but it's irrelevant because in practice people write far more powerful applications in C than in assembly.  Likewise, extensions to Word done in C, while technically feasible, are way beyond the ability or cost-effectiveness to make it a sensible choice.

This is why Word and Excel dominate the market -- they are extensible and powerfully so.

A nuclear reactor is more powerful than a jet engine but a nuclear reactor is not a suitable power supply for a fighter jet. The fighter jet is more powerful and attains greater results with the 'inferior' jet engine because the nuclear reactor comes with too much overhead and maintenance costs and weight to be a suitable solution for aeroplanes.

Likewise, in the domain of making software 'extensible' by the end user, C is one of the worst choices imaginable. If you don't believe me, look at sales numbers of Word and Excel to see what the real world numbers of the marketplace have decided.

Tony Chang
Monday, December 08, 2003

Let's say a user, even a 'technically advanced' one such as myself needs to extend the functionality of a word processor.

Now the least expensive way to do that is to buy a better word processor that has the features you need since even a couple hours development time is more expensive than a fresh new license for a top o the line word processor.

But lets say that we have the word processor that is most advanced already. So, what's the best way to extend it?

Use a specialized domain language designed by experts like Joel specifically for that purpose?

Or modify a source base with ten million lines of code, hope you understand all the things the original authors did, recompile and hope for the best?

Which is faster? Which is more reliable? Which has more high level abilities?

A few times there has been stuff I could not easily do in VB with Excel. I wrote to microsoft and the features were added in the next release.

For most people in the world, this is the way to go.

Those who advocate recompiling word or excel with your new features have yet to demonstrate a case where doing so was more efficient or more powerful. It's all a lot of hot air, while real developers delivering practical results are instead using tools appropriate to the job at hand rather than joining some pseudo-religious 'movement' asking for things that would be of no use.

Tony Chang
Monday, December 08, 2003

Mike Swieton,

Please I beg of you -- join a team working on a ten-million line code base and tell me how easy your precious C extesions are for you as an expert developer with years of experience.

If you truely believe that end users, even technically brilliant ones can easily add modify a ten million line code base that they were not part of the development of, you are seriously delusional or perhaps just plain ignorant.

Tony Chang
Monday, December 08, 2003

My point is that there is nothing more powerful than the machine language, and to argue counter to that is downright foolish.

Your statements are of the same vein as the statements made by OSS fanatics that are criticized here. I find such fantastic arguments annoying irregardless of the side that makes them.

I continue to be boggled. The boggledness is not dissapating. D'oh!

Tony Chang
Monday, December 08, 2003

Mike, let's have a content.

I'll work with Word. You work with the open office code base. We'll have a set of 10 non-trivial extensions to be made to each program. The programmer who completes their task first without introducing bugs to the program will be declared the winner.

Tony Chang
Monday, December 08, 2003

--
I'll work with Word. You work with the open office code base. We'll have a set of 10 non-trivial extensions to be made to each program. The programmer who completes their task first without introducing bugs to the program will be declared the winner.
--

I don't think he's arguing that access to the source is better than access to a domain specific language that was created specifically to add extensions to a program.

I think that he's arguing that access to the source is better if there isn't a well-defined, well-thought-out API to work off of.  All else being equal, it helps to get to the source.

Andrew Hurst
Monday, December 08, 2003

(contest not content)

--

"Anyone who bugs me to write free code for them, no matter how noble the cause and how much I agree with it, will go on my permanent shit list. If I want to give or lend or invest source code, *I'll* call *you*. (Sigh...)" - Tony Chang

Tony Chang
Monday, December 08, 2003

Having worked on a lot of projects, I agree with Tony Chang.

Having functionality exposed as a professional, robust, well thought out API is far preferable to fiddling with source code. It means no-one has worked out a good access regime; it means no-one has tested and verified the dependent functionality.

Show me a project that extends by hacking the source code and I will show you a project that crashes, that needs hand-holding every time it gets installed and probably isn't that terrific anyway.

analyst
Monday, December 08, 2003

In a recent commercial project I'm working on, I have a real need to have a 1ms accurate time-of-day clock accross a group of machines sitting in the same room.

If you look at the MS API, you'd think you can get that resolution. You can't - I've measured 10ms and 15ms and I don't even know what controls if I get one or the other.

My choices now? Forget all Microsoft Time Management code, and write my own clock and SNTP client. Or use an operating system that has sensible time handling.

Linux gives me the precision I need - but even if it didn't (and, at version 1.2 it didn't, if I recall correctly), what I would have had to do was change one #define somewhere (clock resolution), and recompile the kernel. Some of my colleagues actually did that at the time.

Can the "you don't need the source" people please direct me towards a solution? Because the project calls for both 1ms resolution and Windows, and I can't drop the first requirement under any circumstance.

Ori Berger
Monday, December 08, 2003

> As I understand it the FSF's point is that it is
> unethical to distribute programs without source
> code as this prevents the users of the programs
> from making the changes they would like to make.

Unfortunately the GPL license gives the receiver of the software redistribution rights.

If programmer A sells or gives away software to company B, then company B can sell and give away the software, too.

This is unfair for programmer A, and I belive it deeply hurts programming jobs.

Unfortunately, many programmers got seduced by the FSF rethoric.

This is why we need to tell them exactly what I written above.

The programmers that write GPL code should pause for a while and think about what they are doing to other programmers.


> Closed source allows the programmers to exploit
> the users by effectively holding them to ransom,
> free software (freedom, not beer) prevents this,
> if the user doesn't like the service provided by
> programmer A they can take the source to
> programmer B.

...and thus increasing the competition between programmers (or programming firms), and so decreasing programmer salaries!

And the crazy thing is, somebody convinced programmers that this is best for them! :)

John
Monday, December 08, 2003

"1ms accurate time-of-day clock accross a group of machines sitting in the same room."

Then you'd better put a nuclear clock in them, because it is unlikely that they're going to be that closely time synchronized. Even presuming that they're using NTP for synchronization (which Windows 2000 and above includes by default, and was an extra tool prior to that), network latency / software interrupts, and then physically setting the system clock all make a 1ms timeout unlikely.

Ignoring that, the implication that time is a software thing, which you allude to by claiming that a #define is the solution, is foolhardy - no matter what the imaginary, easy to change #define sets, time on PCs are controlled by clocks in the _hardware_, not in software (for a wide variety of reasons, such as the fact that PCs aren't real-time deterministic, and higher priority IRQs can come in and pre-empt a software clock routine -- software clocks quickly start skewing out of reality -- a #define might give you the illusion that you have super accurate time, but it most likely is not).

So why does the time in Windows have a 55ms click? Because the legacy hardware system clock has a 18.2 hz clock that clicks every 55ms -- it has nothing to do with Windows. In more modern PCs there is also a high performance clock (again it's on hardware -- the same hardware queried by whatever OS is installed), queried by QueryPerformanceCounter and QueryPerformanceFrequency -- baselining these to a real time is hardly rocket science.

Dennis Forbes
Monday, December 08, 2003

(and if that bit of reality isn't enough, you can also grab yourself the DDK and make yourself a cleanly delineated, black boxed driver to communicate with hardware, or as its own software interrupt handler, to handle whatever situation you want)

Dennis Forbes
Monday, December 08, 2003

"...and thus increasing the competition between programmers (or programming firms), and so decreasing programmer salaries!"

Competition? That's immoral. Next you'll be trying to tell us that communism isn't the best possible economic system.

"And the crazy thing is, somebody convinced programmers that this is best for them! :) "

You think that's crazy? Take a look at the free-market capitalists and libertarians - they won't stop going on about how competition is the best possible thing for absolutely every aspect of any business. They actually oppose having government legislation or volunteer programs that reduce competition. Instead they want all programmers to selflessly cooperate in a cartel that will ensure no programmer ever competes against another programmer.

Um, er, oh dear. Oddly enough these same people are whining that open source is destroying their livelyhood and they really want everyone else to be nice to them. Ah well, I'm sure someone will eventually explain why free-market capitalists are so opposed to competition when they realise that they're not on the winning side.

at least this isn't slashdot
Monday, December 08, 2003

Writing OSS software is a form of competition similar to "dumping": a company has a lot of money, and sells products bellow the cost needed to make the products, in order to make the competition bleed.

In open source, the same thing happens: some folks have a lot of spare time, and are sustained financially by their parents, and so they write lots of code for free.

There was a recent survey by the OSS community itself, which showed that most contributions to open source software is made by the students in West Europe.

This explains a lot, because in West Europe:

- university is free, so the student doesn't have to pay a loan after finishing the university

- the social security system is very good and you simply can't become homeless, even if you don't have a job

So, the students in West Europe have a lot less financial worries than students in other countries.

Does this give them the right to hurt other programmers who don't have the same living conditions?

John
Monday, December 08, 2003

Also, about the fact that competition is good.

Yes, competition is good.


But there are lots of degrees of competition, between zero competition and lots of cut-throat competition.

This is NOT a black and white issue - there are a lot of shades of gray.


Try coming up with that argument about competition to your nearby medical doctor or lawyer.

Tell them that, in order to have total competition, they should remove the barriers to entry for their professions.

See how they react - please.

And if the doctors and lawyers don't like these ideas and fight them, why do programmers embrace them?

I tell you what the end result will be: medicine and law will continue to be highly paid professions, while programming will be a low paid profession.

And why? Because total, cut-throat competition is sooo good!!!

John
Tuesday, December 09, 2003

Ori,

Regarding your synchronization example, we have a network of Macs running OS X networked here in the studio and they have 0.0104 ms (0.0000104 s) synchronization accuracy. I am wondering if you know if I can get this accuracy on Linux because if I could it might be worth it to switch.

Tony Chang
Tuesday, December 09, 2003

Looks like slashdot

> Competition? That's immoral. Next you'll be trying to tell us that communism isn't the best possible economic system.

Has it occurred to you that we who are examining the realities of open source are doing just that - we're engaging in competition against those who seek to have us work for free.

We're saying, no, I'm going to charge you for that, just as you charge me to do my web site, to write my contracts, to build my house and to sell me food.

analyst
Tuesday, December 09, 2003

JasonB, just about everyone in the world charges for their time and work. It's sort of fundamental because to survive you have to pay for other people's work and produce.

Then along comes this idea that says: Hey, it's noble for programmers to contribute their code for free, to work for free. Programmers get their kicks out of it. Wow. Of course we have to interject that we're not really saying that; you can still charge for your free software. But actually it IS what we're saying.

Now, you've just a few us here saying, hold on, this doesn't sound like such a great idea, us being programmers and all.

analyst
Tuesday, December 09, 2003

umm...actually....no one is asking you to work for free.  Im not.  Ori isn't.  No one here has even suggested it.
Most of the active/important OSS developers are working for money, not for free.

...You closed source nutters seem to be skipping lightly over that particular fact in this thread rather often...

FullNameRequired (Im a closed source developer and owner of my own business, but frankly you lot are embarrassing to read)
Tuesday, December 09, 2003

I think it's pretty harsh to tell a hobbyist who is doing something they enjoy, and is then generous enough to offer what they've worked on away to people at no charge, that they're being a bunch of pricks. Which is pretty much what some people in this thread are doing.

Yes, it may drive the salaries of programmers down. Personally, I don't think it will do a great deal, because it basically only impacts on the shrink wrap market, which is a very small section of the programming workforce. It may have some effect, but I can't see it devastating the whole industry.

But even if it did, so what? Why does anyone have the right to get paid to program, just because they enjoy it? I enjoy getting shifaced and watching the football on TV, but I'm not going to demand to be paid to do that. I'll get paid if people think it's worthwhile.

It's much like the P2P debate, where some people say that musicians deserve to be paid for what they do. What makes the musician so special that they deserve to be paid?

The only answer is that they provide something society finds valuable, so we want to encourage them to make more music.

Taking the same approach to software, we should pay programmers in order to have them further society by making more software. However, if open source does win out, this seems wasted, as it will have been shown you don't need to pay people to write software.

Copyright is a social contract - society gives someone money to encourage further works. If the money is not required to encourage people to produce further works, I think it's rather wasteful to require it. Instead let people be paid to do something else.

Sum Dum Gai
Tuesday, December 09, 2003

I never claimed that changing the C code to a product was better and easier than using published extension APIs. I said that having the source was more powerful. That is, that the complete source code is inherently more extensible.

Your contest is foolish, Tony. And it completely misses the mark. The reason your contest works is because OpenOffice is not the same as Word. I argue that the source is inherently more powerful, by definition.

Consider this contest: you have Word, closed, and you are to write your 10 extensions. I have Word, and the source to it. Which programmer has more resources available? You do?

I would like to reiterate that I have never claimed that it was easier to write a Word extension in C in the executable than with VBA. What I have argued (and I stand by my statement that this is so basic that it needn't be argued; it should be accepted without question) is that having the source is always, each and every time, more powerful than not having the source.

As I reread your original post, I realize where I was unclear: it seems to me that you argue that the goals for Free Software (development as a service, not a product, and that software should be modifiable and extensible) are better met by Microsoft products, such as Word, than a given OSS product.

I could suggest that you are intentionally mangling the intent of the statements; if I have to explain why, say so. I don't feel like writing that right now.

Instead I'll say this: we are talking about open versus closed, two ideologies. We are not talking about specific products. That is, I am not arguing the power of OpenOffice with source against Word without source, I am arguing the power of having the source versus not having source in general terms.

I'd like to repeat that: I am not comparing the relative usefulness and extensibility of a specific product, because we are discussing the merits of a methodology. It would be more fair to compare, say, Word with source to Word without source. Which is more extensible then?

I hope I've been clear this time around. I think I ought to apologize for the rather abrasive tone of my previous postings, having re-read them now.

Mike Swieton
Tuesday, December 09, 2003

> umm...actually....no one is asking you to work for
> free.  Im not.  Ori isn't.  No one here has even
> suggested it. Most of the active/important OSS
> developers are working for money, not for free.

In fact, this is not the case.

Yes, a few very high profile OSS developers for money on OSS products: Linus, Alan Cox, etc.


It's like in the rock music world: AC/\/DC and Metallica make a killing, make lots of money... but they are maybe 0.5% of the people who produce rock music.

The other 99.5% barely scrape by, and struggle to earn their living.


Just take a look at the income (not profit, but income) of Red Hat, Mandrake, and other companies that are Linux-related.

They have an income much lower than a lot of closed source software companies, therefore afford to pay fewer programmers.

John
Tuesday, December 09, 2003

On the other hand, have a look at the income of the biggest Linux company in the world: IBM.

The RedHats of this world sit in a very small part of the market for programmers anyway.

One thing I will say is that if Open Source destroys the profitability of that section of the market, it will destroy the ability to get rich off software.

You'll still be able to make money as a programmer, but the money you make from consulting is a good living but not mega rich. To get mega rich you have to convince thousands or millions of people to pay for your software, not a single client.

Sum Dum Gai
Tuesday, December 09, 2003

"In fact, this is not the case.

Yes, a few very high profile OSS developers for money on OSS products: Linus, Alan Cox, etc."

in fact it is. Im sorry John but you are just plain wrong.

Look at the main developers of php, linux, apache, firebird (database), firebird (web browser), wxWindows, and on and on and on.

_all_ the projects that are actually important in this world of ours are worked on maybe 90-95% by developers that are actually paid for their time by various companies.

there _are_ of course thousands on thousands of projects lined up on sourceforge, but Im fairly happy that most of those are pretty much not going to have a noticeable affect on the world around us.  (although in any poll one can guarantee that the the majority of developers who declare themselves to be OSS developers are going to be working on on all those wee things are going to be the dirt poor developers you have in mind...OTOH they do it for a hobby, so more power to them.)

"They have an income much lower than a lot of closed source software companies, therefore afford to pay fewer programmers."

Im _really_ interested in exactly which companies you are comparing redhat to.....perhaps you would care to expand on this point?

FullNameRequired (Im a closed source developer and owner of my own business, but frankly you lot are embarrassing to read)
Tuesday, December 09, 2003

"It seems to me that you argue that the goals for Free Software (development as a service, not a product, and that software should be modifiable and extensible) are better met by Microsoft products, such as Word, than a given OSS product."

This is correct. This is what I am saying. The 'movement' says that it is not about getting code without paying the programmer - it is all about these noble goals from the free software manifesto about its benefit to the users. I have pointed out that these stated goals are not met by 'free' software as well as they are met by microsoft products. I am not evaluating the noble goals themselves, I am merely pointing out the microsoft meets the goals a lot better than 'free' software does, thus removing from the discussion the red herring about these noble goals.

That leaves us to wonder what the advantages of free software are for developers and users.

I do understand very very well how free software benefits free software mega-advocates. it has worked out very very well for them - they have become multi-millionaires by convincing others to work for free. And I see the competitive advantages of free software made by IBM and Sun in an effort as corporations to undermine microsoft's profitability by product dumping (free products developed and distributed by a corporation with the primary intent of undermining a competitior certainly meet anyone's definition of dumping - Sun and IBM would NOT be giving away free software for hardware they don't make if Microsoft wasn't in the picture with an OS and office apps with such a large market share).

On the absurd issue that 'no one is working for free', it has been well documented in surveys that most open source developers are european students who are not paid. IBM and Sun open source coders are a very small proportion of the scene. Very small. Miniscule. So you see, I have shot that assertion of yours that open sources are paid all to hell as well. Leaving you with nothing.

Tony Chang.
Tuesday, December 09, 2003

John wrote:

---
There was a recent survey by the OSS community itself, which showed that most contributions to open source software is made by the students in West Europe.

This explains a lot, because in West Europe:

- university is free, so the student doesn't have to pay a loan after finishing the university

- the social security system is very good and you simply can't become homeless, even if you don't have a job

So, the students in West Europe have a lot less financial worries than students in other countries.

Does this give them the right to hurt other programmers who don't have the same living conditions?
---

John, I do not knwo where you live, but I live in Western Europe. I can assure you that universities are not free (although not as expensive as in the USA).

You seem to be convinced that the socal structure in Europe gives OSS producing students an unfair advantage. My response: if you can't stand the heat, stay out of the kitchen. The whole point is that conditions all over the world are different. American society is based on the concept of competition, Europe less so. This might be an advantage to European OSS programmers, but it comes at a price. I would love to pay income tax at American rates, I like the social security I enjoy over here. I can not have them both at the same time.

OSS is becoming a force to be reckoned with, espacially now that Microsoft has shown its intentions with licensing programs such as Licensing 6.0. That has been a wake up call to a lot of governments and companies: being dependent on a single software vendor is a bad thing. Linux is fully in view now and now that users have seen that it is possible to have excellent software at very low prices, exaclty tailored to your need if necessary.

Jeroen
Tuesday, December 09, 2003

"I have pointed out that these stated goals are not met by 'free' software as well as they are met by microsoft products."

LOL.

bugger me tony, thats the stupidest comment Ive seen for a while.  So you are _seriously_ arguing that closed source software allows the user to make as many copies of the product as open source software?  that closed source software allws me, as a user, to change the way it runs if I have the desire?
you been smoking something good?

"That leaves us to wonder what the advantages of free software are for developers and users."

speak for yourself, I have a very clear idea of what the advantages and disadvantages of free software are.  Mostly I dont use it, sometimes I decide its better for my clients if I do.


"On the absurd issue that 'no one is working for free', it has been well documented in surveys that most open source developers are european students who are not paid."


yada yada yada.  Sourceforge is full of projects started by students from all over, and there are other places to that stack them up.  Every computer science student in the world has prolly played with an open source project at some time, and therefore will prolly declare themselves to have "worked on open source software".

none of those projects will change the world.  <g> you could argue that the only affect those students have had is to stuff up the polls.


" So you see, I have shot that assertion of yours that open sources are paid all to hell as well. Leaving you with nothing."

yah, you surely showed us.  Proved beyond reasonable doubt that everything to do with opensource software is bad and wrong.
I dont know about anyone else, but youve nailed it for me...Im going to start my own closed source business now and..oh, wait..

Im sorry chang old boy, you've proven nothing except that you have the same grasp on the realities of the computer world as a chimpanzee has on the intricacies of modern physics.

Go smoke some more drugs :)

FullNameRequired (Im a closed source developer and owner of my own business, but frankly you lot are embarrassing to read)
Tuesday, December 09, 2003

Tony, how does your OS/X timing answer my question?

Probably a troll on your side, but In my network, the Linux servers are syncrhonized to within 0.05ms, which is sufficient for me. I haven't looked at it recently, but I could probably reduce that to 0.01ms with a kernel recompile and a switch upgrade. I know why my project needs this; why do you?

Dennis, you really should research this matter before responding authoritatively. The Windows time tick is NOT 55ms - it's closer to 1ms, and it's programmable; Back in 1988, I was getting precise 11Khz timing  (that's ~0.091ms) from a plain old 4.77Mhz PCg. It didn't do much _except_ giving me that precise time, but getting, e.g., 1ms precision and still getting things working was a no brainer if you knew your stuff.

The problem I have is with Windows' time-of-day keeping. Despite being advertised in the API as having 100ns resolution (more than enough for me), it has a ~10ms resolution, independent of the hardware. I have access to a 100ns resolution timer - it's called the High Performance Counter. It's not really reliable (google up if you want to know why), but I can deal with these problems. Unfortunately, it is independent of all other Windows time keeping functions. Which means that if I use it, I have to write my own SNTP client.

The Windows built-in SNTP client can't synchronize better than the internal clock freq (10ms), and - despite having been reported to MS in the past, is still not listed as dependent on the Network service, which more often than not causes it to NOT synchronize in the first 7 hours or so after a boot.

Writing a clock driver? PUHLIZ.

The problem with "well specified APIs" is that they can't cover everything. As in my case. I don't expect anyone to expose the clock resolution through an API, as there are not many users who'd find that useful. But as a result, those that DO need it are left out in the cold.

Ori Berger
Tuesday, December 09, 2003

"bugger me tony, thats the stupidest comment Ive seen for a while.  So you are _seriously_ arguing that closed source software allows the user to make as many copies of the product as open source software?"

That's certainly true -- under US copyright law, you have a right to make as many copies as you wish for your own personal use. You do not have the right to make copies for others though, as that is a violatoin of copyright law. Are you saying that being able to give the software for free to others and not pay hte developer for those copies is the key part of the manifesto that must not be subverted? If you are, I agree with you actually, and have even stated that this seems to be the one aspect of the manifesto that is credible. Thank you for proving my point.

>  that closed source software allws me, as a user, to change the way it runs if I have the desire?

Sure, if you buy quality commercial products, your ability to do this as a USER is quite good. Most of the best commercial products have very robust extensibility capabilities built in. Shoot, I can even write C or C++ code and make it part of Word and Excel. You do realize that is possible and something people do?

But as a competitor seeking free code to sell to others without having to pay for what you are benefitning financially from by integrating it into your own closed source commercial products? Well, I agree that that is the one part of the manifesto that seems to be the sticking point, the one point left after all the specious arguments have been discarded - you want something for free. You want to profit from another's work without paying them. That I believe.

"you been smoking something good?"

I do not smoke. Do you?

"speak for yourself, I have a very clear idea of what the advantages and disadvantages of free software are.  Mostly I dont use it, sometimes I decide its better for my clients if I do."

As I said, I really do see why you like getting ahold of free code without having to pay for it and I see hdow that is profitable for you. Your motivations you have laid out quite clearly.

"yada yada yada.  Sourceforge is full of projects started by students from all over, and there are other places to that stack them up.  Every computer science student in the world has prolly played with an open source project at some time, and therefore will prolly declare themselves to have "worked on open source software".

I am glad to see that you have changed your way of thinking and now acknowledge that most of the people writing the code that is benefiting you are not paid.

"yah, you surely showed us.  Proved beyond reasonable doubt that everything to do with opensource software is bad and wrong."

I do see the benefits of not paying for free software to you and other advocates, but I do not see benefits for the unpaid people who are working on it. And as I said, if you want extensible software, microsoft and others create many fine products. If you believe software should be sold only as a service and not as a product, then there is no doubt that you are very comfortable with Microsoft's new licensing rules which charge for software in precisely that way.

"I dont know about anyone else, but youve nailed it for me...Im going to start my own closed source business now and..oh, wait.."

Yes, this is the business in which you profit off of the work of others without paying them. I do understand that dear sir, you have made your interests crystal clear.

"Im sorry chang old boy, you've proven nothing except that you have the same grasp on the realities of the computer world as a chimpanzee has on the intricacies of modern physics."

Yes, the old ad hominem - the tell tale sign of someone who has nothing to say.

"Go smoke some more drugs :) "

I don't do drugs nor do I smoke. Are these things of especial interest to you?

Tony Chang
Tuesday, December 09, 2003

"under US copyright law, you have a right to make as many copies as you wish for your own personal use."

:) Can I use them all at the same time if I wish?  Can I give copies to my friends and let them use those copies if I wish?  I can with OSS.

"You do not have the right to make copies for others though, as that is a violatoin of copyright law"

unless the copyright license that the product is distributed under allows this of course.


"Are you saying that being able to give the software for free to others and not pay hte developer for those copies is the key part of the manifesto that must not be subverted?"

well...its not exactly a manifest, its a license.  _all_ parts of the license any product is distributed under must be adhered to, of course.
We _were_ talking about OSS, not it appears we are talking about the licenses that are used in that area, there are a whole bunch of those of course, but Ill assume that you are specifically interested in the GPL.
So indeed, the part of the GPL that specifically allows copies to be made, distributed and used by others without payment back to the original developer _is_ as important as any other part of that license.

"If you are, I agree with you actually, and have even stated that this seems to be the one aspect of the manifesto that is credible. Thank you for proving my point."

you are welcome if I did :)  The _important_ point is that _all_ parts of the GPL must be adhered to of course.

"Sure, if you buy quality commercial products, your ability to do this as a USER is quite good."

cool.  Im running IE5 on the mac at the moment, unfortunately its crashing when it tries to load a particular java applet.  How can I fix this?



"But as a competitor seeking free code to sell to others without having to pay for what you are benefitning financially from by integrating it into your own closed source commercial products?"

?  are you?  I wouldn't admit to that if I were you.
Technically its illegal, you are knowingly failing to adhere to the GPL license.

"I am glad to see that you have changed your way of thinking and now acknowledge that most of the people writing the code that is benefiting you are not paid."

That is benefiting me?  well, no...I use apache, MySQL and Firebird (the database)  pretty much hall the important developers on those projects are paid for their work.


"but I do not see benefits for the unpaid people who are working on it."

The people who choose to work on OSS for free by their own choice?

I kind of do, I suspect they enjoy developing as much as I do and feel like sharing their work.
Thats a reasonable benefit.


"If you believe software should be sold only as a service and not as a product"

I definitely do not, my business depends on selling software as a product.

"then there is no doubt that you are very comfortable with Microsoft's new licensing rules which charge for software in precisely that way."

yep :)  honestly, MS can do whatever they want with their licensing.  It _is_ their software after all.

"Yes, this is the business in which you profit off of the work of others without paying them."

Right, thats the one.

"I don't do drugs nor do I smoke. Are these things of especial interest to you?"

They _really_ are.  Its the _only_ explanation I can come up with for your apparent stupidity and deliberate misunderstanding of the postings made in this thread.

But its OK, really....you can play all the games you like with words..

FullNameRequired (Im a closed source developer and owner of my own business, but frankly you lot are embarrassing to read)
Tuesday, December 09, 2003

"why do you?"

Ori, the studio has a dozen 1.25GHz dual G4s networked for music audio. They are synchronized to sample-accuracy at 96kHz for real time DSP work, and give the same performance you'd get with a single system running at 30GHz, actually better performance than a single since you have so many different pipes and busses.

I'm a Windows guy myself -- the studio belongs to a friend in my group. The system is expensive though and if we could switch over to Linux we'd save quite a bundle I imagine, based on what the linux guys are telling me. I'm not sure of the availability of pro level studio recording software, or drivers for the various audio hardware though, but a lot of people tell me they are using Linux for pro level studio work, it's just the details that are a bit hazy. Haven't actually seen one of these studios. Would really like to see how they're set up.

Tony Chang
Tuesday, December 09, 2003

"How can I fix this?

Are you near seattle? I can come over this weekend and personally help you with your technical problems. My rate is $90/hr with a minimum of 2 hours plus 30 cents a mile.

Tony Chang
Tuesday, December 09, 2003

"I suspect they enjoy developing as much as I do and feel like sharing their work."

Excellent, please point us to a site where we may download the fine source code that you develop and enjoy sharing.

Tony Chang
Tuesday, December 09, 2003

""How can I fix this?

Are you near seattle? I can come over this weekend and personally help you with your technical problems. My rate is $90/hr with a minimum of 2 hours plus 30 cents a mile."

closer to new zealand actually.  <g> otherwise I admit I would be _extremely_ tempted to see how you solved it.

"Excellent, please point us to a site where we may download the fine source code that you develop and enjoy sharing."

LOL.

Its on here:  www.sourceforge.org

Do a find on "wow, Im obtuse" :)

FullNameRequired (Im a closed source developer and owner of my own business, but frankly you lot are embarrassing to read)
Tuesday, December 09, 2003

Wow! 72,540 projects! you sure keep busy. Unfortunately, 69,800 of them are extremely low quality. Is that your earlier work?

Tony Chang
Tuesday, December 09, 2003

I think you can get the same kind of sync accuracy out of windows if you want. If you go down the directX route (See IReferenceClock in the SDK) .
I do remember that Studer claimed to have gotten this working with NT4 but I can't remember how they did it (plus I think they were only aiming SMTPE timecode accuracy).

However if you're using word clock accuracy then it depends on how you're distributing it, I'd assume it's distributed seperately using AES3 (Which is what DirectX on windows can do) 'cos I couldn't get it work over regular ethernet and FireWire wasn't accurate enough when I looked at it, but I may have misread the FireWire spec as it was relatively bandwidth limited (only three times more than MADI at that time) so I didn't do huge amounts of work.

Peter Ibbotson
Tuesday, December 09, 2003

"Dennis, you really should research this matter before responding authoritatively. The Windows time tick is NOT 55ms - it's closer to 1ms, and it's programmable; Back in 1988, I was getting precise 11Khz timing  (that's ~0.091ms) from a plain old 4.77Mhz PCg. It didn't do much _except_ giving me that precise time, but getting, e.g., 1ms precision and still getting things working was a no brainer if you knew your stuff."

I apologize it has been a while since I've used the low resolution timer for anything on consequence -- indeed it is ~100 clicks per second in Windows 2000 and above (or 10ms). Actually I just did a test on my PC and it's actually 15ms.

"The Windows built-in SNTP client can't synchronize better than the internal clock freq (10ms), and - despite having been reported to MS in the past, is still not listed as dependent on the Network service, which more often than not causes it to NOT synchronize in the first 7 hours or so after a boot."

That (the dependency) is a single registry string value (DependOnService of say "lanmanserver" under HKLM\System\CurrentControlSet\W32Time). It's an error if they don't have it by default, but it's obviously not a big problem if you can go in and set it.

The fundamental disagreement that we have is that you seem to be confusing resolution with accuracy -- The windows time functions have a resolution down to 100ns, but all of the documentation I've seen specifically states that the _actual_ accuracy can be dramatically different. While non-realtime operating systems might present the imaginary resolution of a super-duper high resolution clock, and maybe it even clicks up in some mildly accurate manner, the accuracy of the clock is a whole different matter --- without going to a hard-realtime system, or explicitly and always using a hardware clock, you're giving yourself the illusion of accurate time.

Dennis Forbes
Tuesday, December 09, 2003

FNR,

getting the source to the software has absolutely no effect whatsoever on your rights to install and run it any more than getting the binaries to it does.
See all the RHEL licencing explanations that have been discussed on this board for reference.

Just me (Sir to you)
Tuesday, December 09, 2003

I've just read through this whole discussion. Yes! The _whole_discussion!

*Disclaimer: I am a closed source developer.*

Why is it assumed that were I to be given a _choice_ to pay or not, I will choose _not_? I do not see how paying for something, that is worth it of course, is incompatible with having a choice of not paying for it as well.

Last word?
Tuesday, December 09, 2003

If you had the option of paying or not-paying, and getting the same product with all of the same legalities and ownership rights, you would choose to pay? Either you're an extreme minority, or you're a bald-faced liar.

Interest story from this (and totally off-topic with this thread, which is okay as it's time for it to die): In the retailing/manufacturing industry there is a rule called the "30:3" rule relating to 'morally superior' goods. The basics of the rule is that of a given random sampling, about 30% of the people will assert that they buy products with a main criteria being "social conscience"- they pay more for environmentally sound products made under good labour conditions, etc.

When they actually monitored randomly sampled purchases at retailers where there is a clear demarcation between the products (with one clearly being socially conscious, albeit at a premium, and one not), and apparently using some methodology that assured some sort of correlation between their survey and actual purchases, only 3% bought the socially considerate product. What does this prove? Basically that a lot of people are liars, and while people might recognize that something is right, they'd rather that everyone else shoulders a bill. In fact said liars will often publicly proclaim their support of such products, and how willing they are to support it, as a sort of replacement for actually paying their share. You can see this evidenced on Slashdot all the time when ideological things like "tip jars" come up -- Music artists should just release their music for free and put up tip jars! They'll make tonnes of money, at least if all of the public proclamations of support for tip jars is accurate.

Dennis Forbes
Tuesday, December 09, 2003

Wow, John's upset that a load of Western European students, whose studies are subsidized, choose to give something back to society.

Whatever will they think of next! Helping out at soup kitchens?

Stephen Jones
Tuesday, December 09, 2003

I must be in the minority, 'cause liar or not, I most certainly am not bald!

Seriously, it is true that I'll _pay_, if I feel its worth it. If not for anything else but the sheer ego trip of having _owned_ it. Call it weird or stupid, nothing like a receipt to claim ownership.

Last word? Of course not!
Tuesday, December 09, 2003

I find this whole technical OSS/CSS distinction seriously overblown.
Think about it. If MS would give you the code to their product tomorrow, what would really have changed?
Maybe a tiny tiny franction of the customers would suggest bugfixes instead of now merely reporting bugs, and certainly the blackhats would have a field day and the IP infringers would have a field day, but everything else would be exactly the same as yesterday.

The difference between OSS and CSS as they are often projected are not about source code at all. It is about business models, but I believe even there the distinction is not all that its reported to be.
Yes, most OSS business models are either in custom software for verticals (who cares you can see the source, you don't have a use for it anyway) or commodities (who cares if I give the soft for free, I'll sell you more xxx because of it). But OSS has its normal "the software really is our business and you pay for each seat" businessess too (e.g. RedHat). CSS has the full range of business models too, but maybe there the "shrinkwrap" model of COTS is more prominent as are wide horizontal market packages.

What gets on my nerves:
- Zealots: OSS has many but CSS also (Mac, Amiga, OS/2 anyone?)
- Academics figting the ABM cause from their taxpayer funded cushion seats
- Governements that can't even distinguish a spreadsheet from a compiler thinking they should promote OSS since it is more "leftish" or "provides more opportunities for their local software sector (which curiously when asked wants them to especially butt out of medeling with the sector).

Just me (Sir to you)
Tuesday, December 09, 2003

Oh and one more thing: The "free" software as in the we'll make our money on compliments businessess is bad news for programmers, since you have just signed up for working in a pure cost environment.

Just me (Sir to you)
Tuesday, December 09, 2003

Yeah baby this is #99...who'll take #100?

Dennis Forbes
Tuesday, December 09, 2003

Eric Raymond, Richard Stallman and Michael Moore must be close friends. Preaching water while drinking wine.

Slashbot
Tuesday, December 09, 2003

Dennis:

I may have used the terms accuracy and resolution interchangably in this discussion, me bad - but I don't think they are well defined in this context.

The clock deltas can only be multiples of 15.4ms on my machine. The fact that this multiple is represented as ~154000 units of 100ns does not really increase the usefulness.  It's definitely low accuracy - and in my book, the precision with which the quanta is specified doesn't make it high resolution.

About synchronization precision - I'm building a "soft real time" system. Delays and inaccuracies are acceptable and accounted for, as long as they (statistically) occur rarely. I'm using a 2Ghz Pentium, with CPU usage barely exceeding 1% over 99% of the time. The synchronized clocks are NOT an illusion - the PCs undistrubed for days, and the NTP client/server I use actually works well. If I stop the NTP service, the clocks diverge less than one second per month. (the relative drift was correctly computed, and it is taken into account even without active synchronization).

No, it's not a hard real time system -- I'm not deluding myself. But for my purposes, clock accuracy of 1ms would have done wonders to simplify things -- and it does, when I run what parts I can on Linux. Unfortunately, Windows won't let me have it.

(and ... IRefenceClock is about as easy as writing my own SNTP client and maintaining my own clock... But thanks)

Ori Berger
Tuesday, December 09, 2003

IIRC, back to the original post, ESR actually lost out big time on the stock options.

He got the options.  After the IPO, he was an idiot and bought his options while the stock value was sky high.  He then had to pay taxes on those options as income.  He did *not* sell before the stock value of VA Linux plumetted to barely-listable.

He may be OK now, but the VA stock certainly didn't make him a gazillionaire.

Richard P
Tuesday, December 09, 2003

Actually, I'm really enamoured with Tony C.'s approach:

I will not work for free. 
Don't ask me to work for free.
MF SOB, quit asking me to work for free.
godammit, I told you I won't work for free.
You scum sucking low life bastard, I will not work for free.
Eat shit and die free software.

Hey, people... quit asking Tony to work for free.

hoser
Wednesday, December 10, 2003

Well, Tony has a very valid point about this:

Let's say that you write a program, and you have a question about how to handle selling the program, or piracy, or software protection, etc.

If you talk about this to an OSS programmer, he will tell you to "just open-source the program and get over it".

Many times he will also insult you or say that you are evil because you write commercial software!

I don't know, but in my book, this is not an ok attitude.

Orson
Wednesday, December 10, 2003

Please learn to differentiate between an OSS programmer that actually produces usable code and a slashdot zealot.

The advice you'll get from wannabes and professionals in _any_ field will diverge.

Ori Berger
Wednesday, December 10, 2003

"I don't know, but in my book, this is not an ok attitude."

Fortunately all closed source advocates are caring sensitive and incredibly polite people, thus proving the natural superiority of closed source licenses.

Er. Yeah, anything you say.

Actually, if you ask Mr Chang a technical question, he'll very politely tell you that you're an evil person unless you pay a large amount of money for the valuable advice you want. (For evidence, look at some of his wonderful posts here.)

But that's so much better than being told that you're evil for not sharing the source, I guess.

at least this isn't slashdot
Wednesday, December 10, 2003

1. If someone asks you to Open Source your code and you don't want to, then *don't do it*. Ain't no one making you do it.

If some high-falutin' gurus are telling you how practical (ESR) or moral (RMS) it is to do it, well, you're a grown up who can decide for yourself which way to play it. Ain't no one making you do it.

It never hurts to have options, and so you should at least think, "What benefits would I get if I Open Sourced the code?". If it's not convincing... then *don't do it*. Ain't no one making you do it.

2. If other people are Open Sourcing their code, for their own reasons.... ain't none of yo' business!

Mr Obnoxious Man
Wednesday, December 10, 2003

You see, here's the rub Mr. Obnoxious Man -- OSS fanatics (particularly the "free" software kind) invade every newsgroup, forum, public discussion, and corporate conversation, blathering incessantly about the technical superiority and moral righteousness of OSS, and the inherent evils of "closed source", or copyright/restricted rights (although they always cleverly ignore the heinous restrictions that the GPL imposes, but let's pretend that the paradox isn't there). When called on it, invariably they fall to your line, which we've seen several times - "Don't worry about it! No one is forcing you! It's none of your business!".

The moral of this is that if you can't eat it, don't serve it.

Dennis Forbes
Wednesday, December 10, 2003

" OSS fanatics (particularly the "free" software kind) invade every newsgroup,"

umm...except this one of curse.

At least the last two threads devoted to OSS, this one and the one before it, were begun by CSS fanatics attacking open source in a self righteours etc etc etc fashion.

I dont care enough to go back through the archives and check the ratio, but IIRC the majority of other threads on this topic were started in a similar fashion.

FullNameRequired
Wednesday, December 10, 2003

*  Recent Topics

*  Fog Creek Home