Richard Stallman is responsible for the shrinking economy

Quite a lot is available

The same is true of the vast majority of commercial applications where FOSS is used

I think this is a good essay question for your midterm.

-rick-

Op Fri, 10 Apr 2009 02:57:03 +0200 schreef Jacko :

If you don't have time for children, then don't get them!

|> Toshiba and Microchip give away their toolchains to make their chips |> more popular.

Well, this is a case where my personal opinion is that the software *always* should be given away for free.

If $VENDOR wants me to use their chips, why should I pay extra money for the development software? If internal bookkeeping wants that software paid for, then cross-finance it via the sales price of the chips.

Rainer

I agree, and I think more hardware manufacturers now support this, in a few cases directly financing improvements to the free tools.

It's worse: some of them actually *keep the specifications secret* so as to support the proprietary tool vendors. Luckily the situation seems to be improving, for example the openocd project can debug a vast range of ARM chips, using very low cost JTAG hardware.

Chris H a écrit :

I understand that you want the discussion which have no common sense, a sort of controversy.

Habib

In message , Rainer Buchty writes

Why? Is software valueless? Should your SW be given away free?

ALL MS products should be free when you buy the computer?

Then you get tools that are just good enough to give away free to get the development started. You loose all the high quality tools and you get the perception that Sw is valueless.

Also you only get the tools they want to supply. If their tools are not very good you are stuck with them. The accountants will not let you use a more expensive chip just to get better tools.

If you are a SW developer your software is valueless and should be free... and by implication you are valueless...

Talk about Turkeys voting for Christmas.

In message , Habib Bouaziz-Viallet writes

No I simply asked for some sort of clarification and qualification of the statement

What benefits?

Chris H wrote: [...]

For example, the benefit of being able to understand problems, if not fix them personally, instead of easter-egging for days or weeks why the compiler does not like its dongle today, or why the debugger bluescreens after a 20 meg download.

After weeks of try-and-error, we have finally found a combination that works most of the time. We still don't know why it works, and why it still fails once in a while, but a reboot cures it most of the time. I'm quite certain that if the thing were open source, we had been able to understand and _fix_ the problem during all that time.

Another benefit of open source is you can usually start evaluating the thing right now, without having to deal with NDAs and crap.

Oh, and by "open source" I don't mean "proprietary with source". That's useful as well, but usually comes with other strings attached. It's not _that_ useful to know where the problem is if you don't have the legal permission to fix it (resp. to deliver the fixed version). My favourite example still exists: a driver marshalling an 'unsigned long' from four 'char's, not 'unsigned char's. The supplier still delivers the wrong version because he refuses to fix it without a bullet-proof standalone testcase.

Stefan

Partially, Microsoft tries to achieve that illusion. When you buy a computer, you get Windows "for free".

You want to imply that high-price software is better than low-price or free software. This may be the case for some things, but it isn't for others. Is there a commercial equivalent to Emacs? Valgrind?

Depends on how you talk to the accountants. For example, one probably can fit an MP3 decoder onto a 20 MHz ARM chip, by doing it in 99% assembler and sacrificing some precision, but it will be much easier, and your product ready much earlier, if you can use a 100 MHz chip and a compiler. Likewise, you can argue that the more expensive chip comes with better tools that let you finish the product earlier, leave room for expansion, etc.

The accountant can then figure out whether it's better to use the more expensive chip, or even use the cheaper chip and buy additional tools.

When you go by first-class train, you get a free newspaper if you want. Does that mean newspapers are valueless?

When you go shopping, you get free plastic bags. Does that mean the chemists who make them are valueless?

Oh, and when you buy hardware in form of a TV set, you get the software for free, too.

Stefan

No, because software has more value when used to leverage your core product than when you try to sell it on its own. Companies don't sell their marketing brochures, do they? They don't charge you to visit their web site either. In this case, a company realizes that the software is a tool that they're using to sell more core product, and that their net profits are higher if they just give it away.

Car dealers don't sell gas, but they'll give you a free tank of gas with a car purchase. Why? Because it sells more cars.

This depends on the vendor. In my case, the contracts often have performance metrics in the deliverables that make the "free" tools of comparable quality to the proprietary ones. Again, it depends on what the vendor (customer) wants, and what value they see in it.

You get that with proprietary tools too. The difference is, if you're "stuck" with free (as in speech) tools, you can at least do something about it. With proprietary tools, you can't even fix the problem.

And IMHO proprietary tools can compete with free tools - look at IAR's tools, which cover basically the same chips as GCC does. For a chip vendor, more choice means more chip sales, so they encourage multiple solutions (Renesas has at least three solutions for their chip families).

I don't see how you come to this conclusion. Just because you don't understand where the value is in free software, doesn't make it valueless.

This is the underlying argument for patents. It is a good idea, IP is the exclusive use of the creator for a time than falls into the public domain. It rewards innovation but doesn't allow it to be a monopoly. It also allows the owner / creator the options including releasing it to the public domain but preventing others from controlling its use..

Regards,

-- Walter Banks Byte Craft Limited

In article , Chris H writes: |> |> Why? Is software valueless? |> Should your SW be given away free?

I think you indeed read further in my original posting as to see the point.

|> ALL MS products should be free when you buy the computer?

Personally, I would prefer if this almost implicit "Microsoft Tax" would be waived -- in fact, it can be more expensive to buy a computer *without* MS software preinstalled.

But you like to do some spin-doctoring here, because the OS and basic work software is to the computer what e.g. fuel is to the car.

|> Then you get tools that are just good enough to give away free to get |> the development started. You loose all the high quality tools and you |> get the perception that Sw is valueless.

That may be the way you see it, but I like to object.

Are video game consoles of low quality because they are heavily subsidized by the games? Are inkjet printers or, even more noticeable, mobile phones of low quality because you almost get them for free if you're just willing to pay for ink / minutes through your nose?

|> Also you only get the tools they want to supply. If their tools are not |> very good you are stuck with them. The accountants will not let you use |> a more expensive chip just to get better tools.

Doesn't that leave enough room for 3rd-party vendors?

If the free VHDL compiler and the free place&route tool of $VENDOR is crap, well, then go for an alternative product and pay. I didn't see that Mentor Graphics or Synopsys went into bancruptcy just because of Xilinx Webpack, Altera Quartus, etc.

If the GCC port for some processor is producing crappy code, well, you're free to go for a commercial alternative. Keil and IAR, for instance, are also still in business -- despite a multitude of free compilers for various CPUs.

But: As a *hardware* vendor whose products require and rely on development software, you better do a good job. Think back to AMD's MachXL3 debacle with its inferior fitters compared to MachXL2; or where these inferior fitters nearly killed the Mach 5.

And as a *hardware* vendor who make their living from *selling hardware* nothing speaks against creating a proper development suite for their specific products, making that available for free to the (potential) customer, and subsidize development costs by the hardware product.

No need to do it "cheap".

|> If you are a SW developer your software is valueless and should be |> free... and by implication you are valueless...

Not to sound harsh, but you may want to re-read and maybe even realize what I wrote, instead of falling into such completely unrelated babble.

I don't see why you believe software to be valueless or even unpaid for in the above scenario.

Rainer

Silicon company tool chains have a very interesting business model where the customers pay tools through silicon sales. This model gets base line toolchains easily into customer hands.

Many commercial tool companies successfully compete with Microchip tools.

In general silicon companies who have their own in house tools are able to create stable tools sets with unimpressive performance. It is worth looking at application oriented benchmarks run on tools provided by in house developed tool chains. Metrowerks /Freescale and Microchip for example.

Regards,

-- Walter Banks Byte Craft Limited

This describes a value curve on any innovation. Well regulated innovation will encourage more innovation.

Actually the maps are not free from Google. Google licensees a copyright for a map to attract people to look at advertisements from companies that have paid Google to display ads to a wide base of potential customers.

The maps are there to prevent the ads from sticking together and besides there and not many people who would look at the ads without something to attract them to the web site.

I would love to have the fees that Google pays for editorial material.

Regards,

-- Walter Banks Byte Craft Limited

??? That is the same with all systems

I'm not

Hang on you can usually do that with most closed source programs. And you don't get the highly restrictive GPL license.

What I'm saying is that thanks to Linux & gcc, someone with a good idea can develop a new product without seeing an investment banker first. The decreased up front investment makes it easier for new players to enter the field.

I don't know how you arrive to "but they should not have to pay others for their innovation". Of course if a company makes use of the innovations of others, they should do it in the terms dictated by the original innovator. If the original innovator has (for whichever motives) decided not to request compensation, then it is perfectly ok not to give any.

I'm sorry, I don't really understand your comment.

Let's make this more concrete. I have been working on and off on a embedded product targeting a *very* niche market. I don't know if it will ever produce appreciable income, if at all. However, thanks to avr-gcc and avr-libc, I have been able to do development on my spare time on a shoestring budget. Had I had to spend even in the $k range for development tools, the project would probably been a non-starter. I can justify spending the odd spare hour working on the stuff to myself (and the wife) much easier than I could justify spending significant amounts of money.

If anything ever comes out of the project, I won't hesitate to put a price tag on my work. And I will also not feel that I am in any way ripping off the good people who have given me avr-gcc and avr-libc.

Bravo.

w..