Is "free" always a good thing?

Drop in for a chat and a cup of tea

88 posts   Page 3 of 4   1, 2, 3, 4
by tufty » Thu May 17, 2012 6:46 pm
Mike.

You're conflating "free as in beer" with "free as in speech".

"Free" as provided by gmail (to pick an example) is great in terms of how much you actually pay, but there's 3 costs:
  1. Advertising in your inbox, assuming you use their web client
  2. Advertising everywhere as Google are mining your inbox for data
  3. Your emails are locked in to google's servers, and it's (relatively) hard to transfer everything across.
When a product is "free" in terms of cost, but not in terms of freedom, you have to start looking at where the actual costs lie.

"Free as in speech" is somewhat different. The idea being that you control your data, not some faceless corporation. There's no reason you can't make money in the "free as in speech" world, you just need to look at things differently.

For publicly funded projects such as schools, I tend to the hardline approach. Everything used to create and manipulate data should be free as in speech. That doesn't mean it should necessarily be "free as in beer", and it doesn't preclude the use of (for example) Windows as a platform.

"Free as in beer" software doesn't kill commercial software anyway. If it did, MS and Apple wouldn't exist, neither would Adobe, Oracle and so on. If your commercial software is being killed by a free alternative, well - frankly - tough - it's obviously not good enough to distinguish itself (or at least, not enough to warrant the pricetag you've put on it).

All you need is an edge. People will pay if your product is good enough. Look at Photoshop - it costs over a thousand euros, but people prefer it to, and purchase it instead of, Gimp, which is free (as in speech and beer).
Posts: 1330
Joined: Sun Sep 11, 2011 2:32 pm
by Stateside » Fri May 18, 2012 1:41 am
Python and Scratch

I do not code. What is " KURT" ?

I made a Python library for parsing Scratch format project (.sb) and sprite (.sprite) files.
You can load the files, look at their internal structure — including sprites, scripts, variables, images, etc — make changes, and save them again!
So far, it can read/write sprites and their properties, including scripts, blocks, variables, and lists. It can read costumes and export them to separate image files
It can't write images as costumes or read/write sounds yet, but they're on the to-do list — see below

http://scratch.mit.edu/forums/viewtopic.php?id=92843
https://github.com/blob8108/kurt
Posts: 67
Joined: Sun May 13, 2012 12:26 pm
by Mike Lake » Sat May 19, 2012 7:06 am
Let me give an example of how difficult it is to make commercial decisions in the world of "Open" and "Free".

We have an educational project that requires a DSI touch screen and camera.

The RPi has hardware interfaces for both and Gert has shown a camera module to be released later in the year - but will the low level software interface be documented to enable others to develop drivers/hardware?

Who is "responsible" for the RPi? Whose door do we knock on?

Will DSI/CSI drivers be provided by someone - commercially or free?

If we wish to write our own drivers, who will provide the necessary technical data to enable them to be written?

Is such data available "Open" and Free" or is it commercial? If commercial, is it a one off payment or a prohibitively expensive licence. Many licences from mobile-phone related companies/products assume the licencee is about to produce millions of products - what if we intend to ship hundreds rather then millions? Will the licence adjust accordingly or does innovation get stifled on day one?

Does release of technical data require a Non Disclosure Agreement (NDA)? We do this sort of thing every day with major companies world-wide. How far does "Open" extend with the RPi?

If we write the drivers it will cost money. Work on the basis of a Systems Programmer costing £50,000 p.a. plus 50% overheads - make it £75,000 p.a.

How long does it take to design, write, test, document and certify a new driver? I don't mean "something that will do for now", I mean something properly done and supported into the future.

Is it an evening, a man-week, one man-month, two-months or what?

Of course, if we invest in doing it we will use it for our own project and we will make it available to others for a reasonable fee.

On the other hand we could wait until, in this new world, someone comes up with a driver for free.

If no technical data is available then of course we have no choice but to wait and see.

... and that's just the drivers.

It's not simple is it?
User avatar
Posts: 96
Joined: Sun Feb 12, 2012 11:45 am
by jamesh » Sat May 19, 2012 7:50 am
If you check the home page you will see we have just saved you £75k!

For low level stuff like this, it would be exceptionally difficult to arrange for you to write the drivers yourself - main problem is the colossal amount of knowledge of the GPU required to do it, and problems with access to the right technical documentation - which is why this stuff is usually done by Broadcom or Broadcom people working in their spare time. (Gertboard, Camera board etc). So it not a good example for the point you are trying to make.

The Foundation intend to produce a DSI interface driver, it's on the list of stuff to do. It unlikely anyone else could knock one up any quicker even if we don't have anyone to work on it right now.
Raspberry Pi Engineer
Raspberry Pi Engineer
Posts: 10590
Joined: Sat Jul 30, 2011 7:41 pm
by rurwin » Sat May 19, 2012 10:03 am
Mike, you keep putting up this straw man, so I will take a pot-shot at it.

Who's door do you knock on to enhance an existing OSS project? Generally the lead-developers'. There will be a developers mailing list, forum or whatever. Post on that that you have a need for a certain enhancement. If nobody takes an interest, offer money. It may turn into a discussion about exactly what you need and the philosophically best way to integrate it with the project, but there is a good chance you will get your work done. More chance than with commercial software; you are talking to several different people with their own opinions, rather than one company with one development plan.

If the developers are not interested, then you could try polling for interest on the users' mailing lists to convince them, or you could make the change yourself. Unlike the RPi which is not OSS, all the information is available to do that.

That goes ten-times for bugs. With a commercial library you can easily spend much more time working around a bug that wont be fixed (or wont be fixed on your version) than you would have taken had you dived into the library source and fixed the bug. And even if you don't fix it, knowing exactly what it is can allow you to work around it more efficiently. As an example, see this Python bug.

Producing free-as-in-freedom software does require a lot of re-thinking of your income stream, and I wouldn't think to advise you on what can be made to work. I note that IBM, who you mentioned earlier are one of the biggest contributors to OSS to the tune of $100M, (that was a few years ago so it's probably more by now.) But they also have closed software and the two lines are synergistic, so it isn't the ringing endorsement it could be.

If you ship free software, then your user-base will be larger and you can avoid the costs associated with licensing and piracy-detecting. Even the support is simpler because there are no questions of who is licensed and who isn't and how many licenses of what type they have. Support is better because the advanced users provide support for the newbies. There are lots of teachers in my family, and I know that those support questions are phoned in on the secretary's phone during the teacher's lunch hour. Hanging on the phone for forty minutes is even more annoying when you are not getting paid for it and you are missing lunch to do it. Good, solid, well-informed support is a feature that will sell any product into the educational sector. But it will not necessarily turn a profit, because in OSS the users can and do make their own support structures.

Why did Linux succeed so well and come to rival Microsoft Windows? Because they had more programmers. Nobody can afford to employ as many people as work on a really successful OSS project. For example see the Wireshark AUTHORS file. (And that doesn't include testers etc.)

Companies making money off OSS are like razor manufacturers. They give away the razor and charge for the blades. That's the model I think you should be looking at, but be aware that it is very easy to kill the goose when you try to monetise the golden eggs; OSS must be free in all respects, it doesn't work otherwise.

And finally, I agree with others that there is no reason for you to have to give away your program free. Many companies sell software to run on Linux, Oracle and IBM for two. I don't believe that there is any culture in schools for using free software, in fact rather the opposite. In my (limited) experience, schools buy commercial solutions even when there are markedly better free alternatives. The difference seems to be marketing and a model the bean-counters understand, not cost.
User avatar
Forum Moderator
Forum Moderator
Posts: 2890
Joined: Mon Jan 09, 2012 3:16 pm
by Stateside » Sat May 19, 2012 10:29 am
Mike wrote:
Personally I would like to see the Python equivalent of a Scratch script visible on demand. That way the system can quickly adjust to the abilities of different pupils in the class. Some will stick to Scratch-type stuff longer and others will immediately leap to the Python code. Even then, the level of Python exposed will vary depending on level - some of it can get very scary, very fast and it is not as intuitive as Pythonites would have us believe.

Could Kurt be used as part of this bridge.?
Posts: 67
Joined: Sun May 13, 2012 12:26 pm
by Mike Lake » Sun May 20, 2012 1:46 pm
In relation to the camera module I think it is brilliant that Broadcom engineers are giving their free time to develop more and more stuff for the RPi.

Is the GPU so "hard" that the rest of us could not handle it? Is anyone suggesting that there will be things about the RPI that will never be fully open and therefore we will always have to rely on the good will of people inside Broadcom?

There is a commercial point here. The RPi is already a fantastic success - the back orders at RS and Farnell make that clear - and it hasn't even got started yet.

With this volume of interest, and the amount of global publicity the project has managed to get for free (worth an arm and a leg), the RPi is a FORCE.

In a commercial world one would use that force to twist arms to get what you want. For example, not pick an image sensor simply because "that's what's there" but to specify an image sensor and find someone to do a run of them. I'm like a record stuck on repeat - like many others I would prefer not just a camera module but also image sensors without IR filters that could be used for lots of other things besides pictures - and I think the RPi has the clout to get them.

I am absolutely sure that everyone working on the RPi does their absolute best to get the best possible prices for all components and manufacturing - but maybe now is the time to go one step further and to start specifying what they/you/we want.

That's what non-technical management people would be doing in a commercial setup, using their clout, and I hope that's what is happening within the RPi Foundation.

... and, what's "Kurt"? Anyone got a link about it?
User avatar
Posts: 96
Joined: Sun Feb 12, 2012 11:45 am
by khulat » Sun May 20, 2012 2:02 pm
As Gert has pointed out in the other thread and the comments that is simply not possible. They will use camera modules that have already been used together with the GPU, because it takes 6 months of fulltime work to tune the camera. So it really wouldn't help anyone to get someone to produce a special image sensor.
Posts: 104
Joined: Sun Feb 12, 2012 9:43 pm
by hippy » Sun May 20, 2012 2:08 pm
Mike Lake wrote:Is the GPU so "hard" that the rest of us could not handle it?


Seemingly so, and at 16MB+ of object code ( 'the binary blob' ), the source in some proprietary language, using in-house toolchains, and only Broadcom engineers understanding what that code does and achieves I'm prepared to believe those who say no one outside could handle it and it would largely be useless to anyone if the source and tools were available.

The binary blob is in full view, so if anyone can reverse engineer it, understand it, modify and reassemble it, prove that view wrong then they can go ahead and do it. I'm sure there will be vacancies at Broadcom and many other places for people who could demonstrate such skills.


Mike Lake wrote:Is anyone suggesting that there will be things about the RPI that will never be fully open and therefore we will always have to rely on the good will of people inside Broadcom?


Yes; and that's been the plan all along; "Don't like it? live with it or choose something else".

Of course plans may change and I'm sure both the Founation and Broadcom are amenable to philosophical, commercial and monetary pressure just as much as the next person or company is so anyone who can convince them it's worth their while is able to apply it.

That is what the Foundation did in getting Broadcom, RS, Farnell and others on-board with the R-Pi.
Posts: 757
Joined: Fri Sep 09, 2011 10:34 pm
by tufty » Sun May 20, 2012 4:17 pm
hippy wrote:
Mike Lake wrote:Is the GPU so "hard" that the rest of us could not handle it?

Seemingly so, and at 16MB+ of object code ( 'the binary blob' ), the source in some proprietary language, using in-house toolchains, and only Broadcom engineers understanding what that code does and achieves I'm prepared to believe those who say no one outside could handle it and it would largely be useless to anyone if the source and tools were available.

To put it into context, I've done quite a considerable amount of reverse engineering in my time. I have a whole stable of handbuilt tools to help me. I'm currently looking at reverse engineering the highly compressed data sent by a particular device (itself discontinued in 1999 or so, and not supported on any computers made since 1998). I have dumps of what it's throwing down the wire (pretty trivial to get, all it requires is the device itself, 2x18 year old computers, a couple of little microprocessor boards, some resistors, a bit or rather clever code that a friend of mine wrote, and not being afraid to tear apart some extremely difficult to source cables), I fully understand the wire protocol it's using - all I'm trying to get at is what data is being sent. I have a copy of the last supported driver for the device itself (circa Y2K), understand what the driver itself is doing and how to go about writing drivers for that platform, and am a bit of a whizz at the particular assembly code used on the platform in question (enough to have been able to throw together a disassembler in a couple of evenings).

Even with all that, it's probably going to take me a couple of months before I start understanding what's going on.

The Videocore GPU uses an unknown instruction set, is totally undocumented, and unless you're used to writing assembler code for massively parallel vector machines, you probably wouldn't have the faintest clue what the assembler was doing anyway. The "Binary Blob" ain't gonna get reverse engineered any time soon.

Mike Lake wrote:Is anyone suggesting that there will be things about the RPI that will never be fully open and therefore we will always have to rely on the good will of people inside Broadcom?

Weren't you the one questioning the viability of openness earlier?

Simon
Posts: 1330
Joined: Sun Sep 11, 2011 2:32 pm
by Stateside » Sun May 20, 2012 4:42 pm
What is Kurt ?

see post by Stateside » Fri May 18, 2012 2:41 am
Posts: 67
Joined: Sun May 13, 2012 12:26 pm
by andyl » Sun May 20, 2012 5:54 pm
Mike Lake wrote:There is a commercial point here. The RPi is already a fantastic success - the back orders at RS and Farnell make that clear - and it hasn't even got started yet.

With this volume of interest, and the amount of global publicity the project has managed to get for free (worth an arm and a leg), the RPi is a FORCE.

In a commercial world one would use that force to twist arms to get what you want. For example, not pick an image sensor simply because "that's what's there" but to specify an image sensor and find someone to do a run of them.

I'm like a record stuck on repeat - like many others I would prefer not just a camera module but also image sensors without IR filters that could be used for lots of other things besides pictures - and I think the RPi has the clout to get them.


Well it is a huge success in terms of the project. However that's not get carried away here. In the worldwide scheme of things the Pi is still pretty much a small player - especially when you consider that a camera module, let alone a more advanced image sensor like you want, isn't important to a lot of those users.

Also off the shelf components are nearly always going to be cheaper than custom components so going the mass market camera approach helps keep the costs down and consequently the market size up.

I guess in the end it would all come down to hard headed business decisions.
Posts: 265
Joined: Tue Jan 10, 2012 11:05 am
by Mike Lake » Mon May 21, 2012 12:09 pm
Tufty

[quote="tufty]Weren't you the one questioning the viability of openness earlier?[/quote]

It is my short memory about where the RPi came from - it's age you know <g>

It is not be possible at the moment to produce a truly open RPi-like device at the price the Foundation has managed (though it would be nice). So, the cost of getting a low cost device is that certain parts will remain closed - like the camera CSI stuff because it is proprietary to Broadcom. Fair enough.

I suppose that's a valid argument in favour of mixing the free and open with the commercial and proprietary.

Anyway, I am glad the RPi is a low cost device - despite being not totally open.

One could also argue against me by saying that the original design of the "IBM compatible PC" is open - and we have all benefitted from that - hardware and software providers. (IBM has, of course, pulled out of all hardware manufacture.) On the other side, Apple is an example of a totally closed system - including their monopolistic app Store. (Apple of course, seems to be thriving - though not with any of my money <g>)

I am happy to fit somewhere in the middle.
User avatar
Posts: 96
Joined: Sun Feb 12, 2012 11:45 am
by rurwin » Mon May 21, 2012 12:25 pm
There was a time when Apple was open, and it was selling the most popular personal computer for business use. Then it brought out the next version and made it closed. They were instantly niche.

Then IBM got tired of the clone makers and decided to make their next version closed. Does anyone remember the PS/2? Nobody bought any and only the keyboard/mouse plug survived. IBM pulled out of the hardware business soon after.

The moral seems to be that you may not make money if your product is open, but you will get much more market-share.
User avatar
Forum Moderator
Forum Moderator
Posts: 2890
Joined: Mon Jan 09, 2012 3:16 pm
by hippy » Mon May 21, 2012 3:29 pm
rurwin wrote:Does anyone remember the PS/2?

My monitor is sitting on a PS/2 base unit in front of me - it is however just a monitor riser!

Closed systems can work as long as they deliver what's wanted or have some USP that appeals to buyers but if competitors all go some other way they become non-standard by default; the MCA bus was roundly trumped by PCI.
Posts: 757
Joined: Fri Sep 09, 2011 10:34 pm
by piglet » Mon May 21, 2012 3:57 pm
Mike Lake wrote:I asked my colleagues if they would work for free in the public good - "in the long term interests of children in our schools. They all said no. One said "how the hell do I pay my mortgage if I am expected to work for free?"


That would work if we could get around the problem of selfishness, and all work to our best for free for the common good and consume only to our needs. I believe its called communism. I understand that it's not had a very good track record. Then again, nor has Capitalism. To misquote Churchill, Capitalism is the worst economic system except from all the others.
User avatar
Posts: 538
Joined: Sat Aug 27, 2011 1:16 pm
by tufty » Thu May 24, 2012 7:07 am
rurwin wrote:There was a time when Apple was open, and it was selling the most popular personal computer for business use. Then it brought out the next version and made it closed. They were instantly niche.

...

The moral seems to be that you may not make money if your product is open, but you will get much more market-share.


ORLY?

Apple top mobile computer sales Q1 2012
Posts: 1330
Joined: Sun Sep 11, 2011 2:32 pm
by rurwin » Thu May 24, 2012 7:56 am
To expand my original point, the Apple ][ had an open specification and could accept third-party add-on boards.It was the PC of choice before the IBM PC, and in many cases even after it. I was using one in 1984-85 to assemble code and blow EPROMs. It had all the advantages of the IM PC except for the name, and it had market penetration. Fighting IBM with a new machine would never have been easy, ("nobody ever got fired for buying IBM,") but they had significant advantages.

But the new Apple Macintosh was completely closed, even writing software for it required buying the SDK from Apple and was not at all easy since GUI was an innovation. They threw away their third-party support at the same time as the IBM PC was as welcoming as the Apple ][ had been. (Even the source-code for IBMs BIOS was published.) As a result the third-parties flocked to IBM, the software writers too, and IBM hardly had to trade on their name to take the marketplace by storm. The Apple Macintosh became a niche product for media people.

In the decades since, the Mac has become a little more mainstream and the iPhone/iPad have been highly successful.

And yes there are counter-examples, such as the iPad, but that is not as closed as the Macintosh was, and Android is winning against the iPhone. All tablets (and phones) are approximately as open as one another, and so openness is not a clear differentiator in the market.
User avatar
Forum Moderator
Forum Moderator
Posts: 2890
Joined: Mon Jan 09, 2012 3:16 pm
by tufty » Thu May 24, 2012 12:13 pm
rurwin wrote:To expand my original point, the Apple ][ had an open specification and could accept third-party add-on boards.It was the PC of choice before the IBM PC, and in many cases even after it. I was using one in 1984-85 to assemble code and blow EPROMs. It had all the advantages of the IM PC except for the name, and it had market penetration. Fighting IBM with a new machine would never have been easy, ("nobody ever got fired for buying IBM,") but they had significant advantages.

Openness didn't come into it. In the '80s , you decided what software you wanted to run, and then bought a machine to run it. Sure, if it could be expanded, so much the better, but if it couldn't - well, who cares. All the machines came with schematics anyway. Software was the driver, and Apple had one really influential piece of software that ran on the Apple II, and which made everything else look stupid. That software was Visicalc. Indeed, it wasn't until the release of Lotus 1-2-3 in 1983 that the PC had anything to beat it; and it was only in 1983 that Apple II sales started going into freefall, despite having earlier kicked it's supposed successor, the Apple III, into touch (although that wasn't hard - the Apple III was badly designed and plain didn't work until its last iteration, launched mere weeks before the entire product line was pulled).

Visicalc was the thing that pulled the "personal computer" out of the hands of geeks and tinkerers , and put it into the hands of business.
Lotus 1-2-3 was the thing that killed the Apple II
Excel was the thing that stopped Windows being a mere graphical launcher for DOS applications

The Mac was something completely new (well, not completely, there was the much unloved Lisa before it), and had its own "killer app" (beyond the awesomely graphic-ness of its interface) - PageMaker.

Apple was (and is) a hardware manufacturer, that's where its profits came from, and that's where they come from today. Trying to compete with the "IBM PC" (which, by 1984, was in fact a morass of more-or-less compatible, budget-basement clones) would have been suicide. The "PC" was bolted together from off-the-shelf commodity parts, the Mac was made, in-house, using specialised parts. Even with the Mac's graphical interface, there was no way Apple could have competed directly on price, even before Windows 3 made the PC and the Mac "indistinguishable" .

IBM didn't like the clone market either. They tried (and failed) to kill it off many times, most notably with the PS/2.
But the new Apple Macintosh was completely closed, even writing software for it required buying the SDK from Apple

Writing software for *everything* that didn't come with a language blown into its ROM required buying the tools to do it. That's largely the case even now.

In the mid '90s Apple tried opening up to the clone market. They nearly died because of it.

What has changed since the end of the '80s-mid-90s is that pretty much all the platforms are equivalent in terms of what you can do with them, so the purchasing choice made is more to do with the hardware / OS combination and not the software you want to run. In most cases, that decision is made purely on cost, unless there is another compelling reason to go with one over another. Apple's compelling reason is that their stuff, quite simply, works as advertised. It's clean, it's integrated, and it's sexy. That goes for the Mac, the iPhone and the iPad.
And yes there are counter-examples, such as the iPad, but that is not as closed as the Macintosh was

:shock:
So, you need an Apple computer in order to develop for it, you need to pay to develop for it, you can only distribute your software through Apple, and Apple have the final say as to whether you can distribute a particular piece of software or not. In what way is that somehow "more open" than the original Mac? At least with the original mac you could distribute what you liked, how you liked.
Android is winning against the iPhone.

Android is winning against the iPhone in the budget sector because :

a - Only one company makes the iPhone, and the iPhone is expensive
b - Android is a no-brainer for phone manufacturers, because it largely "just works" and has no licensing costs
c - Android enables the network operators to push out budget smartphones to people who'd like an iPhone but can't afford one (and thus increase their turnover; if observations of the people I know are anything to go by, this happens twice - once on the initial purchase of a budget Android smartphone, and once when they decide that it's a pile of cark and upgrade to an iPhone)

Apple are doing very nicely in the higher end smartphone market. Android is killing the rest - it's seen as being "as good", but it's pretty much free (MS's extortion racket aside).

Simon
Posts: 1330
Joined: Sun Sep 11, 2011 2:32 pm
by andyl » Thu May 24, 2012 1:06 pm
[quote="rurwin"]To expand my original point, the Apple ][ had an open specification and could accept third-party add-on boards.It was the PC of choice before the IBM PC, and in many cases even after it. I was using one in 1984-85 to assemble code and blow EPROMs. It had all the advantages of the IM PC except for the name, and it had market penetration. Fighting IBM with a new machine would never have been easy, ("nobody ever got fired for buying IBM,") but they had significant advantages.

But the new Apple Macintosh was completely closed, even writing software for it required buying the SDK from Apple and was not at all easy since GUI was an innovation. They threw away their third-party support at the same time as the IBM PC was as welcoming as the Apple ][ had been. (Even the source-code for IBMs BIOS was published.) [/q]

That is a bit revisionist. IBM did not initially publish the source for the BIOS, and brought lawsuits against the company who cloned it, and those computer manufacturers using that cloned BIOS.

As tufty mentioned there was also the PS/2 - or rather MCA. Plus at the time IBM were also pushing Token Ring hard as the right networking technology.

But yes the Lisa (and Mac) were very closed. You even needed special tools to be able to open the case.
Posts: 265
Joined: Tue Jan 10, 2012 11:05 am
by rurwin » Thu May 24, 2012 1:27 pm
andyl wrote:
rurwin wrote:(Even the source-code for IBMs BIOS was published.) [/q]

That is a bit revisionist. IBM did not initially publish the source for the BIOS, and brought lawsuits against the company who cloned it, and those computer manufacturers using that cloned BIOS.


The source code for the BIOS was listed in the Technical Manual (appendix A). Anyone using that code for a clone would still be guilty of copyright infringement, but developers could code against it without problem.
User avatar
Forum Moderator
Forum Moderator
Posts: 2890
Joined: Mon Jan 09, 2012 3:16 pm
by kdakin » Tue Jun 12, 2012 4:20 pm
I agree mostly with what Mike Lake has said.
It is about time someone took the "free" software advocates to task.
I have been a programmer for 40+ years, made a good living and created some excellent debugging products - that blue chip companies were happy to pay for - because it actually saved them money and allowed their programmers to produce more robust software. I could not have done that without being paid for my products.
I have produced similar software for the .Net environment (theoretically executable with the help of Mono under Linux), but I am unlikely to make any serious money from it as everyone will want it for nothing. I won't have any motivation for developing it further and obviously will also not be able to pay anyone to do it for me.
Competition is good - ask Professor Richard Dawkins.
Posts: 54
Joined: Sat Mar 24, 2012 7:20 am
by andyl » Thu Jun 14, 2012 9:01 am
kdakin you are misunderstanding the situation.

First generally there is tons of competition in the open-source world - emacs vs vi, kde vs gnome vs ..., etc. etc. There are a few niches where there is no competition TeX but generally that isn't the case.

Without open-source there wouldn't be much competition in some areas too - browsers for example. Take away Chrome and Firefox's rendering and javascript engines and you are left with IE having approx 97% market share.

Second no-one least of all open-source advocates (who are mostly developers too) want developers to not make any money. They just want to see a different business model. That business model obviously works to some degree - there are plenty of people (both individuals and companies) who make money off of open-source software.
Posts: 265
Joined: Tue Jan 10, 2012 11:05 am
by jamesh » Thu Jun 14, 2012 9:14 am
kdakin wrote:I agree mostly with what Mike Lake has said.
It is about time someone took the "free" software advocates to task.
I have been a programmer for 40+ years, made a good living and created some excellent debugging products - that blue chip companies were happy to pay for - because it actually saved them money and allowed their programmers to produce more robust software. I could not have done that without being paid for my products.
I have produced similar software for the .Net environment (theoretically executable with the help of Mono under Linux), but I am unlikely to make any serious money from it as everyone will want it for nothing. I won't have any motivation for developing it further and obviously will also not be able to pay anyone to do it for me.
Competition is good - ask Professor Richard Dawkins.


If its the only product that does what it does, and it does it well, why can you not charge money for it? Even if people 'want' it for nothing, if they cannot 'get' it for nothing, they have to pay. Or are you saying they won't buy at all?

I want a Lamborghini Gallardo. I want to pay £20k. They won't sell it to me for £20k. Then either I don't have it, or I pay the market price.
Raspberry Pi Engineer
Raspberry Pi Engineer
Posts: 10590
Joined: Sat Jul 30, 2011 7:41 pm
by tufty » Thu Jun 14, 2012 10:52 am
jamesh wrote:I want a Lamborghini Gallardo. I want to pay £20k. They won't sell it to me for £20k. Then either I don't have it, or I pay the market price.

I bet they're a bastard to hotwire, too. Best bet's to hang around outside a posh restaurant or hotel and pretend to be the bloke that parks the cars. Works without fail in the movies.

Simon
Posts: 1330
Joined: Sun Sep 11, 2011 2:32 pm