Not long ago there were comments here that some software just must be sold as a licence to use. This was given as a reason that FLOSS could never succeed in the market. Well, Adobe, one of the flagships of non-FLOSS software has started selling pay-as-you-go instead of pay-up-front. Sweet. It’s still not a cloud-service, but that will come. Now Adobe will have a large increase in customers because they changed their licensing model to pay-as-you-go. That’s proof that paid-up-front licences are not the only way to go. Now little guys/startups will more easily be able to afford Adobe’s products. Because the service is open-ended the profit potential increases. Imagine that.
This licensing model has all kinds of advantages for businesses. Pay-as-you-go is an immediate tax deduction instead of depreciation which is spread over years. FLOSS can use a pay-as-you-go model as well. Red Hat has been very successful with it.
17516
12796
206
3
2
23973 
11881 
11733 
4655 
4286 
1641 
198 
14 
2 
0 
0 
0 
“Not long ago there were comments here that some software just must be sold as a licence to use”
No, that’s not what I said. I said the software couldn’t be FLOSS. “Pay upfront” or “pay monthly” doesn’t matter – the key thing here is that development is funded by software sales. The software is proprietary. Non-free.
You still haven’t explained how professional applications such as Adobe Creative Suite can be developed on the FLOSS model. When thousands of man-years of development are required, at a cost of hundreds of millions of dollars, how can you possibly raise the required funds without having some sort of plan to sell the end product?
Or you could acknowledge that some important software just couldn’t exist without a proprietary, non-Free development model.
Millions of person-hours of labour has gone into developing Linux. One way or another if a product is useful its development will be funded. The world can make its own software. No need to buy licences for what the world makes for itself.
Suppose professionals of type X decide they need application Y. They consult/hire a project leader who organizes a team. They work on contract for several months and produce the application. Using GNU/Linux they don’t have to write any new file systems or OS or utilities. They just write the algorithms the pros need. When it’s done there is a FLOSS app that everyone can use, not just the pros. The pros could decide to sell training/whatever to recoup their investment or just count the savings on IT forever.
An example of this happening was GCompris, an educational tool for elementary teachers/students. A bunch of teachers got together and made it happen. Now they have about 80 apps in a suite that helps them do their jobs.
see http://gcompris.net/-en-. GCompris is used globally.
If you want to add new boards, download and run GCompris, it comes with a full documentation that describes the GCompris internal and a step by step instruction on how to add boards. To give an idea of the amount of work, a board is usually set in a single file of about 500 lines of C code or Python. GCcompris is based on the goocanvas).
see GooCanvas
Pogson, this is a good point, I think you’re alluding to that LINUX is a very early example of crowdsourcing, one where members need a high level of skill to chip in(programmers) but do so by coming together w/o meeting face to face.
No, you **still** haven’t explained how professional applications can be developed on the FLOSS model.
Are we talking about different things? I say there are some applications that can’t be FLOSS and give some examples, and you always respond by giving unrelated examples of FLOSS applications.
It’s as if I say “Some swans are black”. And you respond by saying “But this swan is white!”
In your example, the application Y has to be sufficiently useful to X that they are willing to spend their own money developing it in house. Y must be a net gain for the business. Right?
Now, no matter how much Y benefits the business, it would benefit it **even more** if copies were sold to others. You assert that Y is very useful to X, because you assume that they’ll develop it even if they can’t sell it. You also say that others will seek training to use Y. So whatever Y is, it must be pretty darned useful. (And X are insane not to sell it.)
Now, let’s suppose that Y is slightly less useful, so the benefits of developing Y are only marginal. Once the costs of development are subtracted, the net benefit to the company is almost zero. Or negative.
In **this** case, Y **won’t** exist unless there is some way to increase the benefit. For example, by selling copies.
And there you have it. If every artist had to write his own Photoshop then it wouldn’t be worth doing. If Adobe wrote Photoshop and then didn’t sell it, it would again be a waste of effort. But selling Photoshop means that the cost of development can be shared by millions of users. Impossible with FLOSS, where only the benefits are shared, and the costs are a dead loss.
Another way it can work is a big outfit like IBM, RedHat, etc. forming a project in-house and opening it to the world. It is quite feasible for larger users of IT to create their own software and by using FLOSS, they can cut their development costs and produce a valuable product to make friends or money. Governments should be supporting LibreOffice instead of paying M$ for work M$ did years ago repeatedly. SUN bought the producer of OpenOffice.org for less than the cost of one year of licensing fees to M$.
Software is software. Any of it can be produced as FLOSS. We have games, office suites, operating systems, GUIs, libraries, drivers,… all FLOSS. Where is the problem?
The world is HUGES! Any software the world wants, it can make. It does not need M$ to make it.
e.g. games. Where did all the game software manufacturers come fromt? Some enthusiast started making his own game. He could just as well started as Linus did and invite friends to work on it with him. No software is too sacred or too valuable to develop in the FLOSS manner. Proof is that all manner of software has been produced via FLOSS. Even before RMS and GNU, I was working on FLOSS software that folks at universities were sharing around. you visited, you carried a magnetic tape or a box of cards and made copies of stuff. That was before the Internet. Now everything is faster and easier and FLOSS works even better.
Historically, M$ and others are the aberration, not FLOSS.
“If every artist had to write his own Photoshop then it wouldn’t be worth doing.”
No it wouldn’t. And it wouldn’t be worth it for a Finnish college student to write his own operating system… but he did.
I’m not getting anywhere at all with this. I think you win this argument by attrition because I just can’t be bothered any more.
I thought I was being harsh using the swan analogy, but I’m really not.
I say “Some software can’t be produced using the FLOSS model”, i.e. “Some swans are black”.
And you (and Richard Chapman) respond “But this software *was* produced using the FLOSS model”, i.e. “This swan is white”.
I don’t know what I need to say so that we can even be on the same page here.
Maybe I should close with a story. Once I was working on a project with some guys from ARM, and I was asking them about their product development process. I found out that they were spending some £40million per year on software licences for EDA tools. Surprised by this, I asked them if there might be any lower-cost alternative.
“No,” one replied. “And our business would be impossible without that software.”
In FLOSS world, ARM wouldn’t be able to buy EDA tools. They would have to write them in-house.
In FLOSS world, ARM CPUs would not exist, because ARM would have spent the last 30 years writing the EDA tools to enable the sort of work they do. They couldn’t just buy them from Mentor Graphics or Cadence or whoever. Those are *software* companies. They only sell software – an invalid business model in FLOSS world.
I don’t know what sort of CPUs you would have in FLOSS world. Maybe you would still have Intel CPUs, because large companies would have the resources to develop all their own tools in house and then keep them secret. But there would be no Loongsons and no ARMs.
The world needs CPUs. Someone will build them. SUN open-sourced its design. Anyone in the world can build them. The monopoly on CPUs is fading.
“EDA tools” is a generic term for software and equipment for laying out complex electronic circuits, such as integrated cirsuits/chips.
see Electronic Design Automation
“The earliest EDA tools were produced academically. One of the most famous was the “Berkeley VLSI Tools Tarball”, a set of UNIX utilities used to design early VLSI systems. Still widely used is the Espresso heuristic logic minimizer and Magic.”
Yes. I remember the 1970s. People did write software to lay out their pet project. I know a guy who wrote such a programme to lay out optimally a whole rack of “wire-wrap” circuitry. EDA as an industry did not even start until 1981. It was all in-house and often open-source before then. Folks got along. The first integrated ciruits were developed in the 1960s and the first “modern” microprocessors were developed in the 1970s, long before EDA became a business.
There is FLOSS software to do that, for instance, gEDA, in Debian. It’s for printed circuit boards. GNUCAP can simulate circuitry right down to the transistor. There are many other packages. ARM works closely with a number of EDA companies. FLOSS EDA packages are also listed in Wikipedia
So, the world can produce FLOSS EDA, and does. Magic was written in 1980 and is still in development and wide use.
RedHat / Fedora is in on EDA as well.