I agree that it would benefit RH to increase their marketing but they have more to gain (profit-wise) from spending money on R&D. Redhat is on the brink of profitability already with almost negligible marketing. What sells Linux is the technology and what sells RH distributions is the quality and testing they do. That's why I bought RedHat.
If Joe User's video card doesn't work or if Widgets-R-Us can't get Linux using their VME backplane, they'll switch back to windows98 or LynxOS, no matter how heavily they've marketed themselves. For RH to succeed, they need to have a continually improving product. If the rest of the OS community doesn't provide the software they need, it makes sense for them to do it themselves. It's a lot cheaper to improve the product when you can rely on improvement that other contributors have made. RH and the Linux community could get a HUGE bang for their R&D buck (compared to closed-source R&D). Things improve a whole lot faster in the OS world than in the proprietary world.
RH relies on the OS movement so they aren't going to be able to go proprietary on us. We'd cut them off if they did. Hell we've come close to slaying MS, RH would be a piece of cake (since their entire product is owned by us). Besides, they know that if we they give to us, we will give back to them. People trust RH and they provide a valuble service. If they want new and expanding markets:
use that money to develop new drivers.
Create a driver developer kit and documentation for vendors.
Write some good books or start some education programs
He says that virtually all of redhat's revenue comes from box sales ("take that open source freaks").
Funny, I thought virtually all of redhat's revenue comes from Linux (and the sweat of unpaid thousands). Let's face it: RedHat is primarily a service company and the service they provide is putting together useful distributions and testing them. They also do a fair amount of code cutting too but that's not why I buy RH: I buy it for the months of effort it would take to put together my own GNU/Linux configuration.
He uses RedHat as an example of why the OS movement is bogus but they actually are living proof that you don't need to "own" the code to make money off it!
Getting companies to pay for development certainly is a problem (hell, getting a customer to pay for development is often a problem...) but its not the obstacle you make it out to be.
Something like 80% (does anyone have a citable stat for this?) of the money spent on software development is for customized vertical applications. I've happily worked for almost 10 years in the industry and never once considered working for a company that produced a mass-market product. Look at the traditional sources of UNIX development: the telecom industry, defence industry, universities and scientific research organizations. If they feel that its worthwhile to pay you to develop some tool or module, they'll pay to get it done. If you write a perl module and work for Alcatel or Cisco, they aren't going to spend the money trying to market and sell it to the world. Will they care if you GPL it? Usually not. It's good PR.
If the software needs to be built, it will be built. A lot of times it will be paid for and built by companies that don't sell software. It doesn't affect them one way or another if you give it away.
True there are some companies that specialize in developing software and will go to great lengths to protect any development they do. But many companies develop software even though they don't compete in the software development market. Its just a means to an end and it doesn't hurt them a bit to give it away.
This is how Hoare (the "average slashdot essay writer" in question) characterized his work:
I've written a small page outlining what I think to be wrong with the new "bounty hunter" open source models, i.e. where the developer asks for hackers to code something for a reward. The article is available here.
I'll omit the obvious comment about your time "spent" in academia ("to [sic] much time").
I think every Hoare's aspect analysis will prove to be correct. Although I'd love to see OS developers to be rewarded financially for their work, money will change everything. I can't see a way to make it workable.
Perhaps the best way for corporations to support OS development is to migrate their "intellectual property" to an OS model.
It seems to me that at the root of capitalism is the idea that the worth of something can be measured relative to some scale (money). This seems to work fairly well both in the large and in small markets. We can measure the worth of a product that someone made or the labour that someone invested into performing a task.
What concerns me in the last dime of the 20th century is how rich someone can get without actually doing any work or producing any product. I personally have become wealthier than my entire family (going back 3 generations) simply because I had stock options. Apparently, "risk" is worth more than labour or production in today's economy.
I also observe that one can make money simply by keeping a secret (ie. intellectual "property", etc).
The absurdity of this makes me question the stability of these ideas and this economic system. Making money from nothing (which is what apparently has happened to me) leaves me with a very uncomfortable feeling. I'd have more confidence in the world that I'll leave for my children and grandchildren if the virtues valued by our economy were labour and productivity, not risk and IP.
Slashdot Mirror
If Joe User's video card doesn't work or if Widgets-R-Us can't get Linux using their VME backplane, they'll switch back to windows98 or LynxOS, no matter how heavily they've marketed themselves. For RH to succeed, they need to have a continually improving product. If the rest of the OS community doesn't provide the software they need, it makes sense for them to do it themselves. It's a lot cheaper to improve the product when you can rely on improvement that other contributors have made. RH and the Linux community could get a HUGE bang for their R&D buck (compared to closed-source R&D). Things improve a whole lot faster in the OS world than in the proprietary world.
RH relies on the OS movement so they aren't going to be able to go proprietary on us. We'd cut them off if they did. Hell we've come close to slaying MS, RH would be a piece of cake (since their entire product is owned by us). Besides, they know that if we they give to us, we will give back to them. People trust RH and they provide a valuble service. If they want new and expanding markets:
use that money to develop new drivers.
Create a driver developer kit and documentation for vendors.
Write some good books or start some education programs
He says that virtually all of redhat's revenue
comes from box sales ("take that open source
freaks").
Funny, I thought virtually all of redhat's revenue
comes from Linux (and the sweat of unpaid
thousands). Let's face it: RedHat is primarily
a service company and the service they provide
is putting together useful distributions and
testing them. They also do a fair amount of
code cutting too but that's not why I buy RH: I
buy it for the months of effort it would take to
put together my own GNU/Linux configuration.
He uses RedHat as an example of why the OS
movement is bogus but they actually are living
proof that you don't need to "own" the code to make money off it!
Getting companies to pay for development certainly is a problem (hell, getting a customer to pay for development is often a problem...) but its not the obstacle you make it out to be.
Something like 80% (does anyone have a citable stat for this?) of the money spent on software development is for customized vertical applications. I've happily worked for almost 10 years in the industry and never once considered working for a company that produced a mass-market product. Look at the traditional sources of UNIX development: the telecom industry, defence industry, universities and scientific research organizations. If they feel that its worthwhile to pay you to develop some tool or module, they'll pay to get it done. If you write a perl module and work for Alcatel or Cisco, they aren't going to spend the money trying to market and sell it to the world. Will they care if you GPL it? Usually not. It's good PR.
If the software needs to be built, it will be built. A lot of times it will be paid for and built by companies that don't sell software. It doesn't affect them one way or another if you give it away.
True there are some companies that specialize in developing software and will go to great lengths to protect any development they do. But many companies develop software even though they don't compete in the software development market. Its just a means to an end and it doesn't hurt them a bit to give it away.
I'll omit the obvious comment about your time "spent" in academia ("to [sic] much time").
I think every Hoare's aspect analysis will prove to be correct. Although I'd love to see OS developers to be rewarded financially for their work, money will change everything. I can't see a way to make it workable.
Perhaps the best way for corporations to support OS development is to migrate their "intellectual property" to an OS model.
It seems to me that at the root of capitalism is the idea that the worth of something can be measured relative to some scale (money). This seems to work fairly well both in the large and in small markets. We can measure the worth of a product that someone made or the labour that someone invested into performing a task.
What concerns me in the last dime of the 20th century is how rich someone can get without actually doing any work or producing any product.
I personally have become wealthier than my entire family (going back 3 generations) simply because I had stock options. Apparently, "risk" is worth more than labour or production in today's economy.
I also observe that one can make money simply by keeping a secret (ie. intellectual "property", etc).
The absurdity of this makes me question the stability of these ideas and this economic system.
Making money from nothing (which is what apparently has happened to me) leaves me with a very uncomfortable feeling. I'd have more confidence in the world that I'll leave for my children and grandchildren if the virtues valued by our economy were labour and productivity, not risk and IP.