Business
Eric Raymond: Why open source will rule
Linux and the open-source movement have come a long way since the foundation of the Open Source Initiative in the late 1990s, Raymond says. It now has the credibility it needs to succeed in the business world
Eric Raymond believes Linux is on a roll.
Raymond is best known as the co-founder, with Bruce Perens, of the Open Source Initiative (OSI) to promote Linux and other "free" software to businesses in a language they could understand. The OSI has largely succeeded in its aims, Raymond says, with backers like IBM heavily promoting Linux and big companies adopting the operating system for both back-end systems and desktops. Raymond is also the author of The Cathedral and the Bazaar and other open-source texts.
For evidence that open-source movement now has the mainstream credibility it lacked in the late 1990s, Raymond points to Microsoft's failed attempts last year to discredit Linux and the GNU Public Licence (GPL) on which it is based. Now Linux and the open-source development model are well-positioned to succeed in the increasingly complex world of software development.
Raymond spoke to ZDNet UK during a recent European speaking tour under the auspices of the UK and Danish Unix Users Groups.
Q: You co-founded the Open Source Initiative in late 1998 with the goal of making free software credible outside of the hacker community. What progress has been made since then?
A: I think to a large extent we succeeded in our initial programme, which was in positioning the open-source label as something the typical corporate manager could deal with without being frightened. And I think one of the effects of that has been to empower advocates of open source inside corporations. Now they have a set of resources to point to that aren't as it were ideologically tainted... there are elements of the Free Software Foundation that are not real happy with that. Microsoft has been keen to undo some of that work with its anti-Linux, anti-GPL (GNU Public Licence) campaign.
Thankfully they've failed. It looked a little dicey there early last year, but they've failed. What happened was they floated a few trial balloons, and a group of open-source leaders put out a public statement that basically wrapped a tyre iron around their heads, and the trade press bought our story and didn't buy Microsoft's, and they've kept fairly quiet since. Is there a danger of new laws coming into effect that could protect proprietary software companies at the expense of open source?
Of course there is existing law on the books that is kind of problematic, most notably the DMCA (Digital Millennium Copyright Act), and UCITA (Uniform Computer Information Transactions Act, a software licensing act) in the US. Those are still problems which we're coping with in various ways through lobbying and legal challenges, but I don't think Microsoft succeeded in creating any new challenges. It's hard to say at the expense of open source, because open source doesn't threaten anybody's intellectual property, you don't have to join that game unless you want to. The threats are more from legislation that's intended to accomplish IP protection in areas outside of software that have spillover effects on software development, the DMCA of course being the best known example of such a thing. There have been some encouraging developments recently. The EFF has a brief that makes a fairly strong argument that section 12 of the DMCA is unconstitutionally vague. Under US constitutional law they have a pretty strong challenge going. What do you think of Sun Microsystems' decision to start charging for StarOffice?
In that case StarOffice just died. They just shot StarOffice through the head. It doesn't matter whether I'm in favour of it or not. But if OpenOffice still exists, and it's GPLed, and they're going to start charging for StarOffice, then they just shot StarOffice through the head. Some open-source companies are starting to add proprietary software to their open-source offerings as a way of increasing revenues. Does that threaten the open-source movement?
In fact the company I'm associated with, VA Software, recently announced they're going to be selling extensions on SourceForge that are held proprietarily. I don't see it happening to an extent that really jeopardises any of the core open-source software or applications, otherwise I'd be much more concerned about it. And it's natural that people are going to be more conservative and careful when the economy -- not just in the US but all through the developing world, where most of the software development is going on -- is fairly recessionary right now. That makes managers more cautious and conservative and amplifies any tendency toward over-protectivity that they have. How much does that recessionary atmosphere improve the attractiveness of Linux and other open-source software?
It's a two-edged sword. We're seeing an increasing number of stories about increasing Linux take-up, even on the desktop in large corporations, and it's very clear that what's going on here is that IT managers are under tremendous pressure to cut their budgets and cut expenses. Disruptive technologies really thrive under conditions where people are looking to cut budgets. You get a disruptive technology... that is initially much lower price and reliability, but it's cost-effective in niche markets where low cost is really important. So the technology will get a little bit of a foothold in there, and the manufacturers of the disruptive technology will use the revenue stream they get from that to gradually improve it. And you may get to the point where the disruptive technology improves enough to match actual demand, as opposed to the demand that the sustaining technologies think is there because they're always looking at their highest-margin customers. And when that happens, the sustaining technologies will just fall off a cliff. That market will just collapse. What's happening in software on the broadest scale is that open source in general and Linux in particular is a disruptive technology attack on the traditional proprietary business model and its titans, including outfits like Microsoft and SAP and so forth. And the thing is that under conditions where everybody is under pressure to cut budgets, disruptive technologies get more and more attractive because they cost a lot less. It's a tough row to hoe if you're a sustaining technologist, because there's no point at which it looks rationally appropriate to cannibalise your own business and adopt a new technology, because the margins are lower and your shareholders will kill you. It always looks until the moment of collapse like you're going to be able to continue playing the sustaining technology game and collect fatter margins, but then the world changes, the bottom falls out of your market, and you're gone. It has been argued that in some critical aspects of desktop software, like maintaining a consistent user interface where different applications work seamlessly together, just work better when there's a big controlling company in charge. What do you make of that argument?
I don't think that's a real issue, but there's a closely related issue that is real. I don't think it's necessary to have a single player dominating user interfaces if you have a development community that is alive to the necessity of having a uniform interface, and prepared to make that a priority. In fact, the Linux desktops have already successfully done this. You may note that drag and drop works correctly between GNOME and KDE applications. That's not an accident, it happened because the GNOME and KDE people reached out to each other, said, "We've got to have a common drag-and-drop protocol," wrote a standard, and now the applications on both sides conform to it. That happened in spite of the fact that there was no single player that controlled both GNOME and KDE that was able to compel that interface uniformity. I think we've demonstrated in open source that it is possible to bridge those gaps and create a uniform interface. There's a closely related issue, however that I don't know how to solve yet without a big player with a lot of money, which is doing systematic user interface end user testing. We're not very good at that yet, we need to find a way to be good at it. It's the actual mechanics of setting up large-scale focus group testing with end users. The problem is they're not getting feedback from large-scale end user testing, and that's allowing a certain spikiness in the interfaces to persist that could be smoothed out otherwise. Another argument says that historically speaking, monopolies like Microsoft or the railway companies are favoured when infrastructure is being built, but once the infrastructure phase is over they shrivel up.
I think that's absolutely backwards. It's more important when you're standardising infrastructure that the infrastructure not be under the control of a particular party, otherwise you get critical patents, and trade secret protection and other forms of proprietary control locking everybody else out of that infrastructure forever. I don't think I buy it. Traditionally, when you see cases of horizontal infrastructure being built by monopolies, typically it happened because some player was able to co-opt the government early, and create barriers to entry. That's obviously not the case in the software business, thank goodness. Microsoft has exercised a lot of control over the software market, though.
Well they have, but they haven't been able to, for example, prevent new entrants in the software field. They haven't been able to do things like require government certification of applications with standards that favour big players, which is the classic ploy that you see in monopolisation efforts. It's the classic way of co-opting the government, when you can't buy the government outright, which is frequently what's done. Bruce Perens has argued that big companies like IBM who use open-source software should give some of their own patented intellectual property back to the open source world. Does that seem like a reasonable argument?
Explicitly jaw-boning people on issues like that I think is often counterproductive. The approach I'd prefer is to point it out fairly quietly that there are times when it makes economic sense to give up proprietary control, and lay out all the arguments and let people make their own conclusions. You can't push heads of Fortune 500 companies around; they've got lots of money and they've got big egos. If you try to try and jawbone them, they're going to be counter-suggestible, they're going to dig their heels in. So I'd prefer to take a quieter approach. Next page: Why Linux will take over the desktop, and why Microsoft is doomed
A: I think to a large extent we succeeded in our initial programme, which was in positioning the open-source label as something the typical corporate manager could deal with without being frightened. And I think one of the effects of that has been to empower advocates of open source inside corporations. Now they have a set of resources to point to that aren't as it were ideologically tainted... there are elements of the Free Software Foundation that are not real happy with that. Microsoft has been keen to undo some of that work with its anti-Linux, anti-GPL (GNU Public Licence) campaign.
Thankfully they've failed. It looked a little dicey there early last year, but they've failed. What happened was they floated a few trial balloons, and a group of open-source leaders put out a public statement that basically wrapped a tyre iron around their heads, and the trade press bought our story and didn't buy Microsoft's, and they've kept fairly quiet since. Is there a danger of new laws coming into effect that could protect proprietary software companies at the expense of open source?
Of course there is existing law on the books that is kind of problematic, most notably the DMCA (Digital Millennium Copyright Act), and UCITA (Uniform Computer Information Transactions Act, a software licensing act) in the US. Those are still problems which we're coping with in various ways through lobbying and legal challenges, but I don't think Microsoft succeeded in creating any new challenges. It's hard to say at the expense of open source, because open source doesn't threaten anybody's intellectual property, you don't have to join that game unless you want to. The threats are more from legislation that's intended to accomplish IP protection in areas outside of software that have spillover effects on software development, the DMCA of course being the best known example of such a thing. There have been some encouraging developments recently. The EFF has a brief that makes a fairly strong argument that section 12 of the DMCA is unconstitutionally vague. Under US constitutional law they have a pretty strong challenge going. What do you think of Sun Microsystems' decision to start charging for StarOffice?
In that case StarOffice just died. They just shot StarOffice through the head. It doesn't matter whether I'm in favour of it or not. But if OpenOffice still exists, and it's GPLed, and they're going to start charging for StarOffice, then they just shot StarOffice through the head. Some open-source companies are starting to add proprietary software to their open-source offerings as a way of increasing revenues. Does that threaten the open-source movement?
In fact the company I'm associated with, VA Software, recently announced they're going to be selling extensions on SourceForge that are held proprietarily. I don't see it happening to an extent that really jeopardises any of the core open-source software or applications, otherwise I'd be much more concerned about it. And it's natural that people are going to be more conservative and careful when the economy -- not just in the US but all through the developing world, where most of the software development is going on -- is fairly recessionary right now. That makes managers more cautious and conservative and amplifies any tendency toward over-protectivity that they have. How much does that recessionary atmosphere improve the attractiveness of Linux and other open-source software?
It's a two-edged sword. We're seeing an increasing number of stories about increasing Linux take-up, even on the desktop in large corporations, and it's very clear that what's going on here is that IT managers are under tremendous pressure to cut their budgets and cut expenses. Disruptive technologies really thrive under conditions where people are looking to cut budgets. You get a disruptive technology... that is initially much lower price and reliability, but it's cost-effective in niche markets where low cost is really important. So the technology will get a little bit of a foothold in there, and the manufacturers of the disruptive technology will use the revenue stream they get from that to gradually improve it. And you may get to the point where the disruptive technology improves enough to match actual demand, as opposed to the demand that the sustaining technologies think is there because they're always looking at their highest-margin customers. And when that happens, the sustaining technologies will just fall off a cliff. That market will just collapse. What's happening in software on the broadest scale is that open source in general and Linux in particular is a disruptive technology attack on the traditional proprietary business model and its titans, including outfits like Microsoft and SAP and so forth. And the thing is that under conditions where everybody is under pressure to cut budgets, disruptive technologies get more and more attractive because they cost a lot less. It's a tough row to hoe if you're a sustaining technologist, because there's no point at which it looks rationally appropriate to cannibalise your own business and adopt a new technology, because the margins are lower and your shareholders will kill you. It always looks until the moment of collapse like you're going to be able to continue playing the sustaining technology game and collect fatter margins, but then the world changes, the bottom falls out of your market, and you're gone. It has been argued that in some critical aspects of desktop software, like maintaining a consistent user interface where different applications work seamlessly together, just work better when there's a big controlling company in charge. What do you make of that argument?
I don't think that's a real issue, but there's a closely related issue that is real. I don't think it's necessary to have a single player dominating user interfaces if you have a development community that is alive to the necessity of having a uniform interface, and prepared to make that a priority. In fact, the Linux desktops have already successfully done this. You may note that drag and drop works correctly between GNOME and KDE applications. That's not an accident, it happened because the GNOME and KDE people reached out to each other, said, "We've got to have a common drag-and-drop protocol," wrote a standard, and now the applications on both sides conform to it. That happened in spite of the fact that there was no single player that controlled both GNOME and KDE that was able to compel that interface uniformity. I think we've demonstrated in open source that it is possible to bridge those gaps and create a uniform interface. There's a closely related issue, however that I don't know how to solve yet without a big player with a lot of money, which is doing systematic user interface end user testing. We're not very good at that yet, we need to find a way to be good at it. It's the actual mechanics of setting up large-scale focus group testing with end users. The problem is they're not getting feedback from large-scale end user testing, and that's allowing a certain spikiness in the interfaces to persist that could be smoothed out otherwise. Another argument says that historically speaking, monopolies like Microsoft or the railway companies are favoured when infrastructure is being built, but once the infrastructure phase is over they shrivel up.
I think that's absolutely backwards. It's more important when you're standardising infrastructure that the infrastructure not be under the control of a particular party, otherwise you get critical patents, and trade secret protection and other forms of proprietary control locking everybody else out of that infrastructure forever. I don't think I buy it. Traditionally, when you see cases of horizontal infrastructure being built by monopolies, typically it happened because some player was able to co-opt the government early, and create barriers to entry. That's obviously not the case in the software business, thank goodness. Microsoft has exercised a lot of control over the software market, though.
Well they have, but they haven't been able to, for example, prevent new entrants in the software field. They haven't been able to do things like require government certification of applications with standards that favour big players, which is the classic ploy that you see in monopolisation efforts. It's the classic way of co-opting the government, when you can't buy the government outright, which is frequently what's done. Bruce Perens has argued that big companies like IBM who use open-source software should give some of their own patented intellectual property back to the open source world. Does that seem like a reasonable argument?
Explicitly jaw-boning people on issues like that I think is often counterproductive. The approach I'd prefer is to point it out fairly quietly that there are times when it makes economic sense to give up proprietary control, and lay out all the arguments and let people make their own conclusions. You can't push heads of Fortune 500 companies around; they've got lots of money and they've got big egos. If you try to try and jawbone them, they're going to be counter-suggestible, they're going to dig their heels in. So I'd prefer to take a quieter approach. Next page: Why Linux will take over the desktop, and why Microsoft is doomed
For all your GNU/Linux and open source news, from the latest kernel releases to the newest distributions, see ZDNet UK's Linux Lounge.
Have your say instantly, and see what others have said. Go to the Linux forum.
Let the editors know what you think in the Mailroom.