Although open source software comes with lofty (and sometime altruistic) ambitions, some developers worry that the model just doesn't present a practical plan for the future of the software industry. Shawn Shell argues that proprietary software is here to stay — for a lot of good reasons. Why should you bother with looking at open source software though? Isn't it safer to stick with Microsoft and the other big corporate software designers? David Chisnall helps us to distinguish between proprietary (sometimes referred to as "predatory") software and its open source counterparts.
Post a Comment
Member since:
---
Member since:---
"Open source software is not necessarily non-commercial."
The "commercial" aspect of OSS isn't the software itself. It's usually some "not easily duplicated" add-on. For example the polishing and integration of various F/OSS into something useful were one is paying for labour, and expertise. Even selling it on CD's/DVD's is the paying of materials and convienence. Paying a programmer to work on OSS is paying for the add-on's that he writes. In all cases it's not the "free" software that's being payed for.
Member since:2005-07-06
Open source software is not necessarily non-commercial.
Very true. Conversely, proprietary software isn't always commercial eg. all the freeware out there. To further complicate matters some proprietary software is open because they are built to common standards or publicly available specs (without being patented.) and the code of some open source projects is so obscure the label open seems to be applicable ironically only.
Black and white it ain't.
Member since:2005-07-06
Proprietary is not a differentiating word. In legal sense even GPLed software is proprietary.
Better to refer to it as free v.s. unfree software. The problem is that the word free is a bit ambiguous. We need to point out that we mean free as in free speech.
If we used the terminology free v.s. dictator software that would be a bit more clear, but some people may find that way of defining closed source propriatory software bit trollish.
Member since:2005-07-06
Member since:2005-10-02
Member since:
---
I was thinking of posting a comment, back in the Open Document topic, that WordPerfect has nothing to fear from Open Office if it adopts the Open Document format, because open and professional aren't the same thing. Photoshop versus Gimp is a handy enough example to back this up.
The more I think about it, the more I think the advantages of open and free versus closed and commercial, haven't been as thoroughly examined by both sides of the argument as might be the case, and this article makes an interesting contribution to that discussion.
I'm going to put my head on the chopping block with this one and say that both sides need to move a little closer to the middle. Stallman et al might have to swallow the idea that, sometimes, untrammeled GPL might be a mistake, as Microsoft found out that untrammeled commercialism and secrecy was a mistake.
Member since:
---
The open source model is the typical model of traditional scientific research. It has worked very effectively in the past.
The pharmaceutical industry relies to a very significant on open source. Pure government-funded research is used as a basis for developing drugs. The pharmaceutical industry reciprocates with direct and indirect funding.
In comparison with the computer industry the pharmaceutical industry has no Microsoft, a dozen IBMs , hundreds of Apples and thousands of Red Hat sized corporations. The other difference is that pharmaceutical companies are generally highly profitable.
Member since:2005-08-06
There is a huge different with the scientific research, so your comparison is flawed it seems.
Pretty much all scientifical research is highly dependent on patents. In fact without patents, it's likely much of the commercial research in the pharmaceutical industry would vanish.
Member since:2005-07-06
Right, but in many ways, the patent system (minus software patents) actually has a lot in common with open source. Before the patent system, inventors would keep their inventions secret in order to have an advantage over their competitors. Patents allowed inventors to publish their inventions publicly, while avoid being ripped-off by competitors.
That's what science is all about. If we all pool together our knowledge everyone will be able to build on the work of others ("standing on the shoulders of giants") instead of having to start from zero.
That doesn't mean that the patent system is perfect, but it's not all evil.
Member since:2005-08-06
That doesn't mean that the patent system is perfect, but it's not all evil.
I agree completely. What instead should be done is defining how fast a market moves and then start defining how long a patent should last for each sector.
For instance, in terms of Software, I find 2 years to be a reasonable length for a patent to last. Maybe 2 years for complete protection and between year 2-5 royalty based on a fixed fee or something.
For pharmaceutical research maybe 10 years is more reasonable.
What I'm saying is basically that OSS is not the solution to everything, nor is proprietary stuff. The balance is good with a working patent system.
Member since:
---
Member since:
---
It seems that OSS/free software has been most successful for infrastructure (Apache server, tomcat, linux, tapestry, axis, rails, subversion, mysql). Note that all these products are copies of exiting proprietary products.
However, for non-infrastructure programs (office, CRM, ERP, publishing, photo/drawing software) proprietary software reigns. I think there are a few reasons for this:
1) OSS types are geeks that aren't good a UIs. UIs are key for non-infrastructure software
2) infrastructure is easier than innovation
3) if people have something truly innovative, why would they want it to be free? *Maybe* open source, but not free.
Member since:---
There's one thing I've noticed about F/OSS. A lot of it started under the "Cathedral" model and was then released. e.g. Netscape, Openoffice. Or was purchased by OSS e.g. Blender. A lot of it presently is being sustained by "Cathedral" companies. There however are fewer pure, widely used, "Bazaar" OSS programs.
Member since:2005-07-06
Admittedly I haven't read "The Cathedral and the Bazaar," but when you say that there are fewer "pure" bazaar OSS programs that makes me think of lack of coordination and focus, which OSS need not be. Though I support the merits of FOSS, I don't think the bazaar method works generally. (I'm sure there are exceptions, but they should be just that, exceptions, and not the rule.)
Member since:---
Admittedly I haven't read "The Cathedral and the Bazaar,"
You should. Pay close attention to what it points out. After a few days, read the comments (pro-and-con) that have been posted on it since it was released in 1997.
but when you say that there are fewer "pure" bazaar OSS programs that makes me think of lack of coordination and focus, which OSS need not be.
That's not actually true. I've been on many Cathedral-style projects for the last 20 years, and the ones that work the best are the ones that track closest to the Bizarr-style development model are the ones that are most sucessful. I'm talking closed - propriatory - under - lock - and - key projects...the kind you would expect would benifit most from formality but don't.
Though I support the merits of FOSS, I don't think the bazaar method works generally. (I'm sure there are exceptions, but they should be just that, exceptions, and not the rule.)
The Cathedral model is highly inefficient. It's only necessary for general planning and as long as you have people of good will working on the project and on the client's side, the Bizarr is a much better thing to emphasise.
I've done it both ways. I've pushed for ridgid Cathedral-style methods before. I was wrong.
This is not to say that Cathedral methods are useless...they aren't...though it is a sign of trouble on any project if you have to emphasise them over the Bizarr. Usually, there is fear and (sometimes) lawsuits involved if the Cathedral is over emphasised.
Member since:---
There's one thing I've noticed about F/OSS. A lot of it started under the "Cathedral" model and was then released. e.g. Netscape, Openoffice. Or was purchased by OSS e.g. Blender. A lot of it presently is being sustained by "Cathedral" companies. There however are fewer pure, widely used, "Bazaar" OSS programs.
Bizarr: Linux, Apache, Netscape's precursor Mosaic, Perl, Python, protocols of all types.
There's quite a bit on both sides. The benifits of OSS to both commercial companies who have closed-Cathedral-style products and any group (commercial or not) that joins in the Bizarr for a specific project are substantial.
That said, purity isn't an issue, it's an ideology to any of these groups. The differences between the two methodologies can show how much utility you can expect out of any one project.
Note that I'm a big-time supporter of CMM/CMMI and ISO certification; as Cathedral as you can get. Even though I do support these methods, I have dropped and encourage dropping many of the formalisms that tend to creap in when doing Cathedral software development. Not all of them, just most of them.
The Bizarr methods will always outflank the Cathedral, though if properly used the Cathedral can add structure and certianty when dealing with uninterested and unprofessional developers. It takes a heavy hand on the development team though; something that doesn't inspire anyone to do the best with what they have; something that wouldn't work for projects where people are naturally curious and intersted in doing the right thing.
Member since:2005-08-06
Interesting that I finally see a comment that I completely agree with about the state of things.
Infrastructure is depending a lot on open source, as it helps companies build products on top which gives value. I mean check both IBM and Sun are interested in donating to anything which could be called "infrastructure" while non infrastructure is what they charge for.
This is what it all boils down to, and also why I'm so pro BSD/MIT/PD software. This let's you collaborate around the infrastructure which is in everyones interest and for specific add-ons you can pick and choose from infrastructure and don't have to reinvent the wheel.
What I do think will happen in the upcoming decade though is that things will change a bit in the following areas.
1. Organisations around certain niche industries will Co-develop stuff which will be free for everyone to make it easier for the industry. Example here would be like restaurant business making "menu/order system" or something which everyone gains off.
2. Public Sector will switch to some sort of OSS solution. Tax payers must demand that what ever get's produced by their money (the companies+consumers money) should be available for everyone. I would sincerely hope it get's put into Public Domain where it belongs. Everyone has contributed to it, so everyone should do whatever they want with it. This also the sole purpose of the government in the first place. Taking Tax money in order to serve the publics general interest. The general interest here can only be public domain as far as I'm concerned.
Member since:2005-08-09
2) infrastructure is easier than innovation
I think this is a common mistake. In many "commercial" applications, "innovation" means a different color of buttons, or semi-transparent widgets, or automatic forced installation of newer versions, etc.
There can be just as much innovation "under the hood" in infrastructure, when an O(N) algorithm is replaced with an O(log N) one. Perhaps not as sexy or apparent, but innovative and important nonetheless.
Member since:2005-07-07
There can be just as much innovation "under the hood" in infrastructure
Very good point!
For example I consider using packet switched networks an important innovation over physical point-to-point connections.
Or being able to run more than one process on an operating system.
But maybe your parent poster is still using DOS and a null-modem cable.
Member since:---
"2) infrastructure is easier than innovation
I think this is a common mistake. In many "commercial" applications, "innovation" means a different color of buttons, or semi-transparent widgets, or automatic forced installation of newer versions, etc. "
infrastructure is usually an implementation of a spec (tomcat, apache, axis, mysql). There might be some innovation, but it's not exactly momentous to serve up a web service 10% faster.
innovation is doing something different, clever, useful, with large impact. BAM (business activity monitoring) was innovative. There aren't any open source BAM solutions , it all started from a proprietary source. iTunes/iPod is innovative. The integration between the ipod and itunes and the online store was done elegantly and seamlessly, something that hadn't been done before. The UI on all 3 are brilliant. Again proprietary. Salesforce.com is another good example - take CRM and make it a service. The RSA algorithm was innovative, it certainly revolutionized the encryption world.
Innovative doesn't been completely new, it can be based on something existing, just a new spin on it. And OSS/free software can certainly be innovative (rails is)... but I really don't think that OSS is the major source of innovation in software.
Member since:---
Tomcat, Apache, MySQL are NOT just an implementation of a spec, if they are.
And yes, there is a lot of innovations in these and it is momentous.
But you would have me believe that there was no innovation in Apache since its creation ? Or in MySQL ? You're full of it. Especially if we follow your definition of innovation (doing something different, clever, useful, with large impact). Sorry to make you realize that, but all the FOSS products you cited suit perfectly in this definition.
You say BAM is innovative ? I say CRM : it all started Open Source. You say iTunes ? I say PDR and everything related (TiVo for example, KISS, ...). The UI on all 3 are brilliant. Again Open Source.
And you will have to explain to me how Salesforce.com is an example of proprietary if it just sales services on Open Source ??? This is just what FOSS based companies do. The RSA algorithm did not revolutionize anything, as it was closed-source and very expensive, so few people could afford it. The free PKI revolutionized the world instead, and innovated to bypass the limitations of RSA.
FOSS is very innovative, but as you don't have to pay for it usually, you fail to realize it. The Internet you are using is a huge revolution allowed solely by FOSS, you are completely depended on it. And I would call the Internet a major innovation and an infrastructure.
You still live in the rather new proprietary software world where value is defined by the money the software cost, but you don't realize it.
Member since:---
[quote]It seems that OSS/free software has been most successful for infrastructure (Apache server, tomcat, linux, tapestry, axis, rails, subversion, mysql). Note that all these products are copies of exiting proprietary products. [/quote]
Those are all excellent examples of where open source works best, where developers create tools to do work for them. Often this model of development leads to a higher quality of product, since the end user is the developer, they get exactly what they want. As opposed to the, some guy sitting in a cubicle waiting for Friday, development model of commercial software.
Under this notion, I would tend to agree with you that the commercial counterparts to the afformentioned OSS products are 'exiting' products. Most of the commercial offerings of these tool tend to be inferior (IIS to Apache) and are 'exiting' the realm of relevancy.
-MattK
Member since:
2005-06-29
The author complains about the difficulties of making money on the FOSS model, which are real. But one of the largest benefactors of the FOSS model is in the world where no money is made off of software.
Why does *everyone* forget the non-shrinkwrap developers?
Now, this could actually put people out of work: High quality in house programs shared among people doing similar things; we might just reinvent the wheel less. But with Bill Gates saying there aren't enough good programmers out there...
And in the shrink wrap market FOSS does something very nice: It provides competition for the old dogs who are just making it on lack of competition. I think the competition from Linux has really helped make Microsoft get its butt in gear on security issues.
Photoshop has to keep innovating if they don't want The Gimp to be stepping on their heels.
Member since:---
Agreed.
Surely FOSS is ideal for companies that provide web-services or computing resources as a service.
They will have experts on their team anyway so it doesn't matter if the software isn't so simple to administer. It just has to work properly, scale well, and have minimal unit license pricing. Oh, and interface/integrate satisfactorily with other software.
Just thinking about it, that combination sounds innovative and in my experience FOSS does it as well as anything else around.
Anyone have other specific or generic examples of FOSS innovations?
eg.
generic: Significant portions of the Javascript world
specific: Dean Edwards IE7 -> http://dean.edwards.name/
Member since:
---
I use both open source and proprietary software, and I am pretty happy. OSS seems to make great server software and desktop enviroments, but when it comes to apps? Well, at least you can say it helps keep the commercial guys a little more honest, so I'm glad it's around 
But really, does it have to be a religion?
Member since:
---
I haven't seen a lot of good reasons in Shawn Shell's article. But of course is proprietary software here to stay. That's not the interesting question.
The interesting question is: As a producer, will your proprietary software be able to compete with Microsoft when it comes knocking at your door?
As a consumer, will you be able to pay twice the license fee or wait twice as long for updates when Microsoft decides to do so?
This is what Shawn Shell failed to understand: For infrastructure, there's no other way than Open Source if you're not interested to depend on and compete with a monopoly.
Member since:---
Member since:---
Shawn's article points out that software which has reached a commodity stage is a good candidate for open source; he admits this. However, I think the good reasons are as follows:
1) Developers contributing code don't get paid for their work. Unless they work for a company like RedHat, they're out of luck.
2) The average user (including big companies) are prepared to self support, so they end up paying someone else (why not just buy commercial/non-open source)
3) Beyond the OS, there are few productivity applications that are accessible by the large population of non-technical users.
4) If people were really as upset as the open source community contends about Microsoft, they wouldn't continue to feed the "monster." Dell offered Linux on the desktop and it didn't sell...
Member since:
---
"Does Linus or any other developer contributing to the Linux kernel directly benefit from his labors?"
Linus is definitely not starving.
Where I disagree is that developers are Prima Donnas who should become filthy rich from their coding.
This is a delusional hope too many developers suffer from, thanks to Microsoft and Bill Gates.
In practice, how often do you see this happening?
And why can't developers work as ordinary employees? (albeit good paid ones)
Member since:
2005-06-29
Member since:
---
The author of the first article actually sounded rather intelligent, albeit limited in their understanding of what drives open source (ie. it isn't all about money). But then the author had to go and say something utterly stupid in the context of open source:
There is no guarantee, warranty, or statement of suitability for any application whatsoever.
Very true. Unfortunately, the same is also very true of commercial software.
Member since:
---
In the opening paragraphs the author demonstrates that he has no good working definition for "open source development" or "commercial software development": "Open source development is, effectively, many unrelated or related [so, whatever] developers collaboratively producing software for the free distribution, improvement, and use of anyone willing to abide by the software license [there is non-free software which is free to distribute as long as you abide by the license, such as shareware] — such as [note that the GPL is not part of the definition, it's an example] the GNU General Public License (GPL) or Lesser General Public License (LGPL)."
"Commercial software, by contrast, is developed by programmers employed by one company (or in partnership with many companies) for the purpose of creating a package that will be sold and whose code is protected through various intellectual property mechanisms..." No, many commercial systems are free as in beer and companies make money on support or "adjacent components".
"How is a free model of software development sustainable?" It doesn't have to be. Some guys go to the pub after work, some go fishing on the weekends, some write software and give it away. That's also the reason most of it is not very polished.
"With the lack of a cohesive and coordinated development and support mechanism, why would anyone seriously consider using open source software to run a business?" I work for a large corporation programming CAD software, and we're definitely not coordinated etc. In fact, the bigger the company, the more bureaucracy your request has to get through to get to a developer. If it's a small open source project, you can just email him.
"Does Linus or any other developer contributing to the Linux kernel directly benefit from his labors? No."
The whole Linus example is badly chosen, because using his fame this particular individual got excellent jobs and probably makes good money.
"While the license that governs a lot of open source software—the GPL—does provide for the idea of making money on "distributing software," it doesn't allow developers to receive direct compensation for the act of development." At this point it's getting idiotic. Why on earth does the GPL not allow the developer to get paid?!
His basic point in the next paragraph is that if company A pays its developers, releases software S for free, makes money on support then it will be at a disadvantage when compared to company B which simply makes money off support on software S. This argument has many implicit assumptions, but in my opinion, if these assumptions are satisfied, and the case above arises, then open sourcing the software is simply not what A should do. However, that doesn't mean it can't open the source at a later point, when access to the source code is of no bussiness value anymore. Example: id software.
"However, when a piece of software written by a commercial vendor breaks, you have a vendor on the hook to correct the problem—even Microsoft writes service packs and hot fixes for its products." I think this argument simply does not stand in the real world. Commercial licenses usually have clauses saying that there is no warranty what-so-ever, so in terms of guarantees you're not better off than with open source software. In my experience both non-commercial and commercial projects respond to major bugs by fixing them. And independent developers working on small projects are probably better at fixing small bugs quickly and releasing a new version the next day.
"What the open source software really gives you is a bunch of software components that may or may not work collectively to solve a particular business problem—no guarantees. If you don't like it or want it to work differently, fix it yourself." I could argue that < large software house > is actually worse, since if you don't like it, you can't even fix it.
Member since:
---
"And why can't developers work as ordinary employees?"
This is excactly the point. What we need is big software houses that is just that - a factory that produces software.
Software only have to be written once, so we have to change the economics. How much does it cost to make X software? Put it to litigation. The best bid from a softwarehouse wins. They have programmers which are paid (me 
.
Support when the product is finnished. Guarantees that it works.
When the product is delivered, a new ligitation can be made for extending that product.
The upshot is that the software price is just how many developer hours it costs. This is the reason why Open Source will overtake properitary software. You only pay once for your software.
So repeat after med: Software should be made just like a house is being made. First you need to find out what software you need. Then you need to find others who need it. Then you have to find software companies who can bid on the implementation.
The reason for going through this? The cheapest price.
If you are not carefully when buying software, you can end up paying what you can pay, and not what it costs to develope.
If you wan't to buy properitary software that is OK. You just have to look at the formats. Are you storing data that you need to retrieve again? Then you need an open format. Are you communicating with others, then you need an open format. Else the guy in the other end, needs to use the same software vendor as you (yes, Skype is a messaure of the stupidity in the world).
So when you are not storing or communicating using the software, it can be as properitary as you like.
So to sum up the rules:
1. Never buy software which writes to a closed or patented format.
2. Never buy properitary software which communicates with other in a closed or patented protocol.
3. If you know what the software should do, it is a good idea to create it as open source software. It will be cheaper in the long run, since you then can get everybody to enhance the product, and not just the company which has supplied the original product.
Member since:
2005-09-08
Member since:
2005-10-02
is extremely flawed. I haven't finished reading the second, so I won't post all my thoughts on the issue yet. It will have to wait since I also got a meeting to attend in a couple of hours.
But the first author don't seem to grok the difference between open source and closed source. But mind you. It's just a first view impression.
Member since:
---
one guy on here was saying you should never store/send data in a proprietary or patented format.
I gotta ask: why? why would I care what format it's stored in? If I want to serve up a flash app (perhaps with laszlo)... what's wrong with that?
flash works with 97% of browsers. The format is well understood. So what's the downside?
Same with PDF.
Member since:2005-07-06
"one guy on here was saying you should never store/send data in a proprietary or patented format."
Yes and this guy whas right , you would not believe the crap in windows format and proprietary format from software thats is not accessible anymore that exist in business ( where talking software of 15 years ).
"I gotta ask: why? why would I care what format it's stored in?"
Because you whant access at your data at all time from all solutions. After all its your data.
Flash format is Open source , Pdf format also.
So basically your asking why Open Source software work when your giving an example yourself ...
Member since:---
Member since:
---
Member since:2005-07-06
"Open source attracts everyone"
No , its Open to everyone. Its not the same thing. In Open Source the bad programming is flushed out by evolution , revision and examination.
"who don't necessary know how to code ( or behave ) properly."
You clearly havent seen the closed source people ina ction to say such a statement.
"it's the educational system and payed professionals that guarantee some level of quality."
Educationnal system work with GNU/Linux , BSD and the payed professionnal dont have access to the source code at all time , it explain why so many real flaw in Windows are exploited vs more flaw found in Open Source but wich is found before exploitation in real case.
Member since:---
"Open source attracts everyone, even people who don't necessary know how to code ( or behave ) properly."
Do-it-yourself, home lobotomy kits.
"With closed source, it's the educational system and payed professionals that guarantee some level of quality."
Engineers, Doctors, etc.
Member since:2005-07-24
The free code will get better and better (1000 eyes) and if all else fails you can make it better yourself.
The closed+"in it for the money" code however, even if designed and coded carefully will eventually have a deadline and that's when it usually starts to get ugly. Also after the product is sold the drive to make it better is not exactly great (especially so if you plan on releasing a V2.0 eventually).
Member since:
2005-07-06
Both OSS and closed/nonfree/properitary software are here to stay. neither is going away anytime soon.
And I think, both have there place in the grand scheme of things. altho, I think open standards for things like: file formats, programming langages, etc is a very good thing.
-Nex6
Member since:
2005-07-15
Shawn Shell starts with:
"I've read a number of articles on the open source development model for software.
And exactly that's the problem. He doesn't really talk about free software vs. proprietary software. The real topic is community development vs. commercial development. And i agree, we need commercial development. 95% of the software industrie is about developing special software for the costumer. That's something no community will do. But developing in a community or in a company has nothing to do with the question about the license.
In reality most of the 95% "special-software-market" is already a free software market. Most of the costumers wouldn't pay the huge price if they wouldn't get the source and the right to do with the source what they want.
Member since:
---
How do we know that patents are actually doing anything usefull at all? Yes, there is this nice sounding theory about limited monopolies and money driven motivation and all, but I don't think I've actually seen any raw numbers to back it up.
Did the rate of advancement actually increase after patents where introduced? What about the well being of inventors? How about the public? Is it just administration that is getting fat? We really should have some stats before we go crazy on the system.
Ironically, I recall reading an article recently about the medical innovations, and it seems that there has been a significant decline in productivity since heavy patenting started. Another big hit against the system is the current African situation.
Member since:2005-08-06
Without having any in front of me, I think as you say the medical stuff is indeed a place where numbers do exist (sort of atleast).
Medical reseach is based on a 1/10 concept I believe. This means out of 10 products research are conducted about, 9 get thrown away and 1 works out and get materialized on the market.
The ROI rate of the product is often many many years. We also know that profit declines quickly after patent time has gone out (Sweden has short patent times on medical stuff). So it's harder to make profit on it. The result has been that less research are being done in small diseases (as you can really not calculate any profit here) and only big disases get targeted with research, such as Depression, Cancer and HIV.
