Open-Source Software Development: Doomed From the Start?

[Jason Dunn]

<div class='os_post_top_link'><a href='http://www.pbs.org/cringely/pulpit/pulpit20030424.html' target='_blank'>http://www.pbs.org/cringely/pulpit/...it20030424.html</a><br /><br /></div>"Here is the core argument: There are a thousand Open Source projects that get started out of need or fun, are maintained for awhile for fame, then get abandoned because there is no reason to go on. Eventually, the programmers come to understand that "users" are people who yell at you to fix stuff. So Open Source is inherently flawed. It only works because otherwise unknown programmers can get 15 minutes of fame using the Internet as low-barrier entry into introducing their skill to the world. Since they are introverted nobodies, getting a few emails from unknown users that say "good job!" feels great. But in time, most Open Source projects grind to a halt. The ones that survive are projects like Linux and Apache that have substantial involvement by PAID engineers. One could argue, in fact, that the idea of Open Source software being created by volunteers is a misnomer. Even Linus Torvalds is paid by Transmeta to be the God of Linux."<br /><br />Robert X. Cringely's weekly columns are one of the few "must read" Web columns that I keep track of. This guy is just such a monster thinker - I love watching the gears in his head turn. This is a bit off-topic, but it has a lot of parallels in the Pocket PC software world as well. Give his column a read and tell me what you think (that SIDS info sure is sad...).

[Janak Parekh]

I'll have to say that I don't entirely agree. Just like there are businesses that fail, there are projects that don't succeed. And a lot of people don't realize the breadth of successful open-source software out there.

For instance, my personal server runs Debian Linux (www.debian.org), which uses the Exim mail server (www.exim.org), Apache, the Linux kernel, SpamAssassin (www.spamassassin.org), the Mutt (www.mutt.org) email client, and a bunch of other tools. Most of these aren't corporate-sponsored, have existed for a long time, and work very well for their purpose. And it's all free, which is awesome. (On the other hand, few, if any of them are user-friendly -- they're still geared towards the UNIX geek like me).

Oh, perhaps the best resource out there to see this breadth, btw, is Freshmeat (http://freshmeat.net/). While the number of projects that don't succeed is significant, it's absolutely staggering to see the number of projects that do.

--janak

[dh]

It's heartbreaking that babies still die from SIDS in this age of medical advances. I'm a father of three and I can't imagine being able to handle that situation.

On the Open Source issue, Cringley could well be right to a large extent.

However, in an industry where a few large companies dominate, OS has given us some serious steps forward.

Look at the progress that Linux has made, especially in the server part of the business. Yes, that technology has indeed been hijacked, first by companies like Red Hat and later by IBM. It still provides an alternative to the status quo that might not have evolved otherwise, and has made MS and the other server companies work hard to improve their own products.

In our own PPC world, what is the best media encoding and playing system? Project Mayo and their products.

Commercial interests are always going to win, but Open Source provides great stuff, maybe not for the mainstream but certainly for the enthusiast.

[paris]

well i totally agree with Cringely. Am doing a project based on CMS and open source software and in my summary i will say extacly what Cringely said.

How can someone built the perfect product just by free help? lot of research is needed and a lot of resources which means a lot of money and alot of money do not exist in open source projects. Why do you think linux is far below Windows? just because if they could sell it they would have money to do more developement and reserch on it.

I am a programmer and i prefer to work and get paid, not work just for the joy of builting someting.

[CR]

While there is some truth to that, there is also a great example of it being wrong here in our very own little world of PPC software. PocketMVP. PocketMVP has been in development for well over a year now. Granted it hasn't been under the same developers that started it, but it is still going strong. That's the beauty of open source: if the original author loses interest ( or time ), and there is still anyone who wants the software developed further, chances are someone else will pick it back up. It doesn't just quietly rot away in some obscure folder on someones hard drive. The demand keeps the development going.

Of course, open source development may slow to a crawl, when compared to proprietary commercial development. But that is the nature of the beast. You can't eat if you're not getting paid, and no one wants to pay you for something that everyone else can have.

[dh]

While there is some truth to that, there is also a great example of it being wrong here in our very own little world of PPC software. PocketMVP.

Yep, PocketMVP is far better than all the other media players, including the one from Microsoft.

[BarryB]

I'm not entirely convinced of his argument, but I think he makes a number of valid points. His analysis of the situation with Apache does provide support for his argument.

Coming from the Palm world, where it seems there are over a dozen Bejewelled clones, a hundred calculators, and ten or twenty secure information repository applications, I see many of the same trends on freshmeat. I notice numerous applications that all seem to involve the same things. Now, this may be a function of my newbieness to Linux, but many applications seem to be building on established programs like sendmail or samba and adding their own little tweaks.

One thing that would lend support to Cringley's argument would be a study profiling those who are actively involved with open source (sourceforge could compile data about those who submit projects) and how long they stay involved with both the project they started and open source in general. I suspect that most are high school and college students who do not stay involved too long after graduation, but that's just my hypothesis.

[gdaswani]

well i totally agree with Cringely. Am doing a project based on CMS and open source software and in my summary i will say extacly what Cringely said.

How can someone built the perfect product just by free help? lot of research is needed and a lot of resources which means a lot of money and alot of money do not exist in open source projects. Why do you think linux is far below Windows? just because if they could sell it they would have money to do more developement and reserch on it.

I am a programmer and i prefer to work and get paid, not work just for the joy of builting someting.

Opensource software will always be around, one factor of that is the IS barrier of entry (large companies can afford to spend money on software, small ones like startups cannot) hence there will always be some sort of development (a process which can be contrasted to convection) going on, and will continue to do so as real systems are being built (and needs to be maintained) with OS as the infrastructure. The same business needs will ultimately funnel back $$$ or manpower (in terms of development time) to these projects for they rely on it. GPL is a good thing IMHO as it pretty much gaurantees that improvements made by 3rd parties are disclosed and given back to the source tree.

We can see some evidence of this - just look at the number of smaller players (even consultants) helping out with various OpenSource projects (Linux for one)..

Paris, your developer mindset is totally different to one of a business person (who ultimately hires developers and gives them jobs) - especially to one who wants to get up and running efficiently in a very competitive market place (it's competitive now, just imagine in the future).

OpenSource is the great IS barrier of entry equalizer - people will use it, and people will enhance it due to business need. Simple as that.

[Andrew Duffy]

I'm a professional software developer, working in the network management field. There are mulitple open source competitors for all of our products, but the reason we can sell software is that we invest huge amounts of time (and by extension money) into making software that is easy to configure and use. The only open source project I've seen that comes even remotely close to being easy to configure and use is Lycoris, which is a Windows clone.
On the other hand, our software makes use of open source libraries, which have changed that aspect of software development drastically. Whereas once you had to pay a large per-developer licence and a royalty for a buggy, poorly documented library you now pay nothing for a buggy, poorly documented library that you can fix, and you can pacify your conscience somewhat by contributing back any changes. The difference here is that libraries don't really have to be easy to use; they just have to do what you need.

[bagemk]

Cringely makes a big mistake in his column. He talks about Microsoft and IBM and implicitly assumes that their businesses are comparable. This is not the case: IBM is and always has been, primarily a hardware company. They make software with the purpose of selling more hardware. (The same can be said of Sun or Apple)

Microsoft write software to sell software.

From this simple fact alone, it is clear that Open Source is a gift to the likes of IBM. They get some code for free - they get the benefit of some cheap developers helping them out from all the around the world (perhaps unwittingly, but they are helping). There are no licence costs involved and support costs are reduced. It all helps reduce the costs of producing hardware boxes and increasing their profits.

Microsoft is a software developer. They write/licence all their own code. They dare not go near any Open-Source code because of the licence conditions whereby it extends the Open aspect to any linked/derivative works.

This is the basis of the real battle between open and closed source. It also explains why OS projects that benefit only a few people or that do not feature in any shipped hardware die, while others are encouraged.

Will this situation change in the near future? Perhaps. There are a growing number of companies looking at lifting code from the Open Source community and putting that into embedded hardware. I personally know of several of these and few of them admit to the Open Source basis of their code. However, all of these companies are active on the newsgroups encouraging the development of the code in which they are interested. This creates a very powerful, but hidden force against Microsoft. MS is competing by offering technology (Windows Media 9 for example) and trying to cover all fronts. But Microsoft's biggest strength is their patience (PPCs and phones are perfect examples) - they will keep pushing and trying until they win.

Maybe there doesn't need to be a winner. If there does, I wouldn't like to say who it will be.

[jmarkevich]

I haven't read the article yet...

But here I sit, surfing with five tabs in MOZILLA, using my DEBIAN do-everything server (P166-32 MB), collecting mail through FETCHMAIL and dispersing it through EXIM. I'm probably missing quite a few components...

I consider these to be the best of their breed. Hardly a failure!!

It's very easy to start an open source project, and maybe even easier to let it all drop. There's lots of dead projects on sourceforge, but there are also lots of dead projects behind closed doors at commercial software houses too.

This article is pure flamebait. Replace "Open Source" in the "Here is the core argument:" paragraph with "Software Development" and you'll see what I mean.

It's a shame he wrote such a limp antagonistic article in the same column as a sad one about his son. It's almost using that tragedy as a lever to make you believe the next issues.

[BarryB]

I haven't read the article yet...


Yet, in spite of not having read the article, you feel qualified to comment on it? Sheesh. If I had said something like that on here, I'd be flamed to ashes. :)


It's very easy to start an open source project, and maybe even easier to let it all drop. There's lots of dead projects on sourceforge, but there are also lots of dead projects behind closed doors at commercial software houses too.


Yes, but the question is not merely whether dead projects exist, but whether software produced through open source is more likely to die than software produced through commercial methods.


This article is pure flamebait.


How do you know? You haven't read it.

[Janak Parekh]

Yes, but the question is not merely whether dead projects exist, but whether software produced through open source is more likely to die than software produced through commercial methods.
I think there's no easy way to be able to come up with a straight answer to this, as dead commercial projects rarely make it to the forefront (and don't think these don't exist; I've read historical statistics that perhaps 50% of commercial software projects are failures and end up on the cutting-room floor). Dead open-source projects are clearly visible, on the other hand.

--janak

[/dev/niall]

The idea that only open source projects are prone to subversion by commercial interests is just plain silly. How many times have we seen mergers or acquisitions that result in a larger competitor buying a smaller one "for their technology"? Should every small company stop what they are doing because Microsoft might buy them just to get them out of the way?

In fact, open source offers an advantage in this regard. When commercial software is acquired by another company it's often the last time anyone hears of it. If it's open source and a viable product, the opportunity to continue development still exists.

Could it be that (shock, horror!) there are advantages and disadvantages to both open and closed source development!?

[trepine]

I'm a professional software developer, working in the network management field. There are mulitple open source competitors for all of our products, but the reason we can sell software is that we invest huge amounts of time (and by extension money) into making software that is easy to configure and use. The only open source project I've seen that comes even remotely close to being easy to configure and use is Lycoris, which is a Windows clone.


I too am a developer, albeit not fulltime, and use an open source framework to build proprietary products on top of. However the beginning of your argument I must respectfully disagree with.

It is not that linux is truely harder to use/manage, but that windows is more mainstream. This is the same lame argument used why palm is better than PPC, just because more people were familiar with palm than PPC makes is 'easier' and 'simpler' to use. I have 7 computers at home, linux workstation, linux server, windows 98, windows XP, iBook w/ OSX (unix rocks), a win2k server, and my zaurus (though technically a handheld, I consider it a computer since I can run apache, php, and a gob of other high end programs). The easiest by far to use is my RedHat 9 workstation. There is even a network samba share browser now. No more smbclient or smbmount. My wife, who has traditionally been windows only, now prefers redhat 9 over winXP. One simple reason, stability. This is largely because IE locks up quite frequently, and destroys her shopping session (I know mozilla for win, but that kinda puts a crimp in the opensource dying thing so I will leave that alone). I have introduced about 15 people to linux, of which about 10 still use it on a >= 50% basis.

Linux is not dependant on, but does benefit from corporate sponsorship. However, the real beauty is that as corporate sponsors come and go, the big important projects keep on cooking. Do you think if IBM left that apache would suddenly go away....HRM...No!!!

However, if IBM went away, what would you do with all your IBM servers if there was some vulnerability that was discovered and needed a patch?

[Janak Parekh]

It is not that linux is truely harder to use/manage, but that windows is more mainstream.
Of course, no argument is simple. A properly configured Linux box is easy to use, as is a properly configured Windows box. That said, most people nowadays don't know how to operate a command-line interface, nor edit configuration files; and most Linux distributions require you to drop down to this level for anything but usage or simple configuration tasks. Yes, RedHat 8 and 9 have gotten better, but it's taken years, and I still frequently go to the command-line to do tasks (for example, replacing the built-in Mozilla with the xft-linked version that supports antialiasing). Windows, on the other hand, is frequently preinstalled on computers, and the OOBE (out of box experience) is simpler for someone who only knows point-and-click.

--janak

[BarryB]

Yes, but the question is not merely whether dead projects exist, but whether software produced through open source is more likely to die than software produced through commercial methods.
I think there's no easy way to be able to come up with a straight answer to this, as dead commercial projects rarely make it to the forefront (and don't think these don't exist; I've read historical statistics that perhaps 50% of commercial software projects are failures and end up on the cutting-room floor). Dead open-source projects are clearly visible, on the other hand.

--janak

I agree. I don't think companies would willingly produce figures showing how many of their projects die. In the absence of such figures, all we can do is speculate and render opinions. So I guess any of ours are as good a Cringley's. :)

[/dev/niall]

Though I often disagree with RC, at least he has a clue. From an article (http://story.news.yahoo.com/news?tmpl=story&cid=581&ncid=581&e=1&u=/nm/20030424/tc_nm/tech_microsoft_server_dc) announcing the release of Windows Server 2003:


In response to the threat from Linux, Microsoft commissioned a study last year that showed the previous version of the Windows server software was cheaper to run when considering the total costs involved in the upkeep of Linux systems versus buying off-the-shelf software from Microsoft.


Linux disputed the findings.

Er, who disputed? lol. Makes it sound as if Linux was some sort of hive entity like the Borg.

We are the Linux. TCO as you know it is over. We will add your features and user experience distinctiveness to our own. Resistance is futile.

[laner]

I've used a mix of open and closed source for years. The web has really what made open source viable and open source made the web viable. I've been playing with linux since the redhat 5.1 days, so that would be either 1997 or 1998. Sure I could have used another distro, but it was the best balance of time, money and usablity.

Sure some corporations have helped fund open source development. It doesn't mean they own the products, they can help move the project forward in the direction they would like though. If a developer doesn't like the what's going on he can leave and find other employment. The source is still available to him. If the company goes under the source is still available if the developer wished to continue working on the project.

An example of how I've used open source is when Microsoft hosed the ftp code in ppc2k2. I took Kenney Goers library that's posted on codeprojects.com and fixed a couple of bugs and added activex interface to it. I had purchased the one library on the market that didn't rely on Microsofts built in routines, but it turned out to be buggier then hell and the vendor wouldn't fix the problems or even admit they existed. They wouldn't even refund the close to $1000.

If open source is doomed is this web site doomed? Take a look at the http headers for pocketpcthoughts..

HTTP/1.1 200 OK
Date: Fri, 25 Apr 2003 14:01:52 GMT
Server: Apache/1.3.22 (Unix) (Red-Hat/Linux) mod_jk/1.2.0 mod_perl/1.24_01 PHP/4.1.1 FrontPage/5.0.2 mod_ssl/2.8.5 OpenSSL/0.9.6b
X-Powered-By: PHP/4.1.1
X-Accelerated-By: PHPA/1.3.3r1


How many open source project can you count... Apache, linux, php, perl, jakarta, openssl. Don't forget phpbb. Since it's showing as redhat linux you could assume a lot of other open source projects are involved.

PHPA is free but not open source.

[jmarkevich]

I haven't read the article yet...


Yet, in spite of not having read the article, you feel qualified to comment on it? Sheesh. If I had said something like that on here, I'd be flamed to ashes. :)


I wasn't though. Interesting, eh? :) You didn't notice that I had five tabs open in Mozilla? I started the comment while reading. I would have thought that apparent when you read the rest of my highly qualified comment.



This article is pure flamebait.


How do you know? You haven't read it.

Ok, go back to my post, and get out your white-out. :lol:

[Kirkaiya]

well i totally agree with Cringely. Am doing a project based on CMS and open source software and in my summary i will say extacly what Cringely said.

How can someone built the perfect product just by free help? lot of research is needed and a lot of resources which means a lot of money and alot of money do not exist in open source projects. Why do you think linux is far below Windows? just because if they could sell it they would have money to do more developement and reserch on it.

I am a programmer and i prefer to work and get paid, not work just for the joy of builting someting.

I think that people here are, in fact, misrepresenting what "Open Source" means. People often associate open-source with "free", but that is NOT part of the definition of Open Source.

Open Source means exactly what it says: the source code is open (that is, you can obtain the source code to the product, and thereby modify it yourself).

There are several different licensing schemes some people use to sell (for money - as in, GET PAID) for open-source software. Some of those licenses require end-users who make changes to the source-code that get commercially used to contribute that source-code back to the originator of the code. One common license is the GPL (GNU Public License).

However, there's nothing to stop somebody from selling you a piece of software, and they provide you with both the compiled (ie., executable) code plus the source-code, under a different licensing model.

So - my point is just that, Open Source does not nessarily mean free, and of course many open-source programmers get paid. On the other hand - some of the best coding I ever did was probably in 8-bit assembly for the 6502 chip in my Commodore 64, which I wrote part as part of a of a BBS system I created (which I ran for 2 years) - and I gave the software away to friends, who also ran BBS's on it.

I had never heard the word "open source", and back then, a lot of software was free anyway, and I *did* get phone-calls from friends using it who found errors, and fixing them was a lot more fun than doing my 10th grade homework, as I recall... Don't forget that a lot of software gets written by people with a love for solving problems and programming, it is not, in fact, all about the money.

So I think the author of that article is actually pretty wrong - there are many many sucessful open-source projects, and to assume the model is inherently flawed, when most of the web runs on FreeBSD, Linux, Apache, Perl, sendMail, etc., is a bit of a flawed argument. Oh - I *have* to plug the awesome SpamAssassin - i have a free implementation called SAProxy, which filters 90% of my spam out)

[Kirkaiya]

&lt;snip>
This is largely because IE locks up quite frequently, and destroys her shopping session (I know mozilla for win, but that kinda puts a crimp in the opensource dying thing so I will leave that alone).

Hey, now - I use Mozille 1.3 as my primary browser now, and have for some time (once you set it to load components during startup, there's no annoying lag time to start the browser anytime you launch it).

It's more stable than IE on my laptop, runnin Win2K, and I think is hardly indicative of a dying Open Source project - it's actually advanced quite a lot, and surpassed the heavily-monied Internet Explorer in terms of functionality, in many ways. I think people just had unrealistic expecations when it first went "open source", which led to rumors of it's demise, when, in fact, it's updated very frequently, and has NEVER infested my machine with a malicious worm (which IE has done)

Otherwise - I agree with you fine post!! :lol:

[Tienshiao]

OK.. the second part of his argument is the possibility of hijacking a project by some evil Microsoft backed group.

Now most (large) projects have maintainers who are manage the contributions and in some cases are the largest contributers. I don't think it is likely for somebody to submit new code that takes the project in an undesirable direction, and the maintainers accepting it.

Fine.. you payoff the maintainers. The other core developers of the group who disagree with the decision can fork the code (see the *BSDs). If the project heads in an undesirable direction, those that find it undesirable can fork the openly available code base.

[Coordinator]

The logic behind original author's argument could even seem compelling if only it didn't totally contradict hard facts on the ground. Open source projects are thriving, huge amount of work is being done, many projects have completely thrashed their closed source competition, and if anything the tide is rising.

Now, why this is happening is indeed hard to explain. But trying to explain real facts is more useful then inventing logical explanations for imaginary situations like "demise of open source software".

[BarryB]

The logic behind original author's argument could even seem compelling if only it didn't totally contradict hard facts on the ground. Open source projects are thriving, huge amount of work is being done, many projects have completely thrashed their closed source competition, and if anything the tide is rising.

I'm not qualified to offer an opinion on the rising tide of open source software development.

But the validity of the article's hypothesis does not hinge on how many projects are ongoing or how many new ones are starting up. The validity hinges on whether open source projects are more likely to die than commerical ones. The situation looks something like this:

Is Open_source(dead)/(open_source(live) + open_source(dead)) > Commerical(dead)/(Commerical(live) + Commerical(dead))?

As was pointed out above, it's unlikely that this question will ever be answered due to the tendency of failed commerical projects to be buried.

Granted, I'm avoiding the whole question of what constitutes a "dead project." But I stand by my argument that the hypothesis of Cringly's article is not whether there thousands of live open source projects or more open source projects than commerical. His hypothesis has nothing to do with live open source projects or live commercial projects (except as part of a total number of projects). His hypothesis has to do with dead ones.