« Information wants to be free (as in speech) | Main | Moronic online business practices number 678 in a seemingly infinite series »

April 13, 2005

Lessons Proprietary Software Can Teach Open Source

[Rereading this just before publishing, I became aware that it is full of generalizations. And that it could be read as a strong criticism of Free/Open Source Software (FOSS). There are generalizations certainly. But it is not meant as a criticism of FOSS. Rather an attempt to answer the question above, and think a little more about the differences between what I do (develop commercial software products) and what many far better coders do (write open source software).]

At Slashdot today someone asked what lessons proprietary software can teach OSS.

Now, despite being involved with commercial software, I don't have anything against OSS. I'd have to say that on first reflection I only use OSS peripherally, say Firefox a bit (Safari is still my main Mac browser, though on windows Firefox is my default browser). But of course many many sites I read are served by Apache, using PHP, on Linux or BSD boxes, and other open source solutions. Indeed westciv.com is served from a BSD based server, using Apache. In fact the more you look, the more you'll find OSS - Safari relies in part on the open source KHTML, and the underlying OS in Mac OS X has much to do with the open source BSD version on UNIX. Similarly in Windows, aspects of the operating system use OSS, as far as I am aware.

So open source software is a fundamental part of our computing experience, whether we choose to use an open source application or not.

What is interesting is that so few people, outside the technological "hard core" explicitly use an OSS application, even though we all use OSS. What's going on here? By answering that, I think we get an answer to the question at Slashdot.

See, people don't use software, they don't even use applications. They use products. Whether at a price, or for free, people are looking for a complete solution to a need - what Geoffrey Moore, hi tech marketing guru (author of "Crossing the Chasm") calls a "whole product".
What's telling about the Slashdot discussion is that the term "whole product" does not appear once in the several hundred responses to the article. Which does not necessarily surprise me. Engineers and coders often are quite hostile toward, and generally at least indifferent to, marketing.

OSS has produced many great pieces of software. Whole operating systems, web browsers, networking technologies. What it has yet to produce is many, if any, great "products".
What about Linux, or Firefox? These are certainly great technologies, and Firefox comes close to being a product (I'd even be happy to admit that it is), but certainly Linux is yet to be a "plug it in and turn it on and forget about it" solution like Mac OS X, or Windows. They are whole products. Some Linux distros are getting there, but in a way this just illustrates my point. What companies like Red Hat do is take the technology of Linux and turn it into a product. How? By adding support, by making it "just work".

So, to be shorter than usual, the lesson that OSS can learn from proprietary software is stop just building technologies and start building "whole products".

At westciv, like any remotely successful software company we do a lot more than just write code. We spend a lot of our energy (I'd say the significant majority) on our support forum, on documentation - our manuals and tutorials, on add ons like our CSS templates (available under a creative commons license). We directly support our users when they are having problems. Not to mention user testing, thinking and rethinking about our users experience and so on.
OSS projects, by and large, revolve around writing and testing code. And as a methodology OSS approaches produce an enormous amount of high quality code.
But as yet, it does not look like a great approach to building whole products.

My guess is that companies like Round Two, which Bart Decrem, longtime Firefox dude is associated with will find a business model, as Red Hat has done, in turning open source projects into whole products.
What the OSS community thinks of it, remains to be seen. My guess is that the Stallmanesque "hard core" may not necessarily be positive toward it, perhaps even hostile in parts. Which would be entirely to the detriment of the success of OSS.

But therein lies a difficult problem for OSS. Philosophically, there are those, like Richard Stallman who are opposed to the idea of software as property. And there are those, like Eric Raymond, who are much more pragmatic.

The very idea of a product is part of the culture of proprietary software, a culture toward which many "free software" advocates are quite hostile. Is it possible to adopt tactics and strategies from a culture toward which you have fundamental objections? Philosophically it would strike me as quite difficult. But strategically, it is vitally important, if FOSS projects want to go beyond their current roles as important parts of the plumbing, largely hidden out of the way, and to take center stage on the average user's desktop.
And that will need marketers, writers for documentation, designers who will make these apps look world class (appearance counts for a lot among many many software consumers). That will take user interface experts, all willing to work in the open source model. Firefox has managed something of this, so it is not impossible.
But, the free software philosophy and culture started in the hacker culture. Can it cross over to other communities of practice that are vital to building a whole product? That is FOSS challenge. And it is a significant one.

April 13, 2005 | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Lessons Proprietary Software Can Teach Open Source:


"The very idea of a product is part of the culture of proprietary software, a culture toward which many "free software" advocates are quite hostile"

Absolutely. There's a bad attitude abounding that "if you have to pay for it, it sucks". Coming from a Mac perspective I tend to think 'OSS sucks until proven otherwise"!

I would also say that Firefox is a product, and is a good example of what happens to something when a little 'outward looking marketing' is applied to it.

Posted by: Jon Hicks | Apr 13, 2005 10:58:12 PM

All good points John.

What I think it's going to come down to though is the old adage that "the light bulb has to want to change". While we can point to the exceptions that prove the rule like Firefox, you can't forget that the enormous bulk of FOSS is written by developers with a view to fulfilling their own software needs. What a gaping chasm exists between these developers' needs and the needs of any consumer market segment. Just to give one tiny example of the culture challenge: consumer markets *need* marketing to make them feel safe and comfortable about software purchases. Try selling that idea to the developers behind a FOSS project...

There were in fact quite a few comments at /. regarding support, documentation, easy to use installers, usability testing and interface design. However, the culture at the heart of FOSS is a million miles from all of these considerations.

OTOH I'm a big believer in the idea that any change can happen when it has to. I don't think it's likely that it will be something so simple as the lure of financial reward coming from outside the community which will drive change. If it's going to happen it will be because of some sort of pressure, some sort of change in circumstances, from within the community.

Posted by: Maxine | Apr 14, 2005 8:42:46 PM

Hey, I gather from the comments and so forth that most of you folks use Macs. As a long-time Windows, Unix and Linux user, I think you were right about three years ago, but you are getting further from the mark every day. While the server software has always been better on Unix/Linux, the end user apps are much improved this year. OpenOffice writer is now about as useful as Word, and getting to be less irritating. The Gimp is as good or better than Photoshop Lite or whatever they call it. I haven't found a good Illustrator/Corel/Freehand clone, and the diagramming tool seems pretty buggy too. But these tools all exist and are in rapid development. I am not quite ready to recommend a good Linux distro to unsophisticated users, but I wouldn't hesitate to set one up for somebody. For example, I am moving my kids off Win2k to Fedora this year. If you want to argue about the "productness" I'd be happy to respond.

Posted by: John Fisher | Apr 20, 2005 2:13:34 AM


we develop for both Mac and Windows, though the Mac is my primary platform of choice.

I think you are right in that the apps are definitely becoming products. But the whole product concept also includes things liek training, consultancy, and a whole ecosystm around the "generic product". While counter intuitive, the generic product does not have to be the best one out there to be the market leader. The whole product needs to serve the markets needs best.

The pieces are there, and are developing. Even big companies like IBM are getting in on part of the act.


Posted by: john Allsopp | Apr 21, 2005 8:23:53 AM