26 May 2008

Why Open Source Will Triumph

This is why the traditional software development model is doomed:

Recently, I wrote a review of the note-taking application Tomboy. Though I find Tomboy exceptionally useful, I had a minor issue with the inability to create new notebooks from within a note. Within hours of the review appearing on Linux.com, Boyd Timothy, one of the app's developers mentioned in the article's comments that my idea had merit and said he would add the feature to an upcoming build. True to his word, he did. This is a shining example of one of the most valued yet sometimes overlooked features of open source software: it really is for the people, by the people.


ZZ said...

A much more powerful example would be if the user would add the feature he thought was missing. That would really require open source. As the example stands: a developer adds a new feature in response to a review -- this can happen and I am sure happens fairly often for closed source software too (motivated by more sales).


Glyn Moody said...

I agree that could only happen with open source, but my point was more about the general attitude of open source: the fact that it does respond to users so quickly, and not just in theory.

Closed source could theoretically do this, but I bet there are precious few cases where it has so far, simply because it's not in the DNA.

Dr. Roy Schestowitz said...

Oh, careful there, Glyn. It's a Mono app. Avoid and avoid.

Glyn Moody said...

It's the principle that interests me, not the actual app - but thanks for the warning.

Anonymous said...

Roy beat me to it, but I wanted to concur.

I do find Tomboy really useful and do use it. But I would rather use something else... Mono is tainted.


Glyn Moody said...

Hey, people, don't shoot the messenger....

Anonymous said...

I think that somewhere the point is being missed, as ZZ rightly pointed out. I've had brilliant interactions with closed-source programmers (foobar2000, kookie jar) who are very responsive. At the end of the day, getting a feature on depends on programmer responsiveness, regardless of whether it is open-source or closed. Most open-source projects available on the internet have small core development groups in any case, just as with most closed-source software. Such small groups are a) more responsive & b) more liable to be overwhelmed by requests if the software grows too fast.

The briliance of open-source software really lies in two factors. One is that you yourself can do the programming. The other is the fact that if a project is large enough, there is also a greater chance that someone else will also like your suggestion, and will try to implement it.

Such anecdotal evidence as you cite is very fraught with danger, because there are as many (if not more, simply due to the fact that there's more closed-source than open-source) stories of closed-source software developers being responsive as well.

Software developers do their work for three reasons: a) for themselves (a great many independent, home-based coders); b) for money; c) for users. Responding to users is in software developers DNA regardless of whether they "believe" in open-source.

Glyn Moody said...

I am sure there are responsive closed-source programmers, but my point is about the overall culture, not individuals. The thing is, open source by its nature invites feedback; indeed, it loses much of its point if it doesn't act on it.

Moreover, there is a kind of Darwinian selection among coders such that if a project team is generally unresponsive, others can fork the code and do better. This simply can't happen with closed source: it all depends on whether the team is good or not. If it isn't, there is nothing anyone can do about it.

Anonymous said...

"I am sure there are responsive closed-source programmers, but my point is about the overall culture, not individuals. "

eh? that actually wasn't your original point at all, but now since your "original point" has been disproved, you're re-adjusting.

btw - OpenSource also has a history of abandoned projects, and unfinished and sloppy work.
As soon as people figure out that

OpenSource projects generally don't pay the bills (who's willing to buy something that's free?), they move back to closed source models.

When was the last time you sent an opensource programmer a donation?

wan't opensource and linux supposed to be taking over back in the late 1990's? Since then, linux adoption has actually gone down.

Glyn Moody said...

One counter-example does not disprove my point, which is about the underlying dynamics of closed and open development.

Similarly, I never said that ever open source project thrives – on the contrary, the Darwinian selection that occurs in the free software ecosystem is there precisely to weed out the losers, and to ensure that only the best survive. Again, this is unlike the closed world, where pressures are mostly financial: bad programs can linger on simply because the company producing them can still extract money from hapless users.

As for making money from open source, there are many models. At the personal level, coding free software is a great way to get a job – either with an open source company or elsewhere; companies make money from services, support, customisation etc. I and many others written about these extensively.

And GNU/Linux *is* taking over in many sectors. For supercomputers, for example, it holds 85% of the top 500 systems (http://www.top500.org/stats/list/31/osfam); for servers, it is increasing its market share; for embedded systems it is rapidly becoming the main platform. The only area where its market share is small is the desktop, but it's always been accepted that this would be one of the last to shift.