Gartner Group Debunking Open Source Myths

6j3net sent us a report from the Gartner Group that attempts to debunk Open Source Myths. It looks like they got things pretty much right. It looks as if this might be a great url to send the bossman when he asks questions.

In General, A Good Thing

Lets face it, there are lots of management types out there who don't read/understand/etc places like /. and so on. They fear OSS because of all these myths. Now someone "respected" by many has posted a report saying that OSS should be reassessed, etc.

Hell, irrespective of whether Gartner are paid mouthpieces for big-name companies or not, this kind of report can only help. All that was said is dead-obvious to us, but it's not to people who haven't been a part of the scene. This gives OSS a bit more of a legit status.

The thing to now be aware of is that this kind of report may very well lead to lots more "newbies" coming into the OSS world, from both the IT user/support side and the development side ("Carl, go check out OSS and see how it can help us sell more of product X" :)

It is now the responsibility of those who have said "Isn't this report kinda old-news & obvious" to ensure that the newbies don't go down the wrong path. Ensure that they are guided to reputable companies & examples. Help them avoid the pitfalls of the OSS world.

Gartner are opening the gates. It's up to the rest of us to ensure that OSS doesn't get a bad name as the "clueless hoards" crash the party and start milling aimlessly about.

This is how consulting firms work

They write two pieces. One praising OSS and another bashing it. Then they wait to see how the winds change. A year ago, OSS was just a fad. Now, hey - we knew you'd make it all along. Let us be friends, eh?

Gartner Response

I am a GartnerGroup analyst. In particular, I'm one of the GG analysts's who has covered OSS and Linux for quite some time. Consequently it should be a surprise that I am also an avid visitor to slashdot.
In lives before GG I was a technical architect and developer every bit down in the trenches as the typical slashdot reader and probably more than most.
I'm amazed at the notion that GG has been anti-Linux or OSS in anyway. In fact when you read the bulk of our research its quite bullish on OSS and Linux. Keep in mind that we've published LOTS of information of Linux and OSS; unless you're a GG client you probably haven't seen 99% of our research. Most of the comments I've read on slashdot are based entirely on ignorance or from a small snippet of published material.
I find it comical that advising our clients that Linux WON'T replace Win32 as the dominant desktop operating system in the next three years is somehow Linux bashing. I'd actually expect responses to be no sh** but more often it's the opposite.
Has Microsoft payed GartnerGroup to write Linux or OSS research? No. GartnerGroup does not write research specifically for ANY vendor. Furthermore, as any of our clients know GartnerGroup has consistently been one of Microsoft's strongest critics. Microsoft is a BIG company. Like most big companies they do some things very well and some things terribly bad. Gartner's value to our clients is an unbiased third party analysis and our Linux/OSS research IMO is an excellent example of balanced even handed view of an incredibly overhyped and radically evolving subject matter.
Are Linux and OSS overhyped? Of course they are. Every newly discovered technology trend goes through an inevitable cycle of hype and backlash as it enters mainstream IT environments.
Is there real substance under this hype? Of course there is. Gartner's charter is to help separate the wheat from the chafe. The notion that any negative position is bashing or selling out to a vendor is not only ridiculous but childish.

I thought long and hard before posting this response insofar as I'm sure it will bring inevitable flames but then again this will only prove my point. Bottom line, GartnerGroup is not the enemy, we will continue to track OSS trends, vendors, success stories (and failures) in an impartial and even handed manner. We will analyze and report on the good, bad and the ugly with regards to OSS. You'll probably like what we have to say in some cases and you'll hate it in others.
Do you have particular success stories of OSS projects in your IT organization? Do you have a specific OSS subject that you think we should be covering but aren't? I'd love to hear about them ([e-mail]). Do you have some asinime flame or insult to throw my way? Don't bother.

It will be interesting...

It will be ineteresting to see what posters say about the Gartner group in this article. In the past, when Gartner has commented negatively on Linux, there were roundly cursed as great fools. Now that they are at least mildly pro-linux, I wonder what people will say?

Bias is not how we get to truth guys.

Something to keep in mind here:

Open Source is not necessarily the same as Free Software, though the two overlap much of the time. Gartner is speaking in favor of Open Source and, frankly, I see absolutely no reason to keep source code closed off from people. Free Software, though, is not necessarily the same.

Only the GPL and its descendants guarantee Free Software, while there are many Open Source licenses (the Sun license, the BSD licenses, and the MPL, for instance) that are similar but don't necessarily accomplish quite the same thing. Rather than getting into a detailed discussion of licenses per se, I'd just remind everyone that Open Source comes in many flavors, and hopefully the market will reward the freer varieties over the less and non free ones. Gartner is on the right track, though.

- -Josh Turiel

Re:Myths.

Sig11,

I agree with your points about OS superiority being a myth, and overall I agree with the spirit of your post. However:

In reference to your points about proprietary software having some incredible success stories, your examples are all right on the money, and I guess that's my point. Until OSS developed momentum as a business model, there wasn't much percevied value (i.e. profit, i.e. money) in either releasing your code or open sourcing it. It took OSS success stories to convince companies that had long histories in the closed source model (the only model they knew that worked) to open up.

A parallel example is open-source news. A lot of newspapers were slow to get online because they were closed-source (pay-for-core-content) and that was the business model they knew. Now look at them clamber to get online and give their information away free, and develop other revenue sources. The example is not an exact parallel, but I hope my points come through anyway:

• Closed source success stories aren't successful simply because they were/are closed source.
• They were, instead, closed source because they didn't have any other business model to follow.
• Thus they were successful, I am reasoning, not because of OSS/CS, but because they were good, quality software.

But i think that's your point anyhow; that CS can produce good stuff too, right? But maybe now we'll see more of these CS companies open up their source as well? Or maybe not. That's a different question, and I won't extend this post with it right now.

Today's Weather Report for Hell

Cold, highs in the low 20s. Expect lows in the single digits tonight, and up to 40 below with the wind chill factor. Snow tomorrow, four to eight inches.

Good words for OSS from Gartner? I don't believe it.

Re:Myths.

That is to say they're impatient. The result is hundreds of releases of software each day. Some people think this is because we "release early, release often" - I think it's because the programmers didn't know or didn't care enough to make it work correctly the first time and then need to go back and rewrite the code again.

I am involved both in closed source, comercial software developement, and open source developement. And due to the visibility of the code, open source programs some time has to get it right, or someone will complain or fix it. Closed source on the other side, stopas at release two of your list; they get it working better, but the bugs are still there (Losers don't complain about bugs, they just want to play Quake. If the computer BSODs now and then - that's commonplace).
You are fairly right that different OSes are good at different things. But then you turn the world upside down and concludes Linux is not nessesarily better than Windows. Here is my personal conclusion of some OSes of today: Linux (and other UNIXes or clones): Stability, networking, programming, multitasking. OS/2: Flexible GUI, multitasking, (programming?). MacOS: easy to use GUI, _easy_. BeOS: video, multitasking, programming. Windows: widely used, many applications. From this list we learns that in fact anything is better than windows for any particular application, as long as the specific user application does exist on the said platform.
And last, but not least, Open Source is better because an OSS program can not by any means trade code beauty (maintainability, functionality, robustness, generality) for UI beauty. Of course, the OSS movement still have to learn to code good UIs, too.
--The knowledge that you are an idiot, is what distinguishes you from one.

It's the users...

As we all know, the freedom in free software is the *user's* freedom: to use, to combine, to know, to debug, to extend. As such, it should be no surprise that the development progress to date meets up with the crucial needs of the system's user/developers, thus the focus on serving and networking: OS developers may prefer to play and write games, but they're getting paid to make sure the bosses' networks and apps run correctly.

As the user base extends to those with other needs: non-software industries' apps, games, personal software, PC drivers, ease-of-use, ease-of-development, etc, the quality and availability of those applications will improve.

Although this doesn't refute your point that free software is not always better right now, it does imply that someday it will at least catch up.

And, if you believe in the power of empowered users, it's easy to see that it will likely surpass proprietary software, as it has already done in the areas of greatest need. I suggest that it also will or already does serve needs that proprietary software would have taken forever to get to: child-simple programming languages (what's the prorietary advantage in that?); truly portable applications; unobtrusive, ubiquitous, standard networking; small, reusable, interactive tools.

In general, free software will be much better at creating a marketplace of commodity software: no lock-ins, no single-source, no platform-specificity, no buying the unnecessary to get the needed, no forced upgrade paths, better future-proofing, better accountability, better responsiveness, real choice.

The day when "open source is always better" is not here, but it's coming.

TCO is good

Actually, the TCO of ownership is a point *for* free software, in this context. If only 20% of the TCO is buying the software, it means that there still is the remaining 80% to extract from the customer, so giving away the software as a "loss leader" may make perfectly good sense.

!FUD

It is actually encouraging that it isn't really hype. Just the facts. It doesn't tell you to or not to use OSS models, just why some people do.

Re:Myths.

On the hardware side - have you been watching the source checkins/checkouts for hardware drivers? Then you'll notice they follow a peculiar pattern - Initial Release (aka, it works, but it's slow). Revision 2 (it's buggy, but faster), and finally Revision 3 (finally get it right). The reason is that most people in the OSS community a) don't have access to debuggers to catch this stuff earlier and b) Often don't wait and properly engineer their drivers prior to implimentation. That is to say they're impatient. The result is hundreds of releases of software each day. Some people think this is because we "release early, release often" - I think it's because the programmers didn't know or didn't care enough to make it work correctly the first time and then need to go back and rewrite the code again. Now, if you listen quietly for a moment you can hear people already hammering at their keyboards crying foul.

That's kind of missing the point entirely. What you describe isn't even a side effect, it's the whole point. What makes open source "open"? You release the code, and you conduct development out in the open where everybody has the opportunity to participate. You release sub-1.0 versions and later unstable versions in the hope that others will try it out, make useful suggestions about adding features, and help you to fix the bugs.

In the majority of cases the developers will clearly mark which versions are stable and which versions have more features but are still buggy. Obtaining free software is a case of caveat emptor just as it is with any other kind of software.

But the *difference* with open source (as opposed to closed source) is that (i) the developers won't try to pretend that bugs don't exist; (ii) being subject to greater scrutiny, they'll put more effort into fixing them; and (iii) if you can't wait, you have the option of trying to fix it yourself, or paying someone else to do it.

I know that access to the source won't be much use to those without the resources to make use of it. But, if stability is your sine qua non, then obviously, get the stable version and leave development versions alone. And with closed source you don't even have the option.

May I further point out that my experience is that closed source companies often opt not to fix some bugs at all. In those cases it'll *never* get fixed. Which kind of leaves you in the lurch, doesn't it?

I'm surprise that you of all people would say something like this. I think you must have left your brain on your pillow this morning! Do I hear the sound of Signal 11 smacking his forehead in chagrin?



Consciousness is not what it thinks it is
Thought exists only as an abstraction

The real reason most software exists...

..is to solve someone's vexing problem. The majority of software is NOT written to allow it's creator to drive around in a Ferrari but to get some work done. If my motivation for writing some code is to solve a personal problem or more likely a problem I'm having at work, then open sourcing the best thing I could possibly do. Others who can use a similar solution can make suggestions or even give me patches to make it even better. There is a tangible reward for open sourcing a project. The developer gets patches back! Come to think of the submitters of patches know that others are contributing so they too enjoy what the primary developer/maintainer enjoys....better software. If you look at it as working for patches rather than money then a lot of the mystery of just what motivates people to write open software goes away. The fact that non-coders can enjoy the fruits of this labor is secondary. Even non-coders can note bugs so they can get better software too. Remember: most software will not earn scads of money even if it were closed up and sold for profit. I would rather be an open sourcer getting patches and bug reports than a shareware coder getting two or three registrations a month. It's obvious which tends to produce better software.

Myths.

While we're discussing myths.. how about we debunk another one: Open Source is superior to everything by virtue alone. Listening to RMS, slashdotters, and ESR's writings would have you believe OSS will revolutionize the world and proprietary software is /all/ bad.

Let's not get ahead of ourselves. I direct you to several stunning *proprietary* software achievements which the OSS community has yet to duplicate.

• Autocad. For that matter, any serious CAD program.
• A desktop my mom can use.
• web plugins (until recently, we had no flash!)
• Games. Quake and Loki's offerings may be cool, but they're a small subset of what the Windows world has. I'm still waiting for a Red Alert clone.

Just to name a few. On the hardware side - have you been watching the source checkins/checkouts for hardware drivers? Then you'll notice they follow a peculiar pattern - Initial Release (aka, it works, but it's slow). Revision 2 (it's buggy, but faster), and finally Revision 3 (finally get it right). The reason is that most people in the OSS community a) don't have access to debuggers to catch this stuff earlier and b) Often don't wait and properly engineer their drivers prior to implimentation. That is to say they're impatient. The result is hundreds of releases of software each day. Some people think this is because we "release early, release often" - I think it's because the programmers didn't know or didn't care enough to make it work correctly the first time and then need to go back and rewrite the code again. Now, if you listen quietly for a moment you can hear people already hammering at their keyboards crying foul. But I'm not done yet.

Let's take another myth: that we're somehow superior to windows or macheads. Comeon people, this is dogmatic and fanatical in the extreme. We have an OS that does * some * things better than the others. But NO OS is superior on the basis of it's name alone.

I'd also like to point out that there are other "source" movements out there that make progress approximately on par with linux and it's related software despite the fact that it has a much, MUCH smaller developer pool to draw from: the BSDs. I'll now turn matters over to my fellow slashdotters for an explanation...

Re:a bit light?

Yes.

This is just an index to lots of documents listed at the end. And these documents actually cost money to obtain (just gessing, but Gartner has to earn its living somehow).

Re:It will be interesting...

Gartner groups usually gets their facts straight, but often there are insufficient facts available to get a full answer to their question, so they guess. When they guess about Linux, they usually lean towards Linux being stable & useful, but away from Linux being successful, widely accepted and profitable.

Here there was nothing to guess at, their report is just a rehash of literature that's been around for a while now. No guessing, no leaning. I certainly wouldn't use this one article as an indicator that they are suddenly "at least mildly pro-linux".

----

OSS Business Model Sustainability

"Myth 3: It is against human nature to work for nothing."

OSS developers do not work for nothing; an OSS project is a training and proving ground exposing young developers to large-scale collaborative development. Another reward is the accumulation of kudos -- this may be sought for its own sake or leveraged commercially, especially within organizations that follow an open-source business model (see Note 3).

This is a very perceptive, succinct write up by Gartner, and Rob got it right when he said this one would make a good executive summary.

My question is how sustainable this is. I worry just a little that OSS business model/business culture is a bubble that could burst (especially after enough IPOs and mergers etc etc). Look at Mozilla... what if Netscape programmers were suddenly reassigned to work on AOL 7 (seeing as AOL 6 [observers.net] is already out ;)? ATW doesn't strike me as anything less than a cold-hearted capitalist corporation bent on world domination. RH is a more encouraging example of an OSS business culture that I would expect to survive.

We can't deny that Linux has fueled a lot of the interest in OSS as a business model. If Linux ever stagnates or reaches a development plateau (gasp-shudder-perish-the-thought), and there isn't another project with it's momentum to "lead the charge", could OSS businesses fade? Or is it here to stay? (I personally feel that OSS business is more long-term profitable than closed-source, but IANAE).

If anyone has any links to research or projected growth models, I would appreciate it, particular ones with positive expectations for OSS business ;) Ethan

Wow

This will be very interesting. The stubby trolls that inhabit the PHB-nirvana that exists on the 16th floor of my office pay big bucks for Gartner research services and love to wave their little reports in front of the techs in meetings where the subjects are beyond their comprehension.

Now I have something to put us closer to equal footing.

I especially love this part:

Bottom Line: Contrary to common perceptions, open-source development is neither a recent phenomenon nor a transient one, and more significantly, it is one that will increasingly be associated with commercial vendors and end-user organizations. We recommend that IT organizations which currently exclude all OSS from their acquisition plans should re-examine this policy.

I just can't wait for the next meeting to bring this up again and show these fat, balding, middle-aged, clueless management drones that they have been laughing off OSS at their own expense all this time. Maybe it'll do some good, maybe not. Just the recommendation alone that we "re-examine" the no-OSS policy will set off a flurry of activity as someone will get this action-item to report on this and most of the points that were cited in our poicy are directly debunked in the Gartner article.

Somebody is going to look even stupider than usual.