A think tank's view of free software

[Posted May 9, 2007 by corbet]

Back in early March, a company called the Olliance Group held a gathering of about 100 corporate manager types at a resort in California's wine country. This "Open Source think tank" has now produced a 16-page report [PDF]. It is, indeed, an interesting look at how a certain part of the corporate world views free software - though, perhaps, not entirely in the ways its authors intended.

When a self-appointed "think tank" gets together to talk about free software, one is right to be cautious. When one of that event's top-level sponsors is Microsoft, an extra degree of nervousness seems appropriate. The other top-level sponsor, naturally, is Novell; the remainder of the list is NEC, Unisys, Jasper Soft, OpenLogic, and SugarCRM. Not the most community-oriented bunch one could have come up with.

LWN readers will be glad to know that "Overall, the CIOs unanimously agreed that open source is viewed as a viable option in software procurement decisions for their companies." Once they made that admission, however, this group started to raise its complaints about open source, many of which could have come from the 1990's. The first was lack of support - evidently there still is not enough commercial support for open source software. The report notes that "this is something the open source industry will have to address to increase adoption by companies." One would think that if there is truly a need for more support these companies would see that need as a business opportunity rather than an obligation.

Another problem, it seems, is interoperability:

CIOs desire greater interoperability built directly into open source products. This is an area where proprietary solutions maintain an advantage over open source, as it is far easier to integrate and use a suite of proprietary applications that are guaranteed to interoperate and that have common interfaces that make it easier for end-users to learn and use the suite.

This is a surprising claim, given that free software developers generally work toward interoperability with everything. The next claim is just as surprising:

Open source lacks compliance with many standards when compared with proprietary solutions. These standards include universal standards such as ISO, and industry-specific standards (financial industry standards, health care industry standards, etc.). It was acknowledged however, that open source offers some advantages in the area of technology standards through its openness and transparency and its ability to facilitate the creation of de facto standards such as Eclipse and ODF.

The description of OpenDocument as a "de facto standard" borders on the dishonest. The various reasons why certain "industry standards" may not be supported as well as others are not examined.

Think tank attendees bemoaned the fact that monetizing open source remains challenging. Then, there is this problem:

Open source generally depends on a corps of motivated volunteer developers to develop features. Often, the features that developers are interested in working on are different then features that customers are requesting. For example, Openoffice customers want more Visual Basic macros to ensure interoperability with Microsoft Office, but OO developers have not been all that interested in building VB macros.

The idea that a company whose business model depends on better VB support could devote resources to the creation of that support is not mentioned anywhere in the report.

Licensing is an issue which is mentioned several times in the report:

Open source licensing is a big source of confusion due to the number of open source licenses, and a lack of understanding on how licenses impact business, as well as how licenses interact with one another. Some licenses require technology to be shared with the community, other licenses require attribution, and numerous licenses have different ways of dealing with software patents. Furthermore, many licenses are incompatible. License proliferation, confusion and incompatibility are barriers to the continued growth and adoption of open source.

Clearly, we would be better off with the simplicity, compatibility, and fairness found in proprietary software licenses. Beyond that:

Think Tank participants bemoaned the lack of a business-friendly license that adequately addresses issues such as copyright, patents, attribution and indemnification. While nobody was suggesting "yet another license" as the solution, the dissatisfaction by commercial vendors and customers with the existing licenses was clear.

It would be most enlightening to see what this "business-friendly license" would involve, but the attendees apparently ran out of time before they could elaborate on that point.

The GPLv3 draft was also discussed, with a generally negative response.

Another problem:

These issues also point to the need for better governance of open source contributions. Currently, projects have many different standards governing code contributions - some communities vet the code, some require contribution agreements to be signed and others have no such requirements. The lack of standards and governance on contributions raises concerns on the source and legitimacy of code that is incorporated into projects.

This is a claim that needs to be backed up: despite the intense attention which has been given to the provenance of code in a number of high-profile projects, the number of real problems has been exceedingly small. If the attendees of this think tank wish to claim that the code found in free software projects is less likely to be legitimate than proprietary code, they need to come up with some evidence to that effect. Sadly, space constraints appear to have prevented this evidence from being included in the report.

Other worries include a lack of open source developers - their numbers are not keeping up with the growth of the industry. The fact that quite a few developers are coming out of universities is considered to be a good thing, but not without reservations: "However a concern was expressed that due to the popularity of open source development at universities, graduates may be lacking key skills such as sound architecture, defining customer needs and product management." We also hear that open source "tends to fragment easily," presenting problems for vendors. "Commercial open source tends to be less fragmented, while 'pure' open source tends to be more fragmented."

All is not bad, though. Open source offers "flexibility in procurement" and "flexibility in deployment" where "companies can mix and match open source software as they please" - despite all of those interoperability and standards compliance problems we heard about earlier. Faster product cycles are seen to be good, as are faster bug fixes. Plus:

In addition, there is "perceived" value in the ability to fix or enhance open source code at the CIOs pleasure even if the vast majority of user organizations do not .

This "perceived" value is as close as this report ever gets to any sort of freedom-related issue.

There is plenty more to look at in this report, but perhaps it is best to finish with this observation:

Finally, OSS and proprietary models continue to converge. Proprietary companies are taking elements of the open source model, including faster development cycles, and free, downloadable trial versions. OSS companies are taking elements of the proprietary model, by offering support, updates and indemnification.

This report gives no space to the developers of all this software, beyond complaining that both their numbers and their motivation to implement Visual Basic macros are insufficient. There is no thought toward maintaining healthy development and user communities, addressing problematic legal issues, or contributing back to the community in any way. These are people who see free software as a well from which they can draw resources for their businesses, but that software is just a raw material. They want to repackage and sell that material in as proprietary a manner as possible.

If this group represents the future of the open source business community, we could be in real trouble. A look at the list of sponsors given at the top of this article is cause for comfort, however, as most of the companies which have found real success with free software chose not to support this event. So there is reason to believe that this "think tank" is not representative of the wider business community, that, instead, it's a group of leaders of businesses who wish they were doing better at "monetizing" free software.

to post comments

a MiStaken view of FOSS

Posted May 9, 2007 15:37 UTC (Wed) by grouch (guest, #27289) [Link] (1 responses)

This thing reads like the bullet points of Microsoft marketing training papers. It's the same old stuff, right down to the tired FUD about "integrate[d]" suites. There must be a regular stampede away from the cages, if MS is down to recycling FUD this old.

More connections...

Posted May 9, 2007 19:03 UTC (Wed) by dmarti (subscriber, #11625) [Link]

The co-organizer of the event is a very large law firm, DLA Piper, which has been hired for lobbying by such anti-freedom clients as the RIAA, Reed Elsevier Inc, and Time Warner.

It is shameful for Novell to participate in this stuff (they've even dropped "agree to disagree" in joint "IP" FUD with MSFT) -- but at least Tim O'Reilly is out there selling expensive reports on the pro-freedom side.

A think tank's view of free software

Posted May 9, 2007 15:37 UTC (Wed) by nigelm (subscriber, #622) [Link]

Oddly enough the previous article in my RSS reader's set of stuff is A Think Tank of One (or Ten Million, take your pick) by Michael Tiemann - a response to the same report

A think tank's view of free software

Posted May 9, 2007 15:43 UTC (Wed) by danielpf (guest, #4723) [Link]

The funny aspect is that what is declared as a drawback is then an explicit advice for the OSS community how to protect itself from these greedy sharks, for example by adopting the GPLv3 widely.

A think tank's view of free software

Posted May 9, 2007 15:49 UTC (Wed) by mmarsh (subscriber, #17029) [Link]

Has there been a "similar" report from a more free-software-friendly group of companies? That is, a group of companies that have been able to make money using (whether as commodity or through additional development) free software, describing how they and other companies have or could benefit from free software. It seems like this should be "easy" for them to do, without really giving away business models. Comments from a company like IBM or Intel on the practical implications of FOSS licenses would be incredibly valuable to companies without experience using free code, as would making the point that if you want more features in free software, and you can afford to pay developers, then paying people to work on free software benefits everyone. The latter does not preclude internal-use-only locked-down modifications, either.

A think tank's view of free software

Posted May 9, 2007 16:24 UTC (Wed) by nix (subscriber, #2304) [Link] (21 responses)

Bravo! The well-aimed sarcasm in this article deserves some kind of award. It read like Private Eye, only without the inaccuracies :)

The `greater interoperability' thing could have had a huge flashing logo over it reading `we're trying to find an excuse that will eliminate all alternatives to Microsoft Office'.

I suspect the report was written by someone who didn't know what they were talking about and were just plugging together buzzwords written by someone else: the bit about `skills such as sound architecture' was particularly bad (and this is just reading the bits you quoted: I can't bring myself to read the original, I have too few brain cells left to kill them with things like that).

And of course support is an `element of the proprietary model'. It's not as though RMS talked about it a quarter of a century ago in the GNU Manifesto or anything.

Truly boggle-worthy stuff.

A think tank's view of free software

Posted May 9, 2007 16:54 UTC (Wed) by epa (subscriber, #39769) [Link] (16 responses)

The sentences from the report quoted in this article are all valid points. Yes, there is a lack of commercial support for most free programs. Perhaps you or I don't need it, but a lot of companies want it. Yes, there are sometimes weak points with interoperability - while you can buy plenty of things that integrate directly into Microsoft Excel, but not so many statistical packages in the free world have an OpenOffice.org plugin. Or if they work with OOo they don't plug into KOffice, etc. Yes, there are industry standards which are better supported by proprietary software. Companies aren't interested in the excuses for why this is the case, only that it is.

If a report published by Red Hat or the Apache Foundation or some other august body had pointed out the need for 'better governance of open source contributions' then we would be treated to a solemn LWN editorial agreeing with this and stressing that we, as a community, must remain vigilant blah blah. Because it comes from a report by users or potential users of free software, who happen to be companies rather than individuals, we get treated to a Slashdot weenie diatribe making sarcastic quips rather than acknowledging that there might be something to look at.

Indeed, even if none of the points had any validity, they are still _perceived_ as valid by these hundred or so business people. And their perceptions of free software will influence what they decide to use. If you wish, you could see it as a marketing problem, a need to correct mistaken views in the world outside. Myself I think that 'the customer is always right' is a good motto to follow.

These CIOs are not 'greedy sharks' as another commenter called them. They are users, just the same as you and me. They have their own businesses to run and saving the world is not their main job. The boring work of word processing, accounting point of sale automation and the hundred other things needed by businesses may not be as cool as setting up peer-to-peer wireless networks or porting Ruby to camera phones, but it's an important part of the world and free software needs to spread there too. We should listen to the conclusions of the report and find some way to address them, rather than blowing them off.

A think tank's view of free software

Posted May 9, 2007 17:47 UTC (Wed) by nim-nim (subscriber, #34454) [Link]

> Indeed, even if none of the points had any validity, they are still
> _perceived_ as valid by these hundred or so business people.

A few points have more or less validity, but this is not an objective or balanced report. It reeks of careful steerage of unsophisticated business people by biaised consultant(s).

While it repeats common misconceptions that may be perceived as valid by some business people, some of the points just lack any relevance to software customers and only make sense if someone "helped" formalising the report, putting an heavy spin on the think tank more confused and neutral reflexions.

"Finally, OSS and proprietary models continue to converge. Proprietary companies are taking elements of the open source model...": no business people I know of would write this kind of drivel by themselves:
those who care about "the open source model" know enough to dismiss this (the business people approach to the open source model is "are we independant from $greedy_big_vendor", and open source fakes typically fail this test)
those who don't care just don't care

A think tank's view of free software

Posted May 9, 2007 17:58 UTC (Wed) by jmorris42 (guest, #2203) [Link] (13 responses)

> Yes, there is a lack of commercial support for most free programs.
> Perhaps you or I don't need it, but a lot of companies want it.

No there isn't. It isn't structured in the rigid fashion you might want it, but it DOES exist, plentiful and sufficient to any purpose.

At the bottom of the stack, if you are a small operation you can buy a commercial support contract, see www.redhat.com or www.novell.com for details. Clue, neither company is selling a 'product', they sell support and at least RedHat is selling enough to jump to the 'big board.' Novell is working hard on catching up. So if 'support' is what ya just have to have give either one of em a call and a sales weasel will be right with ya to power close a deal just like you are accustomed with whoever you are wistfully comparing open source with.

If you have a need for support on a particular package beyond what those options offer find yourself a local nerd/geek/developer and put him on a consulting contract. Or look into one of the many open source / free software consulting operations which exist. Remember, this is open code so you do not have to buy support from the same people developing the package, just from someone with experience with it.

And finally, if a package is vital to the operation of your mega corporation you simply hire a lead developer. For the price of one headcount you get the ultra elite platinum with glowing enriched uranium bars level support contract for unlimited seats.

> Yes, there are sometimes weak points with interoperability - while you
> can buy plenty of things that integrate directly into Microsoft Excel,
> but not so many statistical packages in the free world have an
> OpenOffice.org plugin.

No, what you want isn't interoperability since Excel doesn't have an open published plugin interface. What you want is Excel. If you define the problem as "A product 100% binary, bug for bug, compatible with Excel that releases new versions at the same time as Excel" you kinda stop even the hope of migrating someday in the future, because that ain't possible.

> Yes, there are industry standards which are better supported by
> proprietary software.

Only if 'standards' is defined as Microsoft. And if your 'reasoning' were valid everyone would still be running Lotus 1-2-3. When change comes again it will probably be just as rapid. The question is exactly WHAT would cause a switch.

>They are users, just the same as you and me.

No, look over that list again, they aren't users. The word you are looking for is competitor. A better description might be Microsoft's slaves. What the produced was a marketing document, any valid criticism was accidental and purely an unintended consequence. That's why the snarky reception.

The reference to ODF as a 'defacto standard' was the givaway. That wasn't just wrong, I'll go ahead and say what corbet wasn't willing to risk a lawsuit over, that it was an intentional, knowing lie. The whole basis for all this effort to get governments to adopt it is based on the fact it is a published standard, if everybody at that gathering managed to miss that they are so mentally challenged one wonders how they manage fill in their golf cards correctly.

A bad experience with RedHat and Novell

Posted May 9, 2007 23:12 UTC (Wed) by hazelsct (guest, #3659) [Link] (12 responses)

> At the bottom of the stack, if you are a small operation you can buy a commercial support contract, see www.redhat.com or www.novell.com for details. Clue, neither company is selling a 'product', they sell support and at least RedHat is selling enough to jump to the 'big board.' Novell is working hard on catching up. So if 'support' is what ya just have to have give either one of em a call and a sales weasel will be right with ya to power close a deal just like you are accustomed with whoever you are wistfully comparing open source with.

I regretfully have to say this has not been my experience. I am a Debian developer going on six years, and Linux user and enthusiast for nine years. But I am paid to do engineering not system admin, and in any case we wanted a partner who would be around in case I left. So we went shopping for a Linux shop to provide a server, software and support for our mixed Windows-Mac-Linux (me) shop, and I was confident we'd find just such a "sales weasel" to "power close a deal".

Unfortunately, neither RedHat nor Novell put anything good forward. RedHat referred me to their list of RHCEs. Out of five outfits, only two replied to me: one was a one-man shop (not an option if you hope your business stays around longer than that one man), and the other didn't seem to want to do business with such a small firm as ours.

Novell was worse: they forwarded my inquiry to an internal sales rep who took TWO WEEKS to get back to me, then sent a single email referring me to a local Novell vendor. So I contacted the vendor, which sent a secretary to "assess our needs" -- who had NO CLUE about our technical issues. Totally unacceptable.

The message I got was: unless you're a Fortune 500 company, neither RedHat nor Novell wants to deal with you, nor are their supposed legions of business partners competent to do so. There are a lot more MSCEs (or is it MCSEs?) in the world, and the competition is producing higher quality.

As for our business, we will likely get a Windows server, as the "safer" option. I will remain a Debian developer, as I love the stuff, but am sadly disillusioned by the lackluster performance of the two "market leaders", and now understand why we are getting no traction in the SMB market. For the sake of the community, I hope others' experience has not been as bad as ours.

A bad experience with RedHat and Novell

Posted May 10, 2007 0:32 UTC (Thu) by dlang (guest, #313) [Link] (1 responses)

unfortunantly you experiance isn't unique, I've tried several times over the years to have my company buy support (we actually had a contract with linuxcare for a year, but at the end of the year I was talking with their sales guy and outlined the problems that we had been having and he reccomended that we not renew whith them, points for honesty at least)

if you just want to pay someone money you can buy enterprise licenses from redhat or novell, but if you actually want to be able to get support on anything other then a standard desktop config you are basicly out of luck.

A bad experience with RedHat and Novell

Posted May 10, 2007 23:54 UTC (Thu) by giraffedata (guest, #1954) [Link]

This is great information; I've always wondered just what the deal was on Linux support. It's hard to believe someone like Red Hat can economically support an entire monstrous open source distribution, culled from developers from all corners of the universe, as a proprietary vendor could support something it produces itself. And these experiences seem to bear that out.

It seems to me there's a middle ground between a Linux distribution and Windows. A commercial Unix would come with all the support, guarantees, and dependability into the future, while still leaving the door open for small unsupported open source deviations where necessary. It's pretty easy to plug open source components into AIX; IBM even gives you a boost sometimes. It's not so easy with Windows.

A bad experience with RedHat and Novell

Posted May 10, 2007 4:20 UTC (Thu) by lysse (guest, #3190) [Link] (8 responses)

> Out of five outfits, only two replied to me: one was a one-man shop (not an option if you hope your business stays around longer than that one man), and the other didn't seem to want to do business with such a small firm as ours.

So your problem with the former was exactly the same as the latter's problem with you? Mrs Be-Done-By-As-You-Did wants a hug...

A bad experience with RedHat and Novell

Posted May 10, 2007 9:08 UTC (Thu) by mjthayer (guest, #39183) [Link] (7 responses)

Surely someone looking for business support is allowed to make decisions based on business considerations. In this case, the only available option was found to be unsuitable from a business perspective. Are you suggesting that business perspectives should be ignored here? Or that free software is fundamentally incompatible with business?

Quid pro quo

Posted May 10, 2007 14:30 UTC (Thu) by man_ls (guest, #15091) [Link] (5 responses)

I think lysse's point is that your business considerations (to the one-man shop) were as valid as the business considerations of the second company (to your small business). Put another way, a small company that doesn't want to do business with another small company deserves to be disregarded by bigger companies. Will you disregard Windows if all MCSE's that want to do business with you are one-man shops?

Quid pro quo ignores market reality

Posted May 11, 2007 1:58 UTC (Fri) by hazelsct (guest, #3659) [Link] (4 responses)

So what you're saying is that a small business should not expect to get professional-level support in OSS. Which is why we are not getting any traction in the SMB sector. Like I said.

To answer your question, we already have a 100-strong company full of MCSEs doing business with our four-person shop, which sends techs out to our site on an hourly as-needed basis (installation, troubleshooting, etc.). And that four our TWO Windows users (one Mac, and me on Linux)! And there are multiple other businesses waiting in line behind them who could take our contract. They can set up a Windows server for us any time we say the word. But thanks in large part to me, we want(ed) to give Linux a try.

What a pathetic attitude. If this is how the community feels SMBs should be treated, then OSS will never get anywhere in this sector, while Microsoft continues to eat our lunch. And in the US at least, that's a very big sector to just walk away from.

Quid pro quo ignores market reality

Posted May 11, 2007 6:09 UTC (Fri) by man_ls (guest, #15091) [Link] (3 responses)

You did not answer my question, and it seems you did not understand my point, which is: maybe professional-level support can be provided by a one-man shop as well as by a 100-strong company. I don't know why there are so few RHCEs in your area and frankly I don't care, but maybe it is because SMBs don't want to do business with those who are small. On purely business concerns, the few big shops have to concentrate on the big companies which require their services.

What your comment suggests is that there are untapped business opportunities waiting for entrepreneurs to give them support, but we already knew that. So people, get certified and open your support company!

Quid pro quo ignores market reality

Posted May 11, 2007 12:46 UTC (Fri) by sepreece (guest, #19270) [Link]

Note, though, that "There's a great opportunity here" doesn't really respond to the "think tank" point - support is a problem, still.

Quid pro quo ignores market reality

Posted May 11, 2007 20:28 UTC (Fri) by hazelsct (guest, #3659) [Link] (1 responses)

You're right, I didn't answer your question, sorry. We wouldn't consider a one-man Windows support shop either, for the same reason. In case I wasn't clear, that reason is: there's no redundancy in case something happens to the one man. Whether that's retiring to Jamaica, or getting hit by a bus, or having to spend a week taking care of a sick relative, any of these things which took this person away would present a major problem for us if something went wrong.

Our clients don't depend on us in the same mission-critical day-to-day manner, so there's no valid comparison there.

And I agree that there's a real opportunity here...

Quid pro quo ignores market reality

Posted May 12, 2007 0:43 UTC (Sat) by man_ls (guest, #15091) [Link]

Thanks for clarifying that. I have long thought about setting up a support shop for Linux machines, as I'm sure other LWN readers have; in fact it is only by chance that I did not do it three years ago, but the possibility is always there. Knowing the conditions of "our" future customer base is always valuable, so it was not an idle question.

Maybe a federation or cooperative would be a better model for one-man operations; probably a middle-sized company would inspire more confidence into customers.

A bad experience with RedHat and Novell

Posted May 11, 2007 23:58 UTC (Fri) by lysse (guest, #3190) [Link]

> Surely someone looking for business support is allowed to make decisions based on business considerations.

Ah, yes - criticism IS prohibition. *shakes head in dismay*

A bad experience with RedHat and Novell

Posted May 10, 2007 9:06 UTC (Thu) by jschrod (subscriber, #1646) [Link]

It's not only small companies.

In a recent problem at a Fortune 500 company (one of my customers), neither IBM nor Novell was able to solve the problem. The amount of effort they put in was ridiculous low, if I compare that to Solaris or AIX support. So we addressed Red Hat (we were willing to switch distributions), and got the answer that they don't have sufficient resources to handle our support requests. (We wanted the equivalent of a Sun Platinum Support contract.)

Thus, I agree with your sentiment: The level of support for Open Source is not in the same league as that for proprietary Unix systems. That's why we still have to use Unix(tm) for mission-critical servers, even though the majority of the rest can run Linux.

A think tank's view of free software

Posted May 10, 2007 14:37 UTC (Thu) by man_ls (guest, #15091) [Link]

The sentences from the report quoted in this article are all valid points.

Even if they are valid points they are presented in a too biased way to be of any use. As e.g. this article from Schneier shows, how you present things affects the results you get; so judging from the output, those poor CIOs were effectively FUD'ded to death by Microsoft and Novell employees.

You are right in that I miss one of our editor's balancing paragraphs at the end about "not being smug" and recognizing that "there are valid points here".

A think tank's view of free software

Posted May 9, 2007 21:13 UTC (Wed) by sepreece (guest, #19270) [Link] (3 responses)

"I suspect the report was written by someone who didn't know what they were talking about..."

This comment, and a number of others, seems to misunderstand the nature of hte report. This is not a white paper by a research firm, it's a report from a workshop at which various kinds of participants, from various companies, talked together about FLOSS-related issues from their individual perspectives, which were then summarized by the organizers.

So, the whole point is that if some of these comments are clueless, that reflects the thinking of the participants, who are people with a lot of money to spend on IT and software in general. Their opinions should be interesting to people who want to make money on their FLOSS activities.

If the workshop participants say there is a shortage of support, our editor misses the point in saying the participants should view that as an opportunity. The point is that the *readers* of the report should view that as an opportunity.

A think tank's view of free software

Posted May 9, 2007 21:58 UTC (Wed) by nix (subscriber, #2304) [Link] (2 responses)

It was more that the sentence I quoted was plainly written by someone who
*didn't know what the words they were using meant* or at least not how
they are generally used. It's the same sort of linguistic red flag that
gets waved if I see someone refer to `source codes' (oops, that's a mass
noun).

Perhaps it's simpy that the workshop report was put together by someone
who didn't know what the words meant (which is rather odd but not
inexcusable).

(Yes, it's nasty and prejudiced of me. I know. ;) )

A think tank's view of free software

Posted May 9, 2007 22:33 UTC (Wed) by allesfresser (guest, #216) [Link]

I'm glad someone else cares about mass nouns besides me... ;)

A think tank's view of free software

Posted May 10, 2007 16:31 UTC (Thu) by sepreece (guest, #19270) [Link]

"It was more that the sentence I quoted was plainly written by someone who
*didn't know what the words they were using meant*"

You didn't actually quote a sentence, so it's hard to pin down what your objection was. The quote in the report that had the phrase "greater interoperability" was the editor's summarization of discussion. It followed a quote: "Doug Harr remarked that interoperability is the single biggest challenge for adopting open source at Ingres." He presumably has some experience with the problem.

The complete paragraph in the report provides more comments from the discussion.

[I share the concern about mass nouns, though most of the time I see them in text written by non-native speakers. I'm sure my French makes me sound like an idiot when I'm brave enough to use it.]

A think tank's view of free software

Posted May 9, 2007 16:46 UTC (Wed) by smitty_one_each (subscriber, #28989) [Link]

>Other worries include a lack of open source developers - their numbers are not keeping up with the growth of the industry.

Hm. No mention of Google and Summer of Code projects.
Borderline conspicuous in their absence.
That recent LKML thread about how Google carries a ridiculous share of the LKML subscription addresses might be also be interesting in this context.
It seems like there is in fact an embarrassment of riches in open source. Too many different offerings per category, rather than a lack. "Not another web framework!"
Sounds as though this report is wishful thinking for the shareholders.

A think tank's view of free software

Posted May 9, 2007 18:26 UTC (Wed) by hans (guest, #148) [Link] (8 responses)

The other top-level sponsor, naturally, is Novell; the remainder of the list is NEC, Unisys, Jasper Soft, OpenLogic, and SugarCRM. Not the most community-oriented bunch one could have come up with.

Okay, in my opinion, this comment borders on dishonesty. True, the sponsors may not be the first names you think of in the FOSS community, but some of the participants, such as Eclipse, IBM, MySQL, O'Reilly, Sendmail, Sun, and Trolltech, to name a few, have significant stakes in free and open source software development.

I'm not a big fan of the shenanigans that Microsoft and Novell have been pulling lately. I recently sold my Novell stock, which I had bought after Novell's acquisition of Ximian, because I just don't trust Novell anymore. But on the other hand, I don't think this commentary does a fair job of dissecting this report. Some of the comments in it are pretty asinine, true, but many of the worst comments come out of the brainstorming sessions. Brainstorming sessions often create a range of interesting ideas, not all of which are necessarily valid. Those ideas aren't meant to have the same weight as a well-researched report, but that context was basically jettisoned in the commentary. The commentary also ignores positive comments such as the following from Tony Perkins:

The cost of starting an Internet company plummeted by over 80% from 1996 to 2004. This trend was largely enabled by open source software and powerful, cheap hardware.

I'll just finish my rant by saying that I think that this report is interesting, perhaps even in the ways its authors intended. Although its sponsorship is questionable at best, and it's not as one-sided as the commentary seems to suggest. And I definitely don't think that this is in the same league as some other Microsoft-sponsored FUD reports.

A think tank's view of free software

Posted May 9, 2007 18:42 UTC (Wed) by BrucePerens (guest, #2510) [Link] (4 responses)

A lot of the companies Jon mentioned, like SugarCRM, should not really be referred to as Open Source at all. They use badgeware licensing.

Bruce

A think tank's view of free software

Posted May 9, 2007 19:12 UTC (Wed) by hans (guest, #148) [Link] (2 responses)

Okay, let me clarify what I meant. I am in no way defending the sponsors of the event. However, I do believe that the commentary is misleading in that it ignores the fact that there were a large number of non-sponsoring participants who, in fact, are very much FOSS companies. There is no indication from reading the commentary that any other companies are involved other than the ones listed.

I don't think Jon intended to mislead anyone, but the phrasing of the commentary was nevertheless misleading, at least in my opinion.

A think tank's view of free software

Posted May 10, 2007 4:25 UTC (Thu) by lysse (guest, #3190) [Link] (1 responses)

> I don't think Jon intended to mislead anyone, but the phrasing of the commentary was nevertheless misleading, *at least in my opinion*.

Given that it's fairly clear that the sentence as written was an expression of editorial opinion, is it asking too much that you settle at "I disagree with you" without feeling the need to call LWN's editorial integrity into question?

A think tank's view of free software

Posted May 10, 2007 15:21 UTC (Thu) by hans (guest, #148) [Link]

Given that it's fairly clear that the sentence as written was an expression of editorial opinion, is it asking too much that you settle at "I disagree with you" without feeling the need to call LWN's editorial integrity into question?

No, in this case "I disagree with you" would not suffice. I don't disagree mainly with the opinions expressed, but with the presentation of the facts. Others may disagree that the presentation of the facts was misleading, which is why I added the phrase that you emphasized. But that is not the same as saying that I simply disagree with the editorial opinion.

Should I have used the phrase "borders on dishonesty", even though that was used in the original commentary? Probably not. Although that was my initial impression after reading the report and comparing to the commentary, upon further reflection I believe that it was a reckless accusation. But that does not negate the substance of my original criticism.

However, since I did not intend to start a flame war, and have neither the time nor the energy to maintain one, let me just leave it at this: I believe that the commentary does not live up to the excellent quality and high standards that we've come to expect from LWN content.

Okay?

A think tank's view of free software

Posted May 9, 2007 19:53 UTC (Wed) by mattflaschen (guest, #45185) [Link]

I think this should be emphasized. The key line in the report for me is: "This makes OSI the official keeper of the open source brand. However, many in the commercial open source world believe that their needs and concerns are not adequately represented on the OSI board." I interpret this (based on this report, but also the OSI mailing list traffic of late) to mean the group of companies approximately equivalent to DLA Piper and the Open Solutions Alliance (http://www.opensolutionsalliance.org/) may be considering co-opting the Open Source Initiative (and thus the Open Source Definition)'s role in arbiting the meaning of "open source". Already, SugarCRM and other companies are describing the MPL+Exhibit B license (e.g. http://www.sugarcrm.com/crm/SPL) as "open source" even though it has not been approved by OSI and on the face seems clearly non-compliant. I would ask that people use open source to mean the licenses (http://opensource.org/licenses/index.html) approved by OSI. Remember, we're the market they want to decide.

A think tank's view of free software

Posted May 9, 2007 19:26 UTC (Wed) by nim-nim (subscriber, #34454) [Link] (2 responses)

> some of the participants, such as Eclipse, IBM, MySQL, O'Reilly, Sendmail,
> Sun, and Trolltech, to name a few, have significant stakes in free and
> open source software development.

IBM is less attached to its FLOSS product line than to its closed product line. Its representants typically only use the FLOSS part to pimp the expensive closed variants.

Eclipse is a consortium with a lot of big business members not especially attached to open source.

Sun is embracing open source because the market is forcing the move. I doubt the people Sun would send to a "business think tank" are especially attached to Open Source (more than to the old S for share Sun party line). Even if they were they'd auto-censor themselves, fearing to frighten potential customers.

The others are smaller fishes.

A think tank's view of free software

Posted May 9, 2007 20:51 UTC (Wed) by hans (guest, #148) [Link] (1 responses)

I'm not so much concerned with why the entities I listed contribute to FOSS (or FLOSS) as I am with the fact that they do, indeed, contribute. If market forces are driving their moves, that's perfectly fine with me.

Just out of curiosity, what companies' participation, if any, do you think would add legitimacy to a forum such as this? RedHat seems like an obvious one, but who else?

A think tank's view of free software

Posted May 9, 2007 21:31 UTC (Wed) by nim-nim (subscriber, #34454) [Link]

A company may have a few or even many developers contributing FLOSS code and not push FLOSS in business forums. You have several levels of communication:

1. developer communication: everyone will say they're bullish on FLOSS because that's a good way to attract talented developers, get contributions/partnerships, earn goodwill in customer development teams
2. internal communication: what a company thinks of itself
3. business communication: what its salespeople think will attract customers (buzzword-of-the-day)

In a business think tank you do 3. Very few entities will thing the FLOSS song on this stage. It requires an internal FLOSS commitment (2) even pure Linux players (Caldera, Mandrake) have been known to forget (another example is Novell betting the farm on Linux then doing its recent PR disaster)

Red Hat is certainly one of the few companies that integrated FLOSS from 1. to 3. Mysql is so associated with LAMP it's probably aligned too (but it's a minor player, and I never saw one of their reps). SUN is the big guy closest to Red Hat, though it pre-dates FLOSS explosion, has toyed for years with somewhat-open, and its salesforce would probably need a few great years to be convinced the new SUN FLOSS stance is meaningful (profitable). Also SUN is rather smaller than its main competitors.

The plain truth is FLOSS hasn't succeeded as a business message the way it has succeeded as a developer message (yet). Even a pure tech player like BEA is still at the "mixed source" stage.

Now there is a *large* difference between being skeptical of FLOSS as a business value and writing the kind of hatchet job this report is. It's pure vendor spin and was written by whoever paid most to assemble this "think tank" caution.

disjoint

Posted May 9, 2007 19:59 UTC (Wed) by ncm (guest, #165) [Link] (1 responses)

We have seen "reports" like this many times before. A comment on previous such reports applies equally well to this one: "Not everything in the report is old news, and not everything in the report is wrong. However, everything not old news is wrong, and everything not wrong is old news."

disjoint

Posted May 11, 2007 12:52 UTC (Fri) by sepreece (guest, #19270) [Link]

Again, remember that the report is not trying to teach a point of view, it's reporting what a group of people with large amounts of money to spend on software think about OSS. Think of it as market research, pointing at marketing problems OSS vendors have to solve.

It doesn't matter whether the participants are "right" in their perceptions or whether they are insightful; it matters that they think and feel the way they do, because that affects which kind of software they buy and install.

Note, too, that the report pointed out that this year's panel was much more favorably disposed towards OSS than previous years. I would have thought that would have made a good lede for the story...

A think tank's view of free software

Posted May 9, 2007 21:11 UTC (Wed) by sbergman27 (guest, #10767) [Link]

I must say, Novell's participation in this is disappointing. I've been trying to give them the benefit of the doubt as far as their MS deal goes.

Unfortunately, it seems that the best benefit I can give them here is that they are a large corp that doesn't really understand the concept of open source software. That would be in lieu of assuming they were actively trying to discredit it.

But for a corp that has bet their existence on Linux over the last few *years*, saying that they don't really understand the concept of open source software is pretty damning in itself.

I have this feeling that they might change strategies away from Linux at some point before going down the toilet.

A think tank's view of free software

Posted May 9, 2007 21:23 UTC (Wed) by philterry (guest, #28540) [Link]

The meeting structure and organization is of a Delphi meeting.
http://en.wikipedia.org/wiki/Delphi_method

nothing wrong with that except when the "facilitators" are "not trying to reach consensus but to instead surreptitiously guide the participants to a pre-determined outcome....

A think tank's view of free software

Posted May 10, 2007 12:04 UTC (Thu) by Dom2 (guest, #458) [Link] (1 responses)

Clearly, we would be better off with the simplicity, compatibility, and fairness found in proprietary software licenses.

You jest, yet this could be considered to be the case. Proprietary licences all boil down to You are screwed -- we own you. You don't have to think about them any further than that. Open source licences on the other hand, do require at least a passing familiarity in order to evaluate.

Complexity of proprietary licenses

Posted May 10, 2007 18:09 UTC (Thu) by dark (guest, #8483) [Link]

That may be how it works for the individual home user, who depends mainly on keeping a low profile to avoid license entanglements. In a large company, however, considerable effort is spent just on making sure that all the licenses are being followed. I don't know if this has ever been quantified in one of those TCO studies.