Is "Open Source" Now Completely Meaningless?

Nat is building the OSCON schedule, and asks: Is “Open Source” Now Completely Meaningless? From the post:

I’ve found some very strange companies claiming to be open source.

Can you really call yourself open source if you haven’t opened the source? I don’t think so. There’s a flood of “open source” companies selling things that work on open source but which aren’t open source themselves. I think these are proprietary products, not open source. That’s been the attitude that helped me select talks for OSCON–I only want open source products talked about. My rule of thumb is that the audience should be able to download, compile, and use the software that is talked about.

He then goes into what are definitely some edge boundary cases, of which EnterpriseDB is a good example. I agree with his conclusion that in the end EnterpriseDB and companies like that are good for Open Source and should be included in an event like OSCON. A lot of it comes down to intent. You see, being “Open Source” is the in thing right now. That means some companies are using Open Source as a dumping ground. Donating old code is fine, if there is some real value and a chance that a legitimate project might thrive around the code. Donating old code because it is completely useless and you just want to get into this Open Source thing, that’s not. It’s easy to tell the difference. At this point, we even have some companies simply lying or using the words Open Source arbitrarily. Open Source means slightly different things to different people to be sure, but twisting the definition way out of whack to suit you is not something you should do. You end up with situations like this. Now intent is something that is hard to ascertain in some situations, but I can’t think of a better yardstick. You’ll always have the idealists who think something is or isn’t acceptable vs. the pragmatists. Free Software vs. Open Source isn’t going away soon. We’re all in the same larger camp though, and intent is one of the few things I can see us all agreeing on.


The Free Desktop

Bryce Harrington has posted an insightful commentary on running free software communities. A couple highlights:
It's easy to just copy what the proprietary guys do. After all, we like to think they're doing scientific market research and usability studies to uncover user needs. Or at the very least, users grow accustomed to the way the proprietary apps work, so if you make a good clone, you can get into the “well, at least you won't have to bother learning anything new” niche. That's actually a pretty good niche to be in, but at best it's a nibbling-from-below strategy.
However, in my view FLOSS has a secret power for achieving exactly what users need built right into the model. I'd credit a large share of Inkscape's success to simply being utter whores at taking advantage of this.
Basically, the best way to get what user's want and need into your application is for the developers to BE users.
Users have a gut-feel for what they want and need, and gut-feels are tough to express in requirements documents, bug trackers, and surveys. You know when you're developing some thingee for your own personal use what corners can be skipped, and what ones are worth polishing; you know what compromises can be made, and can immediately notice the value of some unexpected potential as the code's developed.
The reason open source has this special power is because of its openness, transparency, and (usually) ease of getting involved. Proprietary companies by their very nature are characterized as secretive and controlling, and do not consider individual users as able to contribute anything of value to their product but money.
FLOSS projects have a different value model than proprietary ones. Value in FLOSS comes not from users spending money, but on users and developers who are contributing enhancements (code, docs, translations, word-spreading, etc.)

So to the point – how can the free desktop succeed? Just cloning OSX or Windows isn't going to be compelling enough. Competing on a feature matrix against proprietary companies is hard since they can pull some new feature out of hiding at any point.
No, the key to success with FLOSS is its core attribute – it's openness to community-scale collaboration. Focus on energizing a strong community and empowering new users to bring their new ideas and energy in. Give them lots of room to experiment, and assist them with paving a good foundation for them to build atop. Give them invitations rather than rules, and infrastructure rather than bureaucracy. Make it easier for people with little technical skill to participate, and treat them as first class citizens when they do.
Don't think of the free desktop simply as an alternative to proprietary desktops. Instead think of it as a platform for large scale open collaboration. For us technical users sharing code, it's already hands down the best collaboration platform; the next step is to work to make it just as good for non-technical users, who are driven to collaborate on everything from clipart, encyclopedia articles, music, schedules, designs, fonts, education, games, politics, and so on. Take the rules, tools, architectures, and lessons we've gained in software collaboration and recast them for “data development”, and I think the free desktop's success will be impossible to stop.

He's absolutely right on multiple points. It's critical that you make it as easy as possible for people to get involved. Part of that comes from making it clear that you don't need to be a ninja coder to be helpful to a project. Be an evangelist, write some documentation, contribute to the projects wiki or submit good bug reports. A user that can post high quality bug reports that include all needed info and clear instructions on how to reproduce the bug are worth their weight in gold, even if they are technically unable to provide a patch. Getting community involvement is something you'll see in almost all of the so called winners. As Bryce mentions, it's sometimes easy for large projects to become insular. That can be, and in fact has been, extremely poisonous to a project and needs to be something the project leaders are cognizant of.
His point about cloning proprietary software is also extremely apropos. Sure, in some cases it's necessary or even desirable. But, it should not be the default action. Working to our strengths (and we have many) will get us much further. If you are considering starting a new software project and will be in charge of the community, you should read the linked article… digest it… then read it again. Community is that important.
BTW Bryce, I couldn't find a link to your RSS feed anywhere on your site or in the HTML source. If you don't know how blosxom lays things out, you might not even realize you had a feed.

Rob Enderle; Linux: Cuba and the Communist Connection – Promoting "Fear"

(via Matt Aslett): I haven’t blogged about Rob Enderle for quite some time, but he may have outdone himself here. In this post, Enderle warns that the Cuban adoption of Linux could lead to it being considered anti-American. It doesn’t stop there though:

Linux, which isn’t really a product in my mind, as it co-defines the Open Source movement, is a brand. This brand has attributes that have been created over time and generally, at least for those that use it, represents something positive.

For some it represents the freedom to look at and mess with code, for others it represents “free” as in free beer, and for some it represents a strong weapon against capitalism, particularly with regard to software. I’m sure we could add some additional attributes, but the one that concerns me is this new concept that it is anti-American.

Can you imagine the NSA IT manager trying to get funding for a Linux based project right now? We are dropping into an extended election where the war in Iraq, terrorism, and nationalism are likely to be major battlegrounds. On the economic front, China remains a huge concern and in many battles it too is likely to come up, particularly as these battles are fought in areas where unions are strong.

If I’m running against an incumbent (who probably has no clue about software at all) and I know some organization under them deployed Linux and that it is being positioned as anti-U.S., might I not use that in the election? “Ladies and gentlemen I promise that under my administration we will not implement products like Linux that put the nation at risk, which contribute critical technology to the terrorists, and embolden our enemies.” In politics all you need is a grain of truth. In fact, sometimes I wonder if you even need a “grain.”

So he contents that using Linux could lose you an election. If that is not the ultimate “what if” scenario stretch, it has to be close. By that yardstick, Microsoft is responsible for nearly all spam, since it is cracked Windows zombies that is the main source. Quick, someone use that in an election. The basic premise here is so absurd that it’s hard to even argue against it in a rational manner. One statement I do take more serious issue with is:

One of the big problems with Linux, as I see it, is that certain topics (basically anything that may imply that Linux isn’t the best thing for every possible use) is off the table, because that discussion creates FUD.

I don’t see anyone saying Linux is the best thing for every possible use. I’ll go one better – it isn’t. There is no silver bullet in life, no panacea. That includes Linux. I have found that in a large variety of places, from embedded to big iron, Linux has made sense for a variety of reasons from security to stability to scalability. Does that mean that Linux is the only answer or even always the best answer? Of course not.

He ends with:

Linux will likely survive this latest internally created challenge but, at some point, someone better take ownership of the Linux brand or others will position it and supporters will probably not like how it is positioned.
What do you think Linux users should do about this? Who should own the Linux brand? And do you agree that Linux should not be used as a political weapon against anyone?

Linux isn’t a “brand”. Red Hat is a brand, as is Novell or IBM. It is however a trademark and that trademark does have a licensor. There is even a shiny “Report Abuse” link right on the homepage. That being said I don’t think the first question people asked when the US went to war with Iraq was: “What OS are they running”. After seeing Rob speak at DLS, I came away with a sort of new found respect for him. He slang little FUD and was a bit more cogent then in most of the items I had read previous to hearing him. I was so surprised, I even blogged about it. I don’t know what happened since then, but he seems to have fallen back off the deep end.


Hopefully my last Novell Microsoft Agreement Post; or Nobody reads this stuff

I wasn't going to post any more about the Novell Microsoft Agreement. I don't have anything else to add and really what else can be said. Don has posted on this issue after Novell contacted him as a result of him comparing the deal to “eating a bug for money”. A couple items in the post are too good not to share:
From Novell's point of view, the deal is supposed to be about giving customers some peace of mind over being sued for patent infringement. But the document that is supposed to explain this to customers is missing some information, and the documents that are actually supposed to contain the missing information are secret.
I don't think this is some big conspiracy, just the result of rushing something through without reading it very carefully. Novell and Microsoft are too big to pull off a conspiracy, but they're both bureaucratic enough to do paperwork that ends up being full enough of dumb mistakes to be pretty much meaningless.
Simple things first. If you still believe people read this stuff, try reading it. Covenant to Customers has two obvious mistakes: Novell spelled “Novel” and the awkward “on account of a such Customers' use of specific copies”, which looks like the result of someone starting a singular-to-plural edit and not finishing. The page has been up since November 2 of last year. Not a big deal, but like I said, nobody reads this stuff.
Now go back and read “Covenant to Customers” again. You get a bunch of Capitalized Words, some of which show up in the definitions section, and some of which don't. One of the terms is “Clone Product”. You don't get sued for running a Covered Product, but Covered Product doesn't include Clone Product, and Clone Product isn't defined. So which of the products included in the SUSE distribution are Clone Products? I asked, but that list is not available.

Not too encouraging when the company name is spelled wrong in the document. I can appreciate that novel is a word, so will pass a spell checker, but it would certainly seem to validate the nobody reads this stuff claim. He then goes into specifically what the Patent portion of the agreement protects you from:
You have a Covenant, so you're all squared away, right? But the Covenant says it covers Covered Products, and Covered Products doesn't include Clone Products, and you don't get a copy of the definition of Clone Products. Covered or not? You're just as mystified about the answer to “Will Microsoft sue me?” as you were before.
Another good one: “Microsoft reserves the right to update (including discontinue) the foregoing covenant pursuant to the terms of the Patent Cooperation Agreement…” Yes, the Patent Cooperation Agreement that you don't get to see.
Put it all together, and you get, “We promise not to sue you for running some but not necessarily all of the software you get from Novell, unless we stop promising not to sue you, and we won't tell you which software you get from Novell we might sue you for now, or under what circumstances we'll stop promising not to sue you for the rest.”

In others words, not much. His conclusion is right in line with what myself and almost all others have been saying:
That sense of wanting to win by product, not by litigation, is what protects you in the short term. What protects you long term is that the IT industry is just a series of recruiting contests. A company wins recruiting contests by serving good fajitas and letting an employee get a $70 laptop power supply without spending $200 in time on getting approval to buy it. A company loses recruiting contests when it's seen as succeeding through legal attack, not great product. Who wants to work on a product that customers have to be sued to buy? If a company's litigiousness makes it miss a round of recruiting, it misses the round of product that those recruits would have built, and game over.
Anyway, the check is cashed, the bug is eaten, Microsoft seems equally likely to sue you for running Linux no matter where you get it — not very — and if all goes well, a lot of that money will go to the many helpful, hard-working developers at Novell who are doing useful work. I guess what I learned here is that the patent deal doesn't really give Novell customers any special assurances.

Microsoft almost certainly wasn't going to sue you before. This agreement makes them no less likely to sue you. In reality, it probably did give Ballmer more fodder, which will have some impact. If you do see Microsoft start to sue, to me it means they think they are worse off internally than we think they are from an outside perspective. That's not saying a while lot. Companies with viable business plans and sustainable products don't depend on litigation for future growth. Just ask SCO.

Show Us The Code

Just stumbled across the Show Us Your Code site. I think it's well intentioned, but misses a few things. First and foremost (and take this as constructive feedback please), if you are going to publish an open letter that you want to get taken seriously, you really need to grammar check it. My grammar on this blog isn't always great, but that's because I write it in kind of a conversational style – ie. this is how I talk. When I hand in articles that will be published, you may notice the style and tone is quite different.
On to the content of the site. First, it's not really the code we need from the angle I see it. It's specifically the patents that they feel we have violated. That's a small but fairly important distinction. Also, looking at It lacks logic, especially when you consider that there are developers around the world who would be more than happy to work with Microsoft to resolve this issue. From the view point of Ballmer, it doesn't lack logic (although it may be lacking from an ethical standpoint). You see, he can bluster on about this topic without showing any actual code or naming any patents and in the Enterprise just that slight chance of being sued in the end makes a difference. In the past, it made a huge difference. The reality is that people are wising up to this and the impact now is substantially decreasing. In the near future, it should be negligible. Look at the SCO case. It didn't stop many people from adopting Linux (although I'm sure it did stop some). It did however stop many companies from talking about their Linux adoption. That should not be underestimated.
Now, I've said before that I doubt Microsoft will move forward seriously with any patent cases. It simply leaves them too open to a MAD-like patent war situation and they don't want that. I've always found it odd how Microsoft has been positioning themselves around IP and patents in the Open Source world anyway. Looking at past litigation, I'm not aware of any Open Source company that has been found guilty of infringing other peoples IP anywhere near as much as Microsoft. I'd like to be corrected on this if I am wrong, but the recent $1.5B (yes BILLION) judgment is just one of many against Microsoft. You think they'd realize just how broken the system is. You think they'd want to fix it. But the reality is, they have the money to lose. The long term threat of losing their advantageous monopoly like position is surely much more worrisome.

Who writes the Linux kernel?

An informative post over at LWN (via Matt Asay) about the code contributed to the 2.6.20 kernel, followed by the last year of kernel changes. The result? As far as companies go, Red Hat is by far in the lead as far as both number of changesets and number of lines go. Other top names are what you would expect: Novell, IBM, Linux Foundation, Intel, Oracle, etc. It's good to see that Red Hat can now back up the claim that they are doing a significant amount of work in the kernel space. It's interesting to note that “Unknown” is in first and “None” in third. The definitions are: “the line marked “(Unknown)” is exactly that: patches for which existence of a supporting employer could not be determined. The line marked “(None)”, instead, indicates the patches from developers known to be working on their own time.” That does give credence to the often mentioned statement that some real percentage of kernel work is done by volunteers. The article also looks at individual contributions. If kernel hacking is something you're interested in, the full article is something you'll want to read. LWN continues its tradition of putting out quality in-depth articles like this and is worthy of your support.

Ballmer repeats threats against Linux

Novell execs must cringe when they see things like this:
Steve Ballmer has reissued Microsoft's patent threat against Linux, warning open-source vendors that they must respect his company's intellectual property.
In a no-nonsense presentation to New York financial analysts last week, Microsoft's chief executive said the company's partnership with Novell, which it signed in November 2006, “demonstrated clearly the value of intellectual property, even in the open-source world.”
The cross-selling partnership means that Microsoft will recommend Suse Linux for customers who want a mixed Microsoft/open-source environment. It also involves a “patent co-operation agreement”, under which Microsoft and Novell agreed not to sue each other's customers for patent infringement.
In a clear threat against open-source users, Ballmer repeated his earlier assertions that open source “is not free”, referring to the possibility that Microsoft may sue Linux vendors. Microsoft has suggested that Linux software infringes some of its intellectual property, but has never named the patents in question.
Ballmer said: “I would not anticipate that we make a huge additional revenue stream from our Novell deal, but I do think it clearly establishes that open source is not free, and open source will have to respect the intellectual property rights of others just as any other competitor will.”

He almost makes it sound like the real value to Microsoft, and the real intention of the agreement, was simply to posture for further protection money from others. I wonder how Novell feels about that in retrospect. Matt Asay asks: “Steve Ballmer: Was this the friend Novell wanted?” I think the answer to that is now clear.
I still wonder if Microsoft realistically thinks they can sue. The amount of potential litigation that could get thrown back is substantial. Red Hat seems to be one of the more likely Microsoft targets, but don't forget that the likes of IBM depend on Linux sales for large chunks of consulting money. It's even more interesting now that Oracle is selling what is in essence a RHEL clone. At this point in the game, I wonder how effective this kind of FUD slinging really is anyway. A few of years ago, many people were fooled. These days, that's certainly changing. Ballmer, it would seem, is not changing with the times. It's unfortunate, as some parts of Microsoft seem to be be attempting to.