Slashdot Mirror
Open Source Convention 2001 Wrap-up
So, we've gotten a lot of submissions about various things related to the O'Reilly Open Source Convention. Michael Tiemann had a few words before the convention; Dan Gillmor wrote a column about it; Fred Baker made a few flameworthy remarks. For whatever reason the whole conference seemed to be dominated by the Mundie-Tiemann debate. See our earlier story for some good links, or watch the debate video, or just read some post-debate coverage here or here. And if you haven't heard enough from Mundie, you can even read his post-debate letter.
Mundie is basically trying to separate the the people who support open source software into two groups, those that think the GPL is good and will continue to support it, and those who either already don't like it or can be convinced that it's bad.
They're attacking what they see as the biggest threat to their way of business, the GPL. I see this as a sort of proof that the GPL works. It's point is to ensure that software is kept free (as in speech). This is exactly the opposite of Microsoft's goal, which is to control every bit of software you use. Don't be mistaken, Microsoft doesn't like any sort of open source software, they're simply using tactics designed to separate the whole and detroy it one piece at a time.
Check out AbiWord.
I am in the middle of a presentation where Bruce Perens is talking about the Skylarov incident. He is actually re-giving the Skylarov talk.
I want to write more about this later and listen now..
"Yes.. no matter what the culture, folk dancing is stupid." -MST3K
You are confusing "Open Source" with "Hobbyist Programming". Take a good hard look at the Apache Project. They have stable and readable source code. For their later projects, like Xerces, they designed and evaluated the whole project before writing the first line of code. An it's Open Source.
And let's look at all of the closed source crap. My current boss used to work for SGI. They had an internal motto of "we have the world's biggest QA department...our users". I've seen requirement documents written *after* customer ship. And I've seen the opposite as well: thousands of UML diagrams and no code.
Generalities are generally [sic] wrong. The "generalized" Open Source project is a hobbyist's project. But not all of Open Source is hobbyist. A lot of it is front line code written by professional software engineers.
A Government Is a Body of People, Usually Notably Ungoverned
MS complains about how the GPL erects a wall between public and private development of code. Essentially hes saying that No private company can use GPL'd code because then they would no longer be private, all their source would have to be released under the GPL. I think that the GPL is a reaction to the draconian copyright laws now on the books. A good compromise would be a limited monopoly, you can keep code private for say 15 years and then it becomes public domain. Surprise surprise, this very compromise is what copyright holders destroyed in the last century in their quest for more money. Current Copyright law tips the balance completely away from fair use and public domain, the pendulum has swung all the way to the right. Reactionary actions beget more reactionary actions. The GPL is just as unfair to copyright holders as current copyright law is to citizens who use copyrighted material. The compromise is to go back to where we were 200 years ago. Copyright is temporary but well enforced. After 15 years or so, the source gets released. This is a good compromise because it makes nobody happy, yet it allows for free dissemination of ideas, while allowing commercial exploitation of said ideas in a closed source sense.
is that if I release my wonderfull code GPLed it doesn't stop M$ from using it - it just requires them to come back to me and license it under some other terms - ie if the evile empire wants my code to do as they please with they'll have to pay thru the nose - just like they expect their cistomers to do .... but if the local high school can live with GPL I'm quite happy to give them the use of my code - and of course I benefit personally from the exchange of code that the GPL bazzar creates - I get back way more than I put in - and I get to work oin the code that I want to - none of this "OS development is off limits because we know best" sort of attitude that M$ puts forward - I much prefer Linus's "If you do something usefull that will make the OS better I'll add it" - judge me on MY merrits, not where I fit in YOUR business plan
Do you find the Mundie debate remotely interesting?
Or, did you tire of this issue a long time ago realizing that debates are an irrelevant, waste of time and that people will do what is in their best interests regardless of what the academic and marketeers think?
Someone you trust is one of us.
I may disagree with Craig Mundie, but I must congratulate him for his tenacity and bravery. He's taken quite a bit of time to explain the MS position on Open Source, and to engage in dialogue with the Open Source community. It's particularly admirable in light of the lack of respect that the OS community as a whole has given him, in spite of their public gestures.
If the OS people aren't careful, they may make Mundie look like the hero, while coming across themselves as flaming ideologues.
Toronto-area transit rider? Rate your ride.
WHY IN THE NAME OF PETE DOES ALL THIS SHIT MATTER!?!?!?!?!? The open source movement was never started to make money, so why the hell does everybody think that if it doesn't make money it will suddenly up and die!?!?!?!?!?
NEWSFLASH!!
In todays news everybody realized that opensource projects aren't making enough money, so all the programmers in the world decided unanimously to delete all their GNU/Linux tools and pay for closed source tools to make their free software with....
WHAT THE FAK!?!?!
Oh god, that woman is John Romero!
Having worked as a programmer, I agree: everything needs to be well-thought-out first.
What does that have to do with the license under which the program is released? The amount of time/effort I spent designing (or not) has no bearing on whether I distribute source with my binaries (or not). This seems like a straw man argument.
You aren't pointing out flaws in the open source paradigm. You're pointing out flaws in the design capabilities and self-discipline of most of the random one-off K3wl Projects of the Week on sourceforge.net.
No argument there; it's one way, but not the only way.
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
Mr. Mundie and the rest of Microsoft have most people chasing a red herring.
The issue is not whether or not Microsoft's software is open source. The issue is not whether or not they like or will ever use the GPL. Being a capitalist entity, neither publishing open source software nor being GPL-friendly should be expected or demanded of them. It's just not logical.
The problem that everyone is concerned about, that prohibits true competition between Microsoft and Open Source developers, is not whether Microsoft's source code is available, but rather whether interoperability can be achieved. What the world needs, and what open source strives to provide, is something that can effectively compete with Microsoft's "standards."
Currently open source software has to compete in a crippled fashion: The developers don't have access to Microsoft's file formats (Word, Excel, etc). They don't have access to the network protocol documentation (e.g. SMB). They don't have access to many hardware forums and manufacturers, because they're not willing to sign NDAs or pay money for the privilege.
What people should be asking for is a level playing field. Making Microsoft open their source code is one way to do it, but it may take away their competitive advantage in other areas (for example, if they build a faster matrix solver) and arguably isn't fair to them. A reasonable alternative, however, would be to ask them to make all of their communication protocols transparent. Once you can interact with other Microsoft users as well as you could if you used Microsoft software yourself, the game is over.
Without such a concession from Microsoft (forced or otherwise), Microsoft's monopoly power will remain unchecked: the Internet will consist of Microsoft clients connecting to Microsoft servers running undocumented application protocols over TCP/UDP. (And you thought forcing HTML email on everyone was bad enough.) If you're smart, you'll either lobby Microsoft (or better, your gov't rep) to make them open their protocols, or you'll buy Microsoft stock.
If anything, the mere suggestion that it is further proves the man's thesis. If there is one big "problem" with open source these days, it's the unreasoned fanaticism of its proponents. We have met the enemy, and it isn't Microsoft. It is us.
Yes, many times the Open Source community has produced things better than commercial endeavors. Apache and IIS is a great example. So are vim and/or emacs vs. pretty much any closed source editor. etc. etc. Perl kicks the ass out of VisualBasic every day of the week and three times on Sunday. In terms of security and ease of administration, any of the free unixen beat win32 server platforms hands down IMHO. But there are times when, well, face it guys, someone with a clue works for a company and makes something good that is closed source. ASP pages, for example, have a nice object model (just use jscript to avoid that vbscript suckage ;-) ). Anyone who things linux/*bsd is a good OpenGL development platform has never used Irix. etc. etc.
In the end, information technology is only useful in the extent that it makes people's lives easier or more entertaining in some way[1]. Whether the machine code came though gcc or msvc++, the end user does not care one whit. We are carpenters and stonemasons, only our raw materials are bits instead of wood or stone. So just pick the right tools for the job at hand, and leave the fanaticism at the door!
[1] Or to stick it to The Man. But that's besides the current point.
(I fully expect to get modded down for saying this stuff, but fuck it, I have 50 karma so I care not a fig for the slings and arrow of outrageous moderation.)
--
News for geeks in Austin: www.geekaustin.org
News for Geeks in Austin, TX
I found the repeated references to a 'Software Ecosystem' in Mundies letter amusing. I am supposing this goes along with the use of the term 'viral' when refering to the GPL. I can just picture a healthy MS landscape, where everything works together in harmony.. little 'VB script' bees pollinate outlook encouraging the growth of communication ("uh.. hey everybody - Don't open any mail from me that asks for your help, or says I love you..). Text Farmers have heathly Word crops with only the occasional paper-clip infestation.
Then..
GPL WILDFIRE! Good lord! its heading towards our profits! The crops will be ruined! Won't somebody *please* think of the children!!
If only those damn university could leave well enough alone, and ONLY allow certain businesses access to source code, this could all be prevented. Lets not waste anymore public money on projects that can't be closed up and profited from by poor lil' old MS..
air and light and time and space
In other words, Microsoft is mad because it wants to "build" upon free software. Of course, once they build upon it, they will have to use their huge marketing force to crush anyone else who might want to use the same free software to create their own program in the name of "innovation".
I would agree that some OSS projects lack proper design, but the majority are designed quite well - a necessity when you're dealing with colaboration on such a large scale (as opposed to a team of 6 programmers). Such large projects as the Apache web server would never have atained the popularity that they have had they not been designed in a way that would allow management of such large scale colaboration. In fact there are OSS projects that are monuments to exceptional software design.
Granted there are some young projects with bad design but you generally don't have to be worried about having ot use those products because the projects generally don't reach maturity, for reasons of managability.
OSS doesn't just have good software design it requires good software design, much in the same way as OSS promotes an enviroment of cut-throat competitivemess among programmers, instituted through an effective system of massively scales pier review. Toy are only as good as your last piece of code, and the OSS project will only survive if the underlying design is exceptional.
This as distinct from treditional corporate software design where pier review is frequently limited to 6 guys in a small room critiqueing your code, and software design is driven by customer deadlines; where software designers, after failing to succeed in one project will be transfered to another and another. Try that in the OSS comunity. If you fail on one public project, it's highly unlikely that you'll be invited to participate in another project. Such is the world of OSS and this is why it will always be a part of the software universe.
--CTH
--Got Lists? | Top 95 Star Wars Line
I just read Mr. Mundie's message; although I understand where he's coming from, I'm not sure the ecosystem he describes
can really pan out. For instance, it's important under his model that those who take from the "commons" give enough back,
but in 1999, for example, Microsoft paid $0 in federal income taxes--large companies in the U.S. do a very good job of
creating ingenious tax strategies, those are the breaks. But it does seem that there's a disturbing trend with our
society "socializing the risk, privatizing the profit," as Doonesbury put it. Now, I'm a software developer, not an
economist, but it seems to me that with Microsoft (and others) paying little or no taxes, hiring as many talented
academics as will take the offer for working within Microsoft Research et. al., then patenting as much as they can to
wall off new branches of technology...well, sheesh, how long can the commons sustain this sort of aggressive
over-grazing?
Turning to the GPL, I really like its emphasis on making software part of humanity's knowledge base, so that advances are
never lost, and even poor societies have access. This guaranteed openness points the way toward a computing ecosystem
like the medical sciences, where scientists not just in the U.S. but around the world share their findings and build on
each others' work. *Implementing* humanity's open knowledge store of medicine is an enormous field that pays many
millions of people, often quite well (my wife is a physician, so I have special insight there). The key word is
*implementing*. I think we'll see this open ecosystem for computing come to pass, thanks to the GPL and the advantages of
making software development a worldwide, not company proprietary, endeavor. This new world will probably be a little
weird for Microsoft at first, but I am confident that everyone will find their niche, and we'll all wonder why it took us
so long.
Dan Gillmore and others observed that Mundie came off like a true statesman, compated to Tiemann. There is a lot to be said for statesmanship indebates of this sort. There is the PR advantage, the potential to catch your oponatnt off guard, and the potential to build alliances. Mundie and Microsoft as a whole has generall looked reasonable and fair (to outsiders) in the ongoing OSS discussions of the last 6 months. This is potentially vary damaging to the OSS position. If we can't get it together, as was done somewhat back in the beginning when Microsoft launched their attach on OSS through Mundie's first speech 6 months ago, then we might as well pack it in because public opinion is the end all and be all of public debates.
While I completely disagree with Mundie's position, I aplaud him for his professional handling of the debates at this conference. I never thought I'd suggest using Microsoft employees as examples but i this case we should learn from Mundie's example.
--CTH
--Got Lists? | Top 95 Star Wars Line
Miguel Icaza have just post this at mono-list:
Hello guys!
I made a presetation at the O'Reilly Open Source conference on the Mono project, shortly after David Stutz talked about the Shared Source implementation of the ECMA C# and CLI that they will be releasing.
Interesting things from David's talk:
* The terms of that shared source license are still not ready, and will likely be different than those from Windows CE.
* They expect to have something by the middle of next year.
* He confirmed that it will just be the core of the system, and will contain a JIT.
He made my life easier by explaining to the audience what the CLI was, which was helpful as I did not have to go into too much detail on the remaining time.
We had a good talk about the CLI afterwards.
The slides for my talk are available on the Mono site (I believe I already sent a mail about this) but in case I didn't, just go to http://www.go-mono.com, and you will find the link to the talk slides.
There were some good questions, like how we will avoid patents if there are any on the ECMA specification. Our answer is that we will stick to use old technologies: things that have been documented or written about in the past in the various areas where the CLI and C# matter: intermediate languages, standards for type systems, traditional optimization, garbage collection in the ways that Java has done for multi-threaded operation, traditional compiler instruction selection.
For those cases where we incur in a speed penalty, we will research alternative ways to implement things to not infringe on their patents. This is particularly useful for those of you who are studying and need to write a thesis, as we have a research project you can work on.
I also got a chance to talk face to face to Sam, and we discussed a bit about possible ways of improving the runtime. One thing that came to mind is that it would be possible for someone to work on a number of projects: retargetting an existing Java compiler (I am familiar with Guavac, and seems good enough) to generate CIL instead of JVM byte codes.
I am now flying to Ottawa for the Linux Symposium. I will try to make releases of the runtime, the class libraries and the compiler on Sunday or Monday when I get back to Boston.
Best wishes,
Miguel.
useful links
Sorry for such a big submit.
-=-=-=-=
I know life isn't fair, but why can't it ever be un-fair in MY favor!?
Furthermore, copyrighted and GPL'd works are available in source form, which is the most useful form for creating derived works, but since Microsoft's products though copyrighted have hidden source code, even when their copyright expires no-one is likely to be able to make derived works, as the source code was never 'published'.
This is perhaps a legitimate point, and a clear definition of a derivative work should be incorporated into the next version of the GPL.--
The real Webmaven is user ID 27463. I don't rate an imposter, because my ID is such a lame-ass high number.