What is Open Source?

s390 writes "The Inquirer is running an article by Olliance about "What is Open Source?" It appears to be the first of a two-part series for managers about how to engage with the open source community. The writers seem to know their material. Are they on target or have they missed something important? Do PHBs really need to read this sort of introduction to get comfortable with the idea of using Linux and other open source software?"

Comments

Like it or not, managers default to commercial

[Sheetrock]

Which means that Open Source needs to be carefully pitched to them -- commercial, closed source software is how business has been done for thirty years, while free software is still kind of a wierd new hippiesh thing (although having big-time companies like IBM embracing it helps.)

It's like the saying goes; when all you have is a hammer, everything looks like it needs a pounding. Nobody ever got fired buying Microsoft, whereas Open Source seems like a gamble... even moreso now with the bogus SCO lawsuit. Calmly and rationally explaining to the people that make the procurement decisions at your company that free software is a valid alternative and explaining why is necessary, because by default they're going to want to go with what they've always went with.

Re:Like it or not, managers default to commercial

[Cereal Box]

Interestingly the concept of the company fixing its own problems as they hold the source was just unthinkable. No manager would give themselves more work no matter how much money it would save.

See, you have the misconception that simply having access to the source means you can fix problems in an acceptable amount of time. If you've ever had to pore over a massive project you'll realize that you could spend a couple days trying to track down and fix a problem that is preventing you from finishing your real work or you could simply send a problem description off to the guys who originally wrote the code (and who you're already paying for), and they'll probably be able to fix the problem far quicker than you ever could. THAT is the reason closed-source software is so appealing. If you have a problem with a piece of code you bought you can tell them "hey, I've got a problem, I'm paying you, get on this problem and fix it for me ASAP or we'll be re-evaluating our business relationship with you." With open source software, you might get lucky with helpful developers and you might not. You never know. What you do know is that you don't have the ability to really push open source developers to fix problems for you, and that's a major drawback.

I loved antitrust

[zapp]

Antitrust beat the hell out of movies like Hackers and The Net.

Atleast Antitrust used GNOME, real unix commands, etc... had hot girls, and an actual plot.

Maybe it didn't portray the fineprint of the GPL, but it did pretty well (I thought) at showing how corrupt corporations can be.

Sorry

[BillsPetMonkey]

But on theinquirer, this is preaching to the converted, and the linux side banner "blindingly easy" betrays a lack of objectivity in approach. I'm not trolling, but this article won't be noticed by the people who really need to see it.

In the area of marketing the "linux zealot" tag is our own worst enemy and unfortunately that's what this will be labelled.

Not obvious

[Jad LaFields]

As a someone who is relatively new to /., the 'geek scene', and alternative software, I'd just like to point out to many people I don't think that OSS is a very simple, obvious concept. I haven't finished reading the article yet, but I think the idea of explaining OSS simply to 'non-geek' people is a good one, considering the "if you don't know what it is, figure it out yourself, we're not pandering to anyone" attitudes I've seen every once in awhile. Well, off to finish reading...

assumptions

[thoolihan]

The author assumes some things. High Quality? Some software is. OpenSSH for starters. However, anyone who's spent some time installing packages on various GNU or BSD systems knows there are some really awful projects out there too.

Now that I think about it though, thanks to the Java version of ICQ, I think closed source still holds the worst app of all time record...
-t

Missing the Point

[Delphix]

While I read slashdot daily, and have an appreciation for Open Source software, I think a lot of you miss the point. I work for a company that builds 3D simulations. For years we used IRIX. SGI was simply the best graphics rendering stations you could buy. We still have some fridge sized boxes, an old Octane, and a couple O2s sitting around.

However, we've moved to Linux. We use Red Hat 7.2. Why? Because Open Source is great? Beacause we're giving back to the community? No. Because Linux allowed us to reuse all our simulation code that was built using Unix interoperabilty on more modern hardware, cheaper. We can go grab a GeForce4 Ti 4600 off the shelf, get a 8MB Buffered IDE HD, and Dual Athlon chips and they'll eat those O2s / Octanes, and even the old fridge sized Onyx. Slap six of them together and you've got a six channel renderer. We did have to change our IG software. We were using Performer, now we use a third party developed renderer.

For years our the company that developed our renderer has supported both Linux and Windows. Now they only support rendering on Windows. They still support the API for Linux though. So now we're looking three options:

1) Devoting time to developing our own Linux based renderer.
2) Continuing to use their API on Linux, but to control chanels on Windows boxes.
3) Converting the simulation over to Windows and dropping Linux.

At this point option 1 is pretty much out. While we have the in house skills to develop an IG, it would take several years to build something to meet our requirements (needs texture paging, terrain paging, must handle terrain sizes in excess of 2GB, etc).

Option 2 looks attractive because it's the least work. And will probably be what's implemented in the short run.

Option 3 will probably be our long run solution as we've had some trouble with nVidia and ATI drivers on Linux. While they do work in most cases, they don't seem quite as robust as their Windows counterparts. Ie with the Quadro cards...

Now if you go back and read over these things, what we considered in deciding what to use were:

1) time
2) cost
3) ease of implementation

You'll note we never said anything about Open Source. When we first moved to Linux we looked at Open Source issues. We use open source development tools. GCC, GIMP, ImageMagick, Glade... But it had nothing to do with the fact they were open source. If they had been sold to us like IRIX was and performed the same functionality we would have bought them.

We're interested in shipping our product, making $$$, and that's what managers are concerned with. The debate over whether or not we use Open Source software is irrelevent. The questions that have to be answered are how it would impact cost, time to delivery, learning curve. The other thing is, you shouldn't expect businesses to go out and just switch to Open Source because you wrote a persuasive article. If there is a system in place, it probably won't be replaced until it's necessary to do so. It's that whole cost thing again. Why replace something that's been paid for that works until it's necesary to do so.

If you answer that with anything other than it will make the company more $$$, or increase productivity, etc... you won't get very far with the suits.

One misconception I had

[jcsehak]

One misconception I had was that open source meant you had to give the product away for free. This was even reflected in version 1.0 of the Open Source Music License I wrote (that I based on the GPL). But that's not so! You only have to give the source away, not the end-product. And you don't even have to make it available for download, you need only sell people CDs of the source for the cost of the media.

Open source isn't so much for the benefit of the end-user as the developer. Or rather, other developers. So it's just as easy to make money on an open source project as a closed one, as long as someone else doesn't take your code and make their own version that better and cheaper. So for MS, open sourcing Word would be a bad idea. But for a musician like myself, "better" is relative. So making your music open-source does nothing but good.

Education, lots of education

[heironymouscoward]

Sometimes what is most obvious to one person is most unobvious to another. OSS is - to Slashdotters - as obvious as water and sun. To managers, it is a contradictory concept and much of the education has to come in the form of explaining why it is "free". Much of the counter arguments to OSS try to leverage the natural paranoia of business people (GPL virus, OSS patent liability, TCO, etc. etc.).

The simple truth - and this is obvious to any neutral observer - is that the Internet changed the dynamics of the software industry by removing all the previous barriers to organization and communication. What used to require a large organization to produce can now be economically (and this is the key) produced by random professionals in their spare time, by tiny teams working on thin margins, and by organizations who would otherwise do nothing special with the code anyhow.

The key questions about OSS ("who pays?") have been asked before, about the Internet, and we now all know the answers: everyone pays, a little, but since the technology required is so very, very cheap, it comes down to rewarding people's time. And it so happens that for many developers, the product is its own reward.

Software only represents a slice of any business's operation, and whether OSS is free or cheap makes little difference. However, when software is expensive (think SAP or Oracle), it eats so deeply into the business that little is left for other investment.

Managers need first to learn that large swathes of the software landscape are now completely in the domain of "as cheap as air", and only foolish people will still pay for something that their competitors get for nothing.

Managers need secondly to learn that this process keeps on moving. One by one the bastions of commercial software will become commodity items. The businesses with the capacity to be pioneers will always benefit. The rest will follow when the technology curve flattens out and moves into the "general market" and then "late adopter" phase.

OSS is simply a slice of a standard S-shaped technology curve. It's the same curve that drives Moore's Law, and indeed, one can say that software cost has a half-life. One could even estimate this. Allow me to state "Heironymous' Law" of falling software costs: every 18 months, software products fall in cost by half, eventually reaching effective zero.

Which is why we can now get databases, office suites, Beowulf clustering software, etc. etc. etc. for the cost of bandwidth.

Patents, by the way, are a brake on this technology curve, and this is IMHO why those of us who love and feel this curve hate patents so much. Technology wants to be free.