Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Programming Businesses GNU is Not Unix IT Technology

Evaluating Open Source 110

Posted by michael from the but-what-about-phase-2? dept.
CowboyRobot writes "Jordan Hubbard cofounded FreeBSD and now oversees the Darwin implementation of BSD for Apple. He describes open source as 'finally being openly acknowledged as a commercial engineering force-multiplier and important option for avoiding significant software development costs.' And thus, companies need to know how to evaluate open source engineering as an option for them. In a new article titled Open Source to the Core, Hubbard goes through a typical open source adoption process."
This discussion has been archived. No new comments can be posted.

Evaluating Open Source More

Evaluating Open Source

Comments Filter:

  • avoiding significant development costs.. (Score:4, Insightful)

    by js3 ( 319268 ) on Tuesday June 01, 2004 @05:09PM (#9309218)
    at the expense of?
  • It points out a number of stumbling blocks that you might get into and walks you through them.

    It has a few paragraphs on dealing with the various liscenses, and on the effort you should put into giving back to the open source community if you use some of the code.
    • That is key, as I don't think corporations are even considering that aspect of open source software. Considering that, in all fairness, efforts should be made to contribute back to the community when using open source software, I think that the alleged TCO benefits begin shrinking drastically. I hope this has been mentioned somewhere, since it needs to be a reality. I don't want to see anything hindering the overall adoption of open source, but I also don't want to see the open source community being abu
  • I can see this as a great opportunity to both mainstream developemnt and provide more options. Id software being able to more between mac/windows/linux on their releases is a good example of this.

  • But why (Score:2, Interesting)

    by Anonymous Coward

    Why would you want to cut down soft.dev. costs if an engineer in India costs $400/month?

  • Open source marketing. (Score:3, Insightful)

    by demonic-halo ( 652519 ) on Tuesday June 01, 2004 @05:23PM (#9309363)
    The sad thing about open source is there isn't really any marketing control.

    Linux hasn't really taken off into mainstream unti IBM started throwing it's weight and marketing Linux.

    If someone could figure out a open source way of marketing and marketing studies to fuel product development, then we'll see a new era.

    • Re:Open source marketing. (Score:2, Funny)

      by Anonymous Coward
      Christina Aguillera dancing in a see through pink blouse with the tropical paradise on the background should fuel the sales of new kernel, me think.

    • Re:Open source marketing. (Score:5, Insightful)

      by flossie ( 135232 ) on Tuesday June 01, 2004 @06:17PM (#9309891) Homepage
      Linux hasn't really taken off into mainstream unti IBM started throwing it's weight and marketing Linux.

      IBM started throwing its weight behind Linux because it was taking off.

      • Re:Open source marketing. (Score:2, Insightful)

        by Anonymous Coward
        Correct. Linux was already mainstream on the server front before IBM announced its "Billion Dollar backing" of linux in Dec 2000. I will give IBM big credit though for their very public backing Linux. There is no doubt in my mind that they have played a big part in spreading the gospel. I can't however give them credit for brining Linux to the masses. That's just not right.

    • I don't think that's such a sad thing. It'll be actively marketed in one way or another as long as someone sees a way to make money from it. IBM has found such a way (or believes that it has), but even if it stops then linux and open source will still be there for me to use --- complete with all of the enhancements that IBM provided.

      I realise that it's not exactly what you're referring to, but in the past week or so I've been hearing Microsoft commercials on the morning radio, definitely peak time o

      • When I say marketing, I wasn't just refering to commercials, but the whole process.

        Who's doing the studies on what features people want and how to prioritize them? Who's deciding whether the software should be imaged as a high quality/high performance, or low cost alternative. Engineers don't necessarily know what's best for the customer and open source software can sometimes get over bloated really quickly.

        It's pretty cool IBM and other companies like Red Hat are collaborating on the Linux front. Now
  • Odd... In a five page article on open source, he mentions Linux 3 times -- once wrt KDE, once wrt Gnome, and once wrt Slashdot. That's it.

    • The reference to Slashdot was suprising to me being referenced in the ACM without even a footnote because not everyone in IT knows about Slashdot, especially your average .NET programmer:

      Marketing. First and foremost, your marketing people will (or should) want to have a prepared message about your use of open source, even if it's only to respond to any questions that may come up. Make sure that they also know enough to make correct assertions about it, or you may find yourself paying the price on Slashd
      • The reference to Slashdot was suprising to me being referenced in the ACM without even a footnote because not everyone in IT knows about Slashdot

        Hopefully, anyone reading an ACM journal should be able to type "slashdot" into google.

    • Suprise! (Score:5, Insightful)

      by Proteus ( 1926 ) on Tuesday June 01, 2004 @06:13PM (#9309866) Homepage Journal
      In a five page article on open source, he mentions Linux 3 times -- once wrt KDE, once wrt Gnome, and once wrt Slashdot. That's it.

      Maybe because:

      • Linux is already well-known
      • Linux was /not/ the first open-source product
      • There's a lot more to OSS than Linux

  • Sound Planning For ANY Migration (Score:4, Interesting)

    by Eberlin ( 570874 ) on Tuesday June 01, 2004 @05:29PM (#9309427) Homepage
    "investigation, evaluation, adoption, and communication"

    Isn't this true for just about every migration plans?

    Investigate -- find out if this will do what you want it to do.

    Evaluate -- dig deeper into the idea. Get a better feasibility study with numbers and monetary figures. Make cool looking presentations to the higher-ups that sign the checks.

    Adoption -- this is where you SLOWLY incorporate the new with the old. Make sure everything is working well. People may have to do double-duty to work with both systems just so they can give it their blessing (that it all works properly). This is where you train a "core" group of support folks from each department so they burden you less.

    Communication -- this really should be earlier on, before adoption. Find people who run this stuff already and communicate whether it may work for you too. See if you can get a "we'll help you through it" before you even adopt.

    Again, this isn't anything strictly for Open Source. I'm sure there are nuances and cultures, yadda yadda yadda...but a good plan of action helps minimize risk with ANY project.
  • - Be prepared to become an expert on everything. If you have problems with component X, if no one in the community is interested in fixing it or if you're under time pressure, you'll have to fix it yourself. Also known as the "if you don't have a kernel expert on staff, you will eventually" rule.

    - Almost nothing works the first time. OSS engenders infinite flexibility which eventually reaches infinite permutations. The plethora of configuration options to a large project's source can be very daunting. Everything interlocks with everything else for maximum flexibility which means more work up front to understand how the pieces fit together.

    - Forget about binary portability. OSS is designed to support source code across platforms in the same way Windows is designed to support binary backwards compatilbity.

    - Expect complexity and plan for it. OSS supports every platform under the sun which breeds extra complexity.

    - Have lots and lots of patience.

    Just my two cents from having developed embedded x86 and ARM Linux for the last two years.
    • Lest I seem to negative up there, I should also mention that in using OSS you will learn more than you expect. It's a lot of fun to be able to mess around in USB device drivers one week then dig into a linker bug the next week. With time, patience, and persistence, you'll be able to understand how all the stuff actually works ("oh, that's how a shared library is loaded from disk into memory"). It's the keys to the kingdom!

      • Re:a few extra notes from someone using OSS (Score:2, Funny)

        by Anonymous Coward
        It's a lot of fun to be able to mess around in USB device drivers one week then dig into a linker bug the next week

        Great for you. Unfortunately, it's a real pain in the ass to have to mess around all over the supposedly working code. And I have my own problems to deal with, more than full time. That's my job. I don't have extra time to deal with a bunch of buggy messes left by others for the "fun" of doing extra work before a deadline.

        oh, that's how a shared library is loaded from disk into memory
        • You don't build bridges with Erector set parts.

          Well, in a way you do. Most complex products (cars, ASICs, software, maybe even bridges) are built these days by buying/licensing/borrowing/copying designs from someone else and kludging them together. The added value is in picking the right parts and getting them to work together, just as for open source.

          Not that I think the open source market is properly developed. There is no really good way yet for you to pay $x to get the linker quality improved, oth
    • It helps to attract real programming talent too:

      Shop A: Pays $65K for a VB and ASP .NET developer to work on Win 2003 and MSSQL/IIS.

      Shop B: Pays $65K for Python developer to work on Debian with PostgreSQL/Apache.

      Who do you think will get the better programmer?
      • I agree. In my experience over the last several years, I've found people who use OSS projects tend to be more self-starters, curious, and technically adept.

        I joke that you learn a lot with Linux, et al, because you *have* to. Show me someone who is running Linux (or BSD) at home and I'll show you someone who knows and likes computers.

      • Re:a few extra notes from someone using OSS (Score:1, Insightful)

        by Anonymous Coward
        The language used has NOTHING to do with how good of a programmer you are.

        If anything, the time savings offered by microsoft's very good development tools allow you to focus on a better architecture and UI.
        • The language doesn't make you a good programer but the environment does make a difference.

          It's much easier to be a bad programer and get away with it in a good environment. While I wouldn't call Python, Debian and PostgreSQL/Apache a bad environment. If you are able to set it up and program in that environment it means you have to have a certain level of skill. On the other hand to setup ASP.NET, Win 2003 and MSSQL/IIS doesn't require anywhere near as much skill. That doesn't mean that the program isn't as good but it does mean the skill range is larger.

          Eg. If you compare someone who can just do VB with someone who can do C++ and VB you'll find that the C++ person will often be better. Why? Because C++ is harder than VB which means more people can program VB which means your more likely to find someone that's not as good if you look for VB.

          That's why there are alot of Crap VB programs around which is good. Lots of people learning and it's easy to throw a program to gether but it's also bad lots of people that can throw a program together think their good. Compare that with C where it's alot harder to throw a program together an while some of them as still bad most of them a better than the VB ones because it's harder so you have to be better to get it to work. Drop the C programer back to VB and they can still write good code just faster.
      • You might as well say that someone who speaks English is automatically smarter than someone who speaks German, French, or any other language. The task might be better served by a VB and ASP.NET programmer, you can't tell. Many OSS-friendly developers use those (or other) tools by day and OSS tools and software by night. Keep an open mind.

      • Re:a few extra notes from someone using OSS (Score:4, Insightful)

        by reverendslappy ( 672515 ) on Tuesday June 01, 2004 @08:33PM (#9310824)
        While I would say "B", as I'm assuming you would, I don't know that you're right for the reasons you might think you are.

        While I'd agree that a Python/Debian/Postgre/Apache developer is probably more adept, I don't think it has to do with the language. It has to do with the fact that (generally speaking, of course) OSS people are more heavily self-taught amateurs-turned-pros. To me, that displays a passion for the craft that others might not have (though to be fair, many MS-based developers are self-taught too, albeit on systems that are much less in-your-face from a learning perspective... OSS developers have to spend a fair bit time learning the systems first, before the development skills, while MS'ers don't necessarily). Add to that my opinion that autodidacts have skills that are generally more flexible and adaptive, and "B" is definitely preferable.

        But the differentiator is not the language. In reality, while B is better than A, a developer that can excel at both A and B is better than either an A or B; a truly gifted develeper isn't limited by language. Overall, B is more desirable to me because I know a B has likely invested more time and passion in learning and honing their skills, not because they know <insert language here>.
        • Actually, that was exactly my point.

          I do ASP+javascript+CSS+(x)html programming and VB for a F500Corp with ADO and Access/MSSQL on IIS + win2k.

          I also do python and postgres and mysql and php on debian (sarge), redhat (not so much these days) and xml-rpc. (consulting, some custom projects, including multipoint remote backup systems for media (can you say multigig uncompressible files?))

          I can tell you I like the latter better of the genuine mindbending.

          I don't even know where my asp book is, since I've gr

      • > Shop A: Pays $65K for a VB and ASP .NET developer to work on Win 2003 and MSSQL/IIS.
        >Shop B: Pays $65K for Python developer to work on Debian with PostgreSQL/Apache.

        >Who do you think will get the better programmer?

        IMHO, it's anyone's guess... You've neglected a multitude of other variables: understanding the business, usability, clear design, ability to get things done, etc.

    • Wow I also have been working on embedded ARM and Linux for the last two years (and just 'normal' server-side Linux before that, since the minix release), and I was just about to post the same basic points until I read yours... uncanny.

      With regards to linux and fixed-purpose computing (embedded is a bit cliche a word, these days...) I would say that the one thing to always remember is that it ain't running until you've built some hardware ... ;)

  • The Problem With Darwin (Score:4, Insightful)

    by Anonymous Coward on Tuesday June 01, 2004 @05:48PM (#9309622)
    I like the idea of Darwin, a free Unix-like OS specifically designed to work well on a Macintosh, and had hoped that it would be developed as a real and useful BSD but from what I can see, it is simply the core of Mac OS X and no one seems to be interested in making it useful OS on its own right.

    Maybe things have changed since the last time I tried Darwin but there are a few problems with it, such as:

    1. No partitioning/formatting options during install

    2. No way to setup Airport

    3. No way to add users/groups without knowing arcane NetInfo commands

    4. Some commands do not make use of the full console dimensions; probably because no one wants to fix Termcap.

    5. No security announcements lists or patches.

    6. No binary releases have being generated since 7.0.1.

    Furthermore, I have seen people who wanted to use Darwin as a server (on a remote Macintosh) told to use Mac OS X Server instead. It seems to me that this is the wrong attitude, that people should actually want Darwin to be useful as a server and and a Unix workstation. It is a shame.
    • i think the deal is that darwin and os x are kept in sync, and so adding the things you want to darwin (making it a 'real and useful' bsd) would also have to get added into os x, bloating it unecessarily. (yes i know. genie effect. it's all bloat anyway :))

      i mean, why duplicate an airport driver in the core system when the actual product has a really good one? why create an intuitive user/group system when the actual product is really damn good at managing them? i think the point of darwin is to maintain

      • Re:The Problem With Darwin (Score:1, Interesting)

        by Anonymous Coward
        "i mean, why duplicate an airport driver in the core system when the actual product has a really good one? why create an intuitive user/group system when the actual product is really damn good at managing them? i think the point of darwin is to maintain a really slick foundation for an operating system, not an OS all its own."

        Maybe I feel Darwin should useful on its own. Otherwise, there is no point in it being opensource other than to get free engineering for Apple.

      • Re:The Problem With Darwin (Score:4, Insightful)

        by Brandybuck ( 704397 ) on Tuesday June 01, 2004 @07:10PM (#9310296) Homepage Journal
        Actually, Darwin is "an OS all its own". Don't fall into the Microsoft trap and make your definition of OS too broad. Darwin might not be a complete graphical workstation environment for non-technical users, but it certainly is a complete operating system.
    • I like the idea of Darwin, a free Unix-like OS specifically designed to work well on a Macintosh, and had hoped that it would be developed as a real and useful BSD but from what I can see, it is simply the core of Mac OS X and no one seems to be interested in making it useful OS on its own right.

      Because there is no point in wasting time doing that. There is nothing (significant) Darwin could offer that OS X [Server] doesn't already.

      • There is nothing (significant) Darwin could offer that OS X [Server] doesn't already.

        Wrong: Darwin offers freedom. For many folks, that is significant. Those folks might like an OS which offers the features of OS X without the proprietary code.

        Of course, I'd just use Linux, but that's me. I don't know if Darwin offers any value over *BSD.

  • Nothing new about open source (Score:5, Insightful)

    by weekendwarrior1980 ( 768311 ) on Tuesday June 01, 2004 @06:01PM (#9309750) Homepage
    Back in 60s and 70s, the era of huge collaboration that gaves us UNIX, Internet etc, everything was open sourced. Of course, the targeted audience who participated tended to be those in academia (outside the corporations that developed them). Guess what? Open source softwares was then and there and yes, it was viable enough to be an academic experiment and commercial at the same time. I think the only thing different now is that we have the same revolution with a wider audience and a sensibility that will sustain open source movement for a long time to come.

    • Re:Nothing new about open source (Score:2, Insightful)

      by Anonymous Coward
      The reason it was free was because nobody thought of software as a business. IBM certainly didnt care about software, except as a way to sell hardware. For the better or worse, a software industry was born, and if big companies like IBM keep giving away software for free to sell the hardware, a lot of the software companies and those who make a living developing software will have to go away... our high tech economy in the past few years is living proof of this.

      • Re:Nothing new about open source (Score:1, Insightful)

        by Anonymous Coward
        For the better or worse, a software industry was born, and if big companies like IBM keep giving away software for free to sell the hardware, a lot of the software companies and those who make a living developing software will have to go away...

        Sure. Just like the universities all collapsed as soon as people started giving away knowledge and teaching materials over the internet.

        Oh, wait...
    • The target audience is a critical difference between what happened in the early days of computing and what the open source movement brings to bear. The open source movement targets business with its message because the goal is to help business be built (and that is certainly consistent with the message of Hubbard's article here). Early computer software development was available to anyone who wanted to learn but the goal was not to help build a business. As the first sentence of the open source definitio
    • Don't forget that in the 60's at least, the costs (and profits) in the computer industry were overwhelmingly on the hardware side. IBM et al made most of their money from selling mainframes and support services, not software.

  • Alright lads, we're going to get people to develope this for us, open source style... then we're going to charge for it. We'll make millions for pretty much free.

    Sounds like that to me... isn't the point of open source to "Give a little back to recieve alot". A "One for all, all for one" approach to software?

  • Open Source: Best damn thing for the economy (Score:4, Insightful)

    by stealth.c ( 724419 ) on Tuesday June 01, 2004 @06:45PM (#9310113)

    It lowers the barrier of entry. Small local businesses can thrive in an environment like this, and anyone is eligible to obtain the necessary knowhow and skill to get a job or start a business in the field.

    I have pretty much one criterion in my mind regarding economics in the USA. If it ups the barrier of entry: automatically bad. It divides the haves and have-nots into perpetually irreconcilable camps. If it lowers the barrier of entry, any perceived "loss" or "recession" is due to the fatcats getting outdone by nimble startups or their own customers. In other words: automatically good.

    Lowering the barrier diminishes corporate power; diminished corporate power means diminished corporate influence on government; and that means more power to the REAL PEOPLE.

    • Re:Open Source: Best damn thing for the economy (Score:1, Interesting)

      by Anonymous Coward
      What I see wrong in this argument is this. Company A charges $100 for selling a product that paid its developers, marketeers, artists, accountants, sales people, people who burned the CDs, boxed it, handles shipping, etc, etc Now here comes the OSS product, giving it for free, download it and install and off you go happily as a freebee. Lets say the OSS developer got a consulting deal and even earned a few bucks. Now this OSS project has been successful in destroying the otehr software company, which in
      • The F/OS project has helped--it has helped every single user, who now has the freedom to hack on the code. Yeah, most of us don't bother, but some of us do: that's how free software improves over time.

        Just look for the real puppet masters behind OSS before you contribute your own src code to this plot to take the power of the programmer.

        The 'power of the programmer' is the power to enslave the user. As a user, I don't want to be enslaved. As a programmer, I do not wish to enslave.

      • If a company can't adapt to new circumstances, and therefore becomes obsolete, it has no right to be in business.
      • whats the difference in cometition with OSS and a consulant making a dollar and some other comercial businaess coming in and selling prodects and getting a consulating gig?

        cometition is competition, If you complaining about who you are competing against then odds are you have been doing somethign wrong in your business model and arer scared others will find out. (ie. charging too much and getting away with it because there was no competition.)

        This is really no different the a 7-11 store opening up across

  • I fully understand that this article is coming from an open source advocate and therefore it will reflect that movement's philosophy. The same thing could be said of Mark Webbink's article about licensing [groklaw.net] and much (if not all of) ESR's articles. But I don't think that means there is license to misstate history. Hubbard notes:

    The Free Software Foundation launched the GNU project in 1984 with the initial aim of creating a complete operating system environment (the GNU system). It may not have succeeded

  • I describe open source as good enough for IBM, Cisco, Oracle, HP, Amazon, Yahoo, etc etc etc.

    Open Source is good. Get over it.
  • Have Americans (I assume he's American) finally done "leverage" to death?
    Is "force-multiplier" the mot-du-jour?

    FP.

Slashdot Top Deals

"If value corrupts then absolute value corrupts absolutely."

Close