Slashdot Mirror
Succeeding With Open Source
Implementing an open source solution requires a different approach from the buyer's point of view: There is no salesperson you'll get a call from, there are no license agreements to sign and no serial numbers to enter. Access to the software is simplified, but sometimes there are few pointers about what to do next.
Golden takes the reader through different aspects of dealing with open source technologies. As one can see from the table of contents, the information is presented from the business professional's point of view. This title is for an IT manager, not developers or IT personnel who might be using open source products already and feel strongly about them. The basic question that the book explores is this: When does it make sense for an organization to implement an open source product? How do you evaluate the product's maturity, functionality, ease of use, support infrastructure and documentation quality so that running open source within the organization starts making sense?
Golden's answer is the Open Source Maturity Model (OSMM), which the author developed himself. The model asks the IT manager to evaluate the software, support, training options, documentation, integration and professional services on 10-point scale. If the technology ranking reaches a certain score (which highly depends on the userbase), then it will make sense to implement it.
For example, on page 144, when the author discusses software support options, he suggests assigning 6 points for excellent community support, 3 points for available paid support and 1 point for availability of self-support (i.e., an employee who understands the product). So on the next page JBoss gets 6 points for community support (very helpful and respectful forums), 2 points for commercial support (since it was e-mail and phone only, and no on-site support) and 0 points for self-support (since no one within the organization stepped up to claim herself as JBoss expert).
JBoss is the prime example used by the author throughout the chapters, and turns out to be quite a convenient choice -- the company offers commercial support, training and documentation for an open-source product. Golden's model is supposed to help IT managers distinguish high-quality open source projects from 0.0.1 version, so widely available on SourceForge.
The book's primary market is business professionals and IT managers who would probably benefit from having a formal evaluation model instead of relying on pure gut feeling. Despite the book's ambiguous title, it's not a manual on how to create your own business with open-source products. Some chapters will be helpful for figuring that out (Chapter 2 talks about business models in the open source world), but it's mostly for people who are implementing rather than developing open source products. The language is somewhat dry, but if your weekly reading requires CIO Magazine, you're probably used to that.
Something I think the author would have done well to include is a collection of in-depth case studies on open source implementations. There's some data on Sabre and Charles Schwab running successful businesses on open-source infrastructure, but the details are not there. While certain companies publish hundreds of case studies to prove that their products will either save money or allow the customer to make more, the success stories are not that frequently publicized in the open source world. Having such material in the book would provide a confidence booster for an IT manager, I think.
The last chapter or the first appendix is where I would expect to find information on solid open-source products suitable for corporate deployment. I mean, if the evaluation model is introduced, why not list the most prominent projects out there for quick reference? The highest-ranked open-source operating system, office suite, corporate messaging system, accounting and tax package, etc.?
Overall the book is pretty good for a manager who has heard of open source, but has not read too much into it. Chapter 1 in PDF format is available from Addison Wesley site. Golden also wrote an article for OreillyNet that deals with bringing open source into the organization. There's also an interview with the author on TechTarget.
You can purchase Succeeding With Open Source from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
1. Write book about Open Source.
2. Get review posted on Slashdot.
3. ???
4. Profit!
Sorry, I had to.
The answer, of course, is SCO SysV Unix source!
Q: 'Open source is great, but where do we start?' A: Write it.
Does everything include nothing?
Ugh. More seminars for the seminar weary.
-- (Score:i , Imaginary)
This may be an overly simplisitc way of looking at this, but my answer is this. Start by making sure open source software will do what you want. If it doesn't, then stick with the propietary stuff.
I like big butts and I cannot lie.
RTFM for starters.
An Indian-American Hindu committed to non-violent thought/speech/action alarmed by the global explosion of radical Islam
I think that the efforts of many big companies to introduce Linux into the corporate world in the past couple of years and the years to come, are, while greatly appreciated by most OSS developers, could be harmful in the long run to OSS in general. It's no question that every OSS fanatic wants nothing more than to see OSS on everyone's server and home computer, and lately a lot of companies have expressed similar interests. I think, however, that the motives should be questioned here. The motives of OSS developers are quite clear to most, and the motives of large corporations are clear to most as well. And anyone can see that, though joined by the desire to have more secure and publicly avialable software, the two ideological motives are on the opposite ends of the spectrum. The corporate interests are always clearly monetary, and that has always proved to harm anything it comes in contact with. While the GPL, and a slew of other licenses and agreements is there to protect open source, there will be many more cases like SCO if open source fully enters the corporate playground. Let's hope at least that the people at IBM, Novell and Sun understand what "the spirit" of Open Source is about and act with respect.
First, Open Source or Free Software is not alone about free access to the source, it's also about helping each other. Even when in commercial settings this is not an option in all aspects, it can still help.
For example there is the Open Source Observatory -- when you do not trust OS zealots, you may trust more independent sources. There are documents (like the Open Source Migration Guidelines), case studies and events where you may meet others with similar questions.
In short: look around and get a clue. Helps enormously and makes much more things better than just software.
Here, an example of a stable commercial company is IBM. Linux is an example of category #3. OpenOffice is an example of category #2. (Sun is not a stable company, and its survival is questionable.) Many applications offered through the free software foundation (FSF) fall into category #1.
Companies debating whether to implement an open source solution should first determine whether there is open source software that meets their needs and that fall into any category except category #1. Category #1 is just too risky.
As for categories #3 and #4, the backing of a stable commercial company means that any serious problem arising in the software will be resolved.
Some software (category #2) may not be backed by a stable commercial company but is already fully developed, has been stable for a long time, and is no longer being revised. We can trust that such software is stable enough for deployment at a commercial company. The Mosaic browser, the original web client, would be a good example.
I'm expecting that this book is well on its way to being transformed into a broadway musical version: "How to Succeed in Open Source Without Really Trying".
1. Superior customer service/satisfaction 2. Lowest possible overhead/marketplace prices 3. Technological/industry leading features These are successful tenents of an online business and OSS makes these possible. Especially #2 - how can you have lowest possible overhead for a business when you pay thousands in license fees yearly to MS? Or the constant reboots of servers/desktops that waste productivity time in the workplace. MS just doesn't make sense.
Same reason I go to the monkey house at the zoo.
Because I find it endlessly entertaining to rile up a bunch of simple minded creatures and watch them fling feces against the glass in vain.
Now, dance, puppet. I command it.
Hang on a second here. ...
(Sun is not a stable company, and its survival is questionable.)
Sun had a 2003 profit of $312,000,000. But then you go on to say that Linux has the backing of stable commercial companies. What Linux companies, might I ask, are more financially sound than Sun?
I don't respond to AC's.
In most industries, I'd guess, open source is a moot point. There simply is no open source software comprable to the vast majority of business related applications out there. For example, there's no open source point of sale product that even comes close to what's out there in the proprietary market.
I don't respond to AC's.
Perhaps it would help if you entered Sun's actual ticker symbol of SUNW rather than Sunoco's which is SUN. You'll find SUN's earnings were -much- lower.
Yeah, SUN is definitely a good example of Linux success:
Sunoco, Inc. operates through its subsidiaries as a petroleum refiner and marketer and chemicals manufacturer with interests in logistics and cokemaking. Sunoco's petroleum refining and marketing operations include the manufacturing and marketing of a range of petroleum products, including fuels, lubricants and some petrochemicals. Sunoco's chemical operations consist of the manufacturing, distribution and marketing of commodity and intermediate petrochemicals. The Company's operations are organized into five business segments: refining and supply, retail marketing, chemicals, logistics and coke.
I think I'd go even further and say that the ranking SHOULD be applied for any software you do not write yourself - how many times has a company purchased software from a vender and then finds NO-ONE (not even vender professional services!!!) knows how to run/extend it?
In any software you are considering to buy or use, consider the community support above all and see if you are going to be able to get help from anyone outside the company or vender you are getting the software from.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Take a look at Sun here.
Sun's revenue (not profit) was 11.18 Billion. That's in Dollars, not pesos.
Furthermore take a look at cash reserves. 3.60 Billion (Dollars, not pesos). With only 1 Billion in debt, that leaves more than a little bit left over...
Free cashflow was -20 Million. That's just about a rounding error compared to the actual sales volume.
In short, you have to be completley daft if you think Sun is going anywhere for a very, very long time. To call them "unstable" and "likley to fold" is to proclaim yourself ignorant in the strongest terms.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
You forgot the absolutely most common open source project category:
5. New software currently being developed by one programmer without backing from anyone.
This is by far the largest category. Often this one programmer has rather interesting ideas about user interface and the requirements of the user community toward which he is targeting the project. And, yes, "interesting" is meant in exactly the same way as the Chinese curse "may you live in interesting times."
This argument really only applies if you're determined to be dependent on external entities for maintenance and support over the long haul. The main point with Open Source is that the source is, umm... open. Hence the market for M&S is similarly open (and competitive), and you are free to decide whether to handle it internally or externally as you see fit. No forced upgrades or other common forms of extortion that arise from monopolistic control of code. Whether a big, stable company "backs" it may not be as material as you think, and to the extent that such a company comes to control the software, it could even be a liability in some cases.
1: Insert CD or your favorite disto
2: Run install
3: Never pay M$ tax again
4: For support read the online info on your favorite distro.
A better question to ask is: "What provides the best solution to this problem?". If open source provides the best option, great, go with it. If being open source itself provides a solution to a problem, say making sure that the software can be updated in house, for example, that's a different story. In real life, you pick the best tool for the job. Sometimes that means commercial products, other times it doesn't.
RandomAndInteresting.comdefending the world from stupidity since 1979
"Open source is great, but where do we get support?".
Seriously - for the project I'm on now the client has insisted we use Weblogic, even though all we're using is the servlet container. Resin or tomcat would have done the job just as well for a fraction of the cost, but they want the big-name backing and support contract.
(Yes, I realise that that's often a false sense of security - don't tell me, tell the client.)
It's official. Most of you are morons.
it's unlikely that software with that level of maturity would attract the attention of corporate IT - at least the attention of management. And the obvious solution in the rare cases where it does - hire him.
#!/usr/bin/english
This is one of those OSS myths, like the security one discussed here the other day. The fact that a large project's source is open doesn't even slightly mean that just anyone can come along to develop new features or bug fixes. The barrier to entry involved in learning a large project is huge even if it's well-documented and impeccably designed. Without the original dev team around, you're basically sunk. If there is an exception to this rule, I've never seen it, either in OSS or commercial development land.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
Thinking of the tool before thinking of the goal is kind of ass-backwards.
For all I know, you could be operating a $50 a day hot dog stand that may not even need any software. Seriously, whatever business you're into, your business objectives will be unique, and those business objectives should be fleshed out before you invest anymore emotional time into a particular tool.
Yes. People who choose technologies are sheep -- show them the flock and they'll follow.
Google on "____ success stories" where ____ is your favorite open source technology.
Linux, apache, python, and perl all turn up useful case studies.
Check out this website, which contains another OSMM. However that one is Open Source and a year old already.
So where is your savings now running Winders?
Yes I do know about deployment of complex packages on a large scale and haven't had anymore problems than deploying the same on Winders. I guess keeping up with all those client licenses is no problem for you either....