Slashdot Mirror
Key Advantage of Open Source is Not Cost Savings
cmcsonar writes "Computer Economics recently conducted a survey of visitors to its website regarding the perceived advantages in the use of open source software. Although not a scientific sample, the results are nevertheless startling."
Of course, with the way vendors rape their customers with outragous rates for changes, and removing "features" so that the system works the way they asked it to work in the first place; it's no surprise at all that the first thought is about vendor dependence...
Since it wasn't mentioned in the summary, I'll post it here. The key advantage they found was less dependence on vendors. <flamebait> Something Linus recently found out :) </flamebait>
Myself, I use KDE on Linux because it gives me the best environment to code in. I used to use Windows, and have a Mac OS X laptop, and find them both awkward compared with KDE. I really don't get why they are considered miles better for the desktop than Linux. Linux was okay for me on the desktop eight or nine years ago, and it's come on leaps and bounds since then. I'd happily pay for Linux, but I wouldn't pay for Windows.
With all commercial software, I spend huge amounts of time just looking at if things are compliant or now.
Can I move an install to another PC and not break the license?
Can more then one user use the software on a PC without problems?
Will license structure XYZ or ZYX suit a particular company better in the long run?
do i get the lite version or premium version?
will it's copy protection/activation become a problem?
All this is totally gone with GPL licensing, the answer is basically I can do whatever bar sell it (In my case I dont modify and code, so that doesnt come into play).
I also find the quality of open source products much higher then that of commercial software, irfanview I reccomend to anyone wanting to make minor changes to digital pics, and in batches, works well and is free.
I think it is most important that the ROI be measured in an effective method. Such as, not only look at the obvious costs, but look at the hidden savings from changing to Open Source. Such as, we are running Pentium II computers for a year longer since we are running Linux, which extends the life beyond the cycle of expected depreciation. We can cycle in upgrades to hardware in cycles to prevent a one time expense on the balance sheet.
Then cover things like the amount of power saved with the older machines using less watts. For some companies, this could be $100,000+. EnergyStar has statics on this information.
I would also mention the recent losing of the source code for Windows along with the ability to break free of recurring charges with virus software.
In the grand scheme of security, it would probably be beneficial to note that spyware and corporate theft is less likely in a system that is unfriendly to script based theft schemes.
Mention that you don't have to worry about paying for MCSE for employees. You have no fears of employees stealing licenses.
No more formatting when a new employee inherits a machine.
The ability to disable Cd Drives remotely at will.
I guess that covers the basic things. I would give them all copies of Linux LiveCDs that they can take home and use on their home machines. LindowsLive is a good one to use. Let them see for themselves that it is not going to be a foreign OS, but just a slightly different OS.
To confirm you're not a script, please piss in my ear.
It has been at least three times in last 4 years that I have seen our company to struggle with dependence on a software vendor and there has been huge efforts and significant resources (10+ developers working on internal product) just to reduce dependency on unresponsive vendors. Its pain to ask for new features or just simple bug fixes in timely manner. We even offered to do them ourselves, but since there is no access to the code... no luck. Its very frustrating and if its some software that is critical for your company, this can prove to be a major pain.
-
SQL Server 2005 Enterprise
$24999/CPU
- Windows Server 2003 Enterprise Edition pricing:
$3999/CPU
- Google, estimated to have 100,000 dual-CPU servers
Total = 100000 * ($24999+$3999) = $2,898,000,0002.8 BILLION DOLLARS!
What bugs me is paying permium fees and getting all that. It's one thing to be handed the package and knowing I'm on my own. It's another thing to be paying good money for "support" only to still find out I'm on my own.
Unfortunately, the presumption of that lock-in period is what justifies a company's initial startup costs. Without it, it's nearly impossible to get VC funding.
/. readers make in their economic pundrity is thinking that everything ends when a company becomes profitable. But realistically, you're not successful until you've made back all the money from the initial investment plus 10% ROI to cover the opportunity cost. Plus, if the majority of companies will fail then the successful ones need to make twice as money in order to still give the investors a 10% average rate of return.
The biggest mistake that
Let's face it: lock in is just smart business. Ignore it at your peril. If you're not always fighting to keep the customers that you have, then you're going to have a lot more time & money to spend on the ones you don't. If you found a company based on some idealistic notion that lock-in is bad, then you are going to fail just like any other two-bit company with no business sense.
-a
is that I find that I get even better support with open source. There have only been a few times that I could not go to Google, bust out a simple query, and find a whole forum of people who would help me through a problem within a couple of hours.
.conf files, man pages, and other documentation, while Microsoft "support" has a script that they are seemingly not allowed to deviate from.
Sure beats the shit out of sitting on hold with Microsoft for 2 hours, only to get grilled and having to convince them that you are not trying to steal product, only to get charged for support that ultimately ends up with fdisk/format.
Granted that not all of those problems are Microsoft's fault, but in my experience, they could have done some freakin troubleshooting before telling me to backup, reinstall, and restore. At least the F/OSS community will have an extensive reference to
Maybe I am wrong. Maybe the advantage is that F/OSS tends to me more modularized, and thus you are more likely to rescue an installation by fixing one component... Thoughts anyone?
bash: rtfm: command not found
with the Moodle LMS, as opposed to commercial Learning Management Systems's.
With Moodle, the free support has been very much better than the support that comes with a paid Blackboard or WebCT license.
And another nice thing is if you need it you can get paid support from a variety of partners, so if you don't like the paid support from one partner, you can choose another without having to switch LMSs--with the closed source systems there is only one source of support--the license provider. If they cut support to boost quarterly profits, you're SOL.
Since switching LMSs is a huge deal for a school, being able to choose from a range of support services is a pretty nice feature.
But you have to choose the right product--look for one with a vibrant, open, active community where the core developers participate often. With some open source products, the support is no better than Microsoft--they tend to be the ones where the developers don't participate in open discussion, where the community is asking alot more questions than are getting answered, etc.
Other great features are scaling clusters without added license costs, being able to test new versions extensively before putting into production, being able to run multiple versions without having to pay multiple fees, and of course bugs are fixed much more rapidly and generally just by changing the code directly without having to apply a 'patch' or shut down the system.
> If you're talking about, say, a small business
> that needs basic desktop machines, the overhead
> in say, going with an OS OS will far exceed any
> price savings.
And what sort of "overhead" might that be? A modern Linux distro practically installs and configures itself, comes with boatloads of software, and does not require an advanced degree to sit down and start using for everyday (non-development) purposes.
My last Linux installation took about 1/3 the time of my last Windows installation (on the exact same, very recent hardware) and the Linux installation included setting up hardware, networking, and installing many common personal/small biz apps such as office suite, browser, email, IM, etc. The Windows install did not include any of these "extras", all of which must be done *in addition to* the OS install for a Windows box.
The last time I installed Windows, it took me roughly a day and a half to have everything ready to roll so I could get some work done. The Linux installation took maybe a couple of hours to achieve the same goal. In spite of the fact that I have about 8-10 times more experience using Windows than I do using Linux.
Sure some of the apps are a little different, but most of them have a little "Help" clicky-widget in the program menu just like any Windows app does.
My experience is that the myth that Windows has a lower TCO than Linux is just that, a myth. *Particularly* in the SOHO space.
Il n'y a pas de Planet B.
It'd be interesting to compare user-agents with those who voted for vendor lock-in. It may not prove anything substantial, but it'd be an interesting tid-bit.
What's odd is that the headline of that article captured the results perfectly, but the article failed to explain it properly.
I can back up this story with a similarish one, but I don't need to post AC, because we're proud of what we've done.
I work at Friendster, and we have... ah... a really big database cluster. It runs MySQL. Not that Oracle didn't try. They sent out sales people to convince us to convert over. After we looked at the dollar signs, we laughed them out of the office.
I was interviewing a candidate for one of our sysadmin positions. He said something along the lines of: "Well, now you're running MySQL. Once you start making money, do you think you'll start using Oracle or something else that scales better?"
I laughed and said exactly what parent AC said: "Oracle scales in theory. But in practice, 99% of businesses can't afford to scale with Oracle. I can build another couple terabytes of DB storage in a redundant replicated cluster tomorrow for $10k with MySQL. With Oracle it'd be 10x that much, if I were so lucky." That's not to mention the overhead of calling their sales guys, licensing hassle, and other crap. With MySQL, you install and go.
There are other huge advantages MySQL has over Oracle and their ilk. Take this for example... Right now MySQL AB tech support is stellar. Front line support knows when to escalate to the proper engineer (InnoDB problems? Two hours later, Heikki Tuuri is emailing you!). I remember talking to a PHB a year or two ago, and he said: "Well, MySQL support may be good now, but that'll change. It'll get bad."
My response? So what? Then I'll find a MySQL support shop that has good support and use them. They can support MySQL just as well as MySQL AB can.
Try that with Oracle. "No, Oracle, I hate your tech support. Starting tomorrow, I'm going to have Sybase support our Oracle installation." Oracle will laugh at you, then double your support costs for your insolence.
fifth sigma, inc.
Yeah, sorry I posted as an AC; but we ended up going bankrupt largely due to stupid spending encouraged by one of the execs and the VCs. Congrats on Friendster's success and smarter leadership than we had.
"Try that with Oracle. "No, Oracle, I hate your tech support. Starting tomorrow, I'm going to have Sybase support our Oracle installation." Oracle will laugh at you, then double your support costs for your insolence."
On the other hand, I have heard of IBM global services supporting Oracle on Solaris with Intel based Windows clients, despite having competitive products to each of those. Of course if Oracle started crashing they'd certainly be unable to fix the problem (no source code access would do that) and probably just refer you to the DB2 sales team. :-)
And my experience is the exact opposite. So what is your point?
I do agree that most competent distros are usable for standard desktop applications (internet, office, etc). However, I don't think it's fair to whine about Windows not coming with any applications. You can't have your cake and eat it, too. First you complain that Microsoft has an unfair advantage with bundling their apps. You want them to be forced to unbundle their apps. Then after they unbundle their apps you complain that Windows doesn't come bundled with any apps. I think Microsoft did have an unfair advantage in bundling their apps, but I'll be damned if my government is going to force them to bundle competitor's apps.
More often than not the root of these TCO estimations favoring proprietary OS are attributing the cost of switching and / or re-training. But I believe this is blaming the chicken for getting caught by the fox.
Has anybody analyzed the TCO of hiring skilled vs. non-skilled people for the IT department?
I suppose what you might have meant to say is that Linux gives you choices of these extras, and were installed through various packages off the cd, which there were probably several of due to the added bulk of all the software you could ever need.
What sort of work do you do that requires IM btw? It seems to me that it might take 2 hours tops to install xp, an office suite, browser, e-mail, etc. , unless you count the updates, which you didn't mention (if you did I would understand more where you're coming from, updating sucks!).
Umn... that really depends on the distro. Having friends who are gentoo advocates, I can promise you that some distro's aren't the kind of thing you can just download and install and run. On the other hand, mandrake^H^H^Hiva is about as basic as they come.Overall, yes, some linux installs are easier than windows installs, assuming you like the bundled software. But then again, I happen to have Windows XP Install ISO that has been updated to SP2, which would GREATLY cut back on the install time.
You make a good point, but, and I'm sure you know this, transition cost are a one off. Sure once you switch to OpenOffice.org you are in effect locked into it by the training costs you staff would have to under go to switch. But what you aren't locked into is where you get your servicing, your upgrades, bug fixes and perhaps most importantly your archives of documents will always be readable. Don't like the company who is supporting your software? Find another company. They all have access to the source code. Not to mention the support is often better with open source software because it's what geeks use and like to talk about (for example on forums). If a company turns around and says to you (in effect) "your bug isn't a priority, it's part of a small section of our system and we will get round to fixing it in six months". With propriatary software you are locked in. It might be costing you thousands a day because of a simple bug. And you can fix it. What if the bug is simple. With open software if your primary vendor gives you the finger, spend a few grand, fix the bug yourself, heck make a PR exercise out of it by telling your customers you are more reliable because when things go wrong you can fix it. Sure it cost you some cash, but less cash and you can extract some benefit by appearing dynamic or powerful. And thats really the bottom line here. Companies aren't looking to avoid vendor lock in simply because they are power crazy. Vendor lock in hits the bottom line, and it hits it unpredictably in many ways. These people don't want freedom for freedoms sake and could care less about how 'evil' proprietary software is (or otherwise). They want options because not having options costs them money. This being said the parent has a point. We shouldn't recomend open source software without cavets. Switching software is expensive. Companies budget for software expenses. Some expect they will have to upgrade MS Office and it's in their software budgets for when it happens. The message shouldn't go "switch to OSS it's free". (beer and pretzels free or freedom free). The message should be switch to OSS when you can afford the major transition costs and take back the power to make your business more profitable. Power to change support source, power to fix bugs, power to get the features you need by putting them in yourself, power to read your achives without converting them through five different formats. People in business want power because the can use that power to save time and make money. Thats why they don't like vendor lock in, and the advantage exists even if switching from Office XP to Office Longhorn (or whatever it will be called) is cheaper than switching from Office XP to OpenOffice.
Rather than thinking about how wrong these idiot people are and spouting off about how super wonderful your Linux experience has been, let's consider why these answers were presented.
As a Linux User, I would have selected a different list of priorities in the survey:
- Security
- Customization
- Cost of Ownership
- Vendor independence
NOTE: Vendor independence goes on the bottom because you are still hooked into some variation of vendor dependency based on RPM/DEB packaging and configuration approaches. Minor at best.What I find really shocking about this is the idea of Security. Apparently an undertanding of Security is rather lacking with the survey group. It's so contradictory to my experiences that I'm not even sure how they could have gotten there. But it needs a little more noise from the Open Source advocates.
I'm very happy with the results of this survey, it shows people are "getting it". All of you guys thinking this is wrong and cost savings is the key advantage of open source really need to think again. To say so is short-sighted, just like saying "free-as-in-beer is higher priority than free-as-in-speech".
Independence from vendors means you can make your own fork of whatever project you are working on and maintain it without someone else making the decisions for you, that's the closest thing to free speech you can get in software development.
Keep in mind that OSS is not necessarily cheaper: A closed source company can choose to squash bugs and integrate new features into an application without asking for more money (other than the licensing, of course), or charge you to fix/develop a specific feature if you happen to be the only customer with such a special request. With OSS, if there's no interest in the community to fix/implement that feature for free, then you have to pay someone to do it as well.
In both cases you end up paying for the custom code, and the only real difference is the cost of the license. But hey! With licenses like the GPL, not only you pay less, you can actually own the pieces you pay for! Maybe even repackage and sell the whole thing! Again, independence is far more important than anything else.
I'm not saying cost savings are not important, but let's face it, OSS doesn't necessarily guarantee free/cheaper support, maintenance or development. Hell, you don't have to charge less money per hour when working on projects related to OSS, you know? =)
What OSS guarantees, however, is a BETTER development model, which usually brings greater cost savings along with it.
- Otaku no naka no otaku, otaking da!!!
Modern Linux distros (and other similar) operating systems are easier to install, configure and maintain that the MS variants I've observed.
Really? Perhaps you can come over and help me figure out how to get my multimonitor setup working correctly. I've been through Xconfig I don't know how many times. Nor was what passes for a manual for SAMBA very helpful, and I spent more than 5,000 times as long configuring networking alone and it still doesn't work as well as my all windows network. Even getting java working wasn't painless, where it wanted to put things didn't work on my distro and I had to spend several hours online (vs 2 minutes on windows) to figure out how to get it to work correctly and install.
Then there was an issue with my drive controller not being properly supported and corrupting data left and right which forced me to go through the entire process again and eventually swap out hardware, and some quirks with printing that took a lengthy discussion in newgroups to figure out & I still don't get as good of results as in Windows.
In the past I've had far less problems, but with my current needs, and current workstation linux just isn't working out. It's not worth the hassle to fight these things for the few advantages it holds.
As a student, and as a professinal, to me the best thing about FOSS is that almost every information you need is there, also for free.
That way, I don't have to spend all my money on books (they're really expensive here at Brasil) and trainning. I can sit down, and read the free online documentation... I've learnd almost everything I know about linux this way, and how to program Java and Python.
If I wanted to learn anything from M$, I would have to buy their OS, their certified books, their certified trainng, and subscribe to their devellopment network... too much money for me!! The average middleclass can't afford all of this around here, I can't.
I own my knowledge to the FOSS... All this free software would be useless to me if the documentation, foruns, newsgroups and chat rooms doesn't exist, or if they cost money! To me, this is the single best feature to the IT professional, it plays a even bigger role here on Brasil, because Linux, and Unix culture, is almost unknow on the academic circle! Microsoft domminated the academic circle far too long, and most of the professors fear and don't understant Linux and FOSS.
The community, that's the "real good thing" about Open Source.
---- You know how some doctors have the Messiah complex - they need to save the world? You've got the "Rubik's" complex
"You know, screw it, time for a change of pace -- lets switch vendors on our database/customer tracking/data mining/image recognition/OCR/whatever solution. I want to spend a couple hundred thousand in transition costs and cause disruptions in our main business to no purpose whatsoever"
Let me tell you what it's like in the real world, using two real world situations I've had to cope with a number of times in my career:
Here's the deal. You've got a ticketing/dispatch system that isn't cutting the mustard, and what's worse, the fine print of the license says that to be in compliance, you need to cut even more functionality, or pay an extra three mill a year. Not even the database schema is available for examination, so you can't jump ship to another vendor, or more reasonably in this day and age, hire a couple of Java geeks and roll your own web app.
Here's the deal II. You've got a mission critical messaging application that can't keep up with demand, pounding the little windows box it's on so hard it keeps falling over. You'd like to put it on one of the big mama-jama Sun Enterprise clusters you've got sitting around with spare capacity. Too bad, the tiny company who licensed it to you had to auction off the sofas in the break room on ebay to meet payroll, and can't really afford to develop a Sun version. Or the megaconglomerate you licensed it from couldn't be bothered to recompile and test on Sun for a single customer.
If it's open source, it's likely someone's already compiled, tested and put it out as a tarball for Solaris10. It's even more likely it's written in a portable language like Java, PHP or Python, using your choice of OSS RDBMS and web server software, making the platform it's deployed upon irrelevant.
Massive changes to infrastructure happen, happen often, and happen for sound business reasons. Closed source applications get in the way of an agile and profit-making IT environment.
SoupIsGood Food
Support. No, seriously. I've seen both proprietary software and open source software crash on me.
With proprietary software, most home users will not be able to do much more than call a paid support phone number and hope their problem goes away in the next version. Those helpdeskers are usually helpdeskers for a reason- if they could develop, they would.
Compare this with the level of support you often get with open source software. To open source developers, their project is often their baby. Not only do the developers not mind you reporting bugs, they actually seem grateful for it. I've seen "help it crashes!" being responded to by "ok let's fire up the debugger", resulting in a solution the same day. Now that's a kind of support I have yet to see in closed-source.
Visit http://ringbreak.dnd.utwente.nl/~mrjb/growingbettersoftware to download your free copy of the book
I think you make some really good points about "cost" here.
So, basically we can say that when determining the relative "costs" of products we should consider the following:
- necessary for conducting business
- actual price of the product
- cost of labor for the user (productivity)
- labor costs of supporting the product (updates/security fixes)
Some things can supercede all others. If your business requires that you deal with a particular type of document, then its a bad business decision to not use the correct tool, regardless of cost.
Basically, people should look at their needs first, then find the product that best fits them, open source, closed source, public domain, whatever. Different people have different needs. Trying to apply a one size fits all solution, whether open or closed source is probably not the best way to go.
----- If communism is a system where the government owns business, what do you call a system where business owns govern
Like how they san't seem to maintain compatibility between minor realeases. How they fix bugs, then reintroduce the same bugs in the next release. Support that reeks when you call them. IBM actually supports Oracle better than Oracle. I know, I've used them both.
I think the only thing Oracle really has going for them is a great sales team. And getting customers locked in because they write all their stuff in plsql.
And what do we get for using this, I think the last report said its 5% faster than PostgreSQL or DB2. Hundreds of thousands of dollars for 5%? WTF?
For the cost of an Oracle 9i/10g license, you can install DB2 or PostgreSQL, buy a box, and hire a new dba.
----- If communism is a system where the government owns business, what do you call a system where business owns govern
- Reduced dependence on software vendors. Somewhat true. I'm still locked in to the product- switching will be an enormous pain, with lots of conversion costs no matter if we're proprietary or Open Source. Switching would actually be easier with a commercial product- conduits exist for Blackboard to WebCT and back. Nothing of the sort exists for Dokeos and Sakai, the project we'd most likely move to. Plus, I've also had to deal with a fork where the lead developer took his ball and went home. That was a little tense.
- Lower total cost of ownership. Almost certainly untrue. Yes, Blackboard would rape us on fees. But you can hire Blackboard training and support people cheap. Dokeos realistically requires a programmer to support. Luckily I like to program, but my job description when I was hired never mentioned that. (I'm rewriting it this week)
- Easier to customize Very dependent on product. The user interface of Dokeos is vastly less configurable than Blackboard. On the flip side, since I can tweak code I have it firmly embedded into half a dozen systems here.
- Higher level of security Very, very doubtful, again with a few exceptions. Back in the days I installed Claroline (Dokeos' parent) it required register_globals=on. There have been other places where the developers have found SQL and code injection points.
I'm a big advocate for Open Source, but those people checking "reduced dependence on software vendors" probably haven't had a serious fork in a project a year after rolling it out to 1000 users."Seven Deadly Sins? I thought it was a to-do list!"
I think you meant 8-10 times less experience using Windows as opposed to Linux. You must be part of Best Buy's Geek Squad. I've heard several people say these clowns take all day to set up a PC with Windows, security software, and basic apps. I figured it was just for the hours they could bill, but now we've got you claiming that it takes 1 1/2 days to set up a Windows machine. The last couple of machines I've set up took me about 3 hours, and that's including setting up all my web authoring and software programming tools.
Now that is what makes Linux better than Windows. I agree with you, so I run Gentoo. However, when I want to toss a distro on a work computer (in a work environment where people still call hard drives CPUs and tell me they have 3 GHz of RAM) that does everything Windows does... And more... I throw Fedora at it and everyone is happy right out of the box.
Windows could come in different flavors with different pre-installed software, rather than the "Home" and "Pro" versions that won't even network to each other properly, but the MS philosophy has always seemed to be "computers should be easy to use" which is my biggest complaint because it leads down that slippery slope of "user friendliness" and the next thing you know paper cips are insisting that the word oxygen can't be pluralized. I mean, it takes me as much time to turn off all the BS (like desktop cleanup wizards, stupid tours, and freaking 'update me' pop-ups) in Windows as it does to configure all my hardware in a fresh Linux install... And don't get me started on administering a mutli-user Windows box... Chmod let me count he ways I love you. Oh /home directory it's ok; I have enough love to go around.
Actually, I wrote my thesis on life experience.
The pool really did not go into details, but I think that some IT users are smart enough to understand that if a vendor drops a product line they are hosed. They are stuck with technology with no way, either through the vendor or by themselves, to support it.
If you look at technology as an industry, it is very volatile. IBM is the only one around for a substantial amount of time (100+ years).
Sperry/Burroghs - gone
GE computers - gone
CP/M - gone.
Apollo computer - gone.
AT&T computers - gone.
Sun is shaky.
HP is shaky.
SGI is shaky and becoming a Wintel box shifter.
DEC absorbed by HP absorbed by HP, the Alpha is being sunset.
Apple almost died.
A host of competitors bought out or killed by MS.
Not to mention the constant upgrade treadmill you can find yourself on, which can be expensive.
The best way to insure that you are not left with an orphaned technology or forced into an expensive upgrade cycle is to go OSS with an open license (GPL, Berkley, Artistic etc.)
putting the 'B' in LGBTQ+
For me one of the biggest pains in the ass is the license mechanics. I'm cool with buying things, and my job allows me to even expense big expensive things, but most thigns have painful license installs. Some require a license server. Most are nice and can integrate with FLex, but some write their own (badly) driving up support time. One vendor was hitting their license server so bad it made it shut down, stopping all licenses. One server needs to be on a lower port, meaning we have to run some crap as root. One client needs to be installed on every machine, and a a key generated by running some software on the localhost, that talks to the vendor's machines and generates a machine specific key. If you're on a machine behind a restrictive firewall, you need to generate the hsot token, send a request on their webserver, and wait for an email with an attachment (and hope your MTA doesn't scrub the attachment or call the message spam). Luckily hardware dongles are a thing of the past, or at least are not in my world anymore.
I've always thought that having a commercial where someone is installing Word on a few machines, having to contact MS license servers, and have them go through all their frustration, compared to jsut installing OpenOffice, no license hassles. Maybe is a good Linspire ad.