FUD About FLOSS v Non-Free Software

However, when we crunch the numbers – not just the total cost of ownership, which is virtually zero, but also the maintenance and support costs, plus the risk of cutting out any particular vendor – only 50 per cent end up actually saving money, while most of the rest break even, and the remainder even spend more compared to proprietary software.

See Open Source vs Proprietary Software

50% saving money and most of the rest breaking even is somehow a bad thng? They do mention that flexibility is achieved with FLOSS. I’ll give them that, but when I crunch the numbers there is no contest. I can save half the capital cost and a huge chunk of maintenance on a system using GNU/Linux. For instance, most schools need a full-time IT guy if they have over 100 PCs with that other OS but it is only a fractional-time job with GNU/Linux. When I was doing that I spent a few minutes checking that every service was running, no huge loads were evident and resetting a few passwords. What’s that 30s per day per PC? I did my regular job and did not even charge for doing a bit of IT because it was not a significant burden. I have managed 40 PCs running that other OS and it was a constant struggle with malware, re-re-reboots and slowing down. Many organizations break even in the first month and the savings pile up. What other investment returns on investment so quickly?

The article goes on to suggest that one cannot run GNU/Linux without having some programmers in-house. What nonsense! Folks here say I am not a real programmer although I did that for a living. I add value by doing a tiny bit of programming, mostly configuring things but also creating web pages and databases, stuff that might be needed with any OS and an IT system actually putting servers to work instead of generating income for M$ ($1k licence + $40 CALs per PC…). Just the work of maintaining a database of those authentication codes is a burden unknown in GNU/Linux. TFA is just silly.

Later TFA pounds the software-patents drum:
“an IT department can best indemnify itself against by trusting the original provider to take on responsibility for dealing with any potential legal issues or patent violations.”

Yeah, right. M$ has been saying that for 8 years now and they haven’t sued anyone who could fight. Google use GNU/Linux on most of their servers and PCs and is now spreading Linux under Android and M$ has not noticed… No one thinks M$ is serious about losing its software-patents by fighting with Google, a company that can fight back. They would rather sue TomTom over FAT. The world is bigger than M$ and its software-patents wielding buddies. M$ will not go around suing its customers or they will soon be ex-customers like Ernie Ball.

No. There are few disadvantages to any business or organization using FLOSS everywhere they can. It makes no sense not to do that. TFA would have been more correct to suggest that FLOSS should be the default and non-free software should only be used in exceptional circumstances. Suggesting that there are problems with FLOSS all the while usage is exploding and users are happy is really strange. That’s like a throwback to 2003 when SCOGvWorld broke out.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology. Bookmark the permalink.

25 Responses to FUD About FLOSS v Non-Free Software

  1. oldman says:

    ““7″ may be the “best” OS M$ has ever produced but it won’t run on half the world’s PCs.”

    Those pc will simply be age out and be replaced with systems that can run the latest software. Not everyone thinks they are going to keep their computers forever like you Pog.

  2. oldman says:

    “Well, if the world should wait ten years to see if GNU/Linux is OK they should also wait ten years to see if “7″ is OK.”

    You are not going to get anyone to take that kind of statement seriously.

    One does not need 7 years to know that the body of commercial desktop applications that run on windows OS ARE more than OK – they are as a body superior ot anything that exists on Linux. IN fact most of the applications that running on linux that people would want to use are sometimes straight clones of commercial applications.

  3. Well, if the world should wait ten years to see if GNU/Linux is OK they should also wait ten years to see if “7” is OK.

  4. The share of XP and the billions of people still not using PCs will swamp lovers of M$ sooner or later. There is no force pushing people to use that other OS if they find Linux on retail shelves and M$ is forced to interoperate.

    “7” may be the “best” OS M$ has ever produced but it won’t run on half the world’s PCs.

  5. oldman says:

    “ou would delay innovation for a decade? That’s too “conservative” for my liking.”

    Again a subjective statementr What innovation has been delayed Pog. I went from having software that could baqrely notate my music, to having software that non only notates is but allows me to render a performance with human nuance in what sounds like a hall and burn it to disk as mp3 or wav.

    And that just one innovation.

  6. oldman says:

    “I would rather not be always working around M$ when I can get along fine without them.”

    And that is your choice, but being your personal choice does not make it valid for everyone, especially in the light of the fact that the bulk of your experience has been with a version of windows that is now two generations obsolete.

    It strikes me You are going to have a harder and harder time getting taken seriously by anyone outside those who are of like mind with yourself if you insist on trying to tar microsoft today with the shortcomings of now obsolete software.

  7. oldman says:

    “M$ has been good at getting its partners to provide lock-in. By using M$’s tools they produced unportable software.”

    Microsft hasn’t been that good Pog. Most of the software that I consider as part of my core applications are available for instance cross platform on both Windows and OS X. Both of which are commercial operating systems with completely different API’s. It would seem that tha ISV’s involved, (microsoft included BTW) feel that there is bu$ine$$ to be had doing so.

    “The world would have been much better off writing portable software.”

    That is a highly subjective statement Robert. The issue of portability was a topic of conversation when I was actively programming. The issues were already known – make it portable and function suffered as one had to dumb ones application down to the lease common denominator functionality. It was of course moot under the DOS’s (microsoft and apple) because of the relative primitiveness of the environment, one could create fairly portable software. The company that actually tipped the balance away from application portability was apple with the Mac OS, who came along in 1984 and demonstrated what a non portable environment with rich OS features could deliver in application functionality. Microsoft was actually johnny come lately with windows and took 7 years to come up with a version (windows 95) of their environment that even came close to apple’s offering. People voted with their wallets and the rest was history.

    The world is doing that now and cutting M$ out. They were not needed. “

  8. You would delay innovation for a decade? That’s too “conservative” for my liking. If you had followed that principle, M$ never would have gotten in the door.

  9. oldman says:

    “I see Largo as being in the real world and that other OS is not cheaper.”

    What you view is irrelevant. Largo remains an island that is kept going by the efforts of a very dedicated team. The proof of the pudding will be what happens as that team breaks up and/or retires.

  10. Remember XP SP2. M$ slipped that in to our organization and our scanner became dead wood. Folks had worked around M$ and when M$ changed things, our system broke. What kind of innovation was that?

    M$ has been good at getting its partners to provide lock-in. By using M$’s tools they produced unportable software. Suppose M$ committed suicide. All that software would have to be rewritten. The world would have been much better off writing portable software. The world is doing that now and cutting M$ out. They were not needed. The OS does not need to be intertwined with all the applications. That is a poor design.

  11. We had a steady stream of updates for security (critical or not) and other tweaks that changed the behaviour of our systems in ways we did not like. Every now and then M$ would undo some of my configurations. I would rather not be always working around M$ when I can get along fine without them.

  12. oldman says:

    “M$ has created a world of sick IT. IT should be for the benefit of the users, not the suppliers.”

    I have never seen such an astoundingly wrongheaded statement. Its behavior aside, Microsoft the company has over the past 30 years been a major contributor to the desktop computing revolution. Their OS and tools have allowed an army of vendors to bring a rich set of application to the computing public. a set of application, I might add that have been extensive copied by the FOSS community.

    As far as the infamous mips wasting memos are concerned, So microsoft wrote applications that wasted mips – what about it. My experience and the experience of many over the years was that in almost all cases, the increase in function was always greater than any increase in hardware power needed to support.

    I’m sure that there were those for whom these advances were unnecessary – their existing software being good enough. Those users were free to stay where they were, and did so. I know many people running 8 year old windows XP systems and being happy about it. The rest of us simply upgraded our hardware as needed over the years and chalked it up as the cost of progress.

  13. oldman says:

    “Sitting down to study XP is a waste of time. On the next update, M$ can change the behaviour. What I like to do is pave it over with GNU/Linux. Problem gone.”

    Pog, I think you hav been spoiled by the unique situation that has allowed you do this. In the vast majority of other places, you don’t get to just pave over your problems with the solution of choice. as a support person you get to solve the problem, and if you you want to make your life easier, you DO study the problem and acquire the tools that you need.

    Not studying XP because microsoft is going to change it is a copout, especially in your environment where XP is probably going to be around for quite a long time. Assuming your job hiatus is not a permanent one, I would suggest that you hold your nose and start learning how to make any time that you spend with windows products as easy on yourself as you can.
    It will also mak you more employable if you go in knowing both windows and linux.

    Powershell is a good start BTW. Even version 2.0 can be retrofitted on to Windows XP and would allow you to perform maky if not all of the scripting tricks you perform now with Linux.

  14. Well, then it was really wasting time… see http://www.wikihow.com/Make-Windows-XP-Startup-Faster

    Sitting down to study XP is a waste of time. On the next update, M$ can change the behaviour. What I like to do is pave it over with GNU/Linux. Problem gone.

  15. Yonah says:

    “I have even seen desktops that were unusable for 30s or longer when they appeared and any attempt to use them was met with an hour-glass.”

    Yes, so have I, and it is a problem that is fairly easy to rectify. Did you actually sit down and investigate WHY this occurs? You are aware that “pre-loading” was not a feature of XP, right?

  16. Amen. I have often watched that other OS boot. It is amazing how long the CPU usage and disk activity stays high. I have even seen desktops that were unusable for 30s or longer when they appeared and any attempt to use them was met with an hour-glass. Some of that is pre-loading apps that may not even be used. For example, a person may use the browser and not the word-processor or the word-processor and not the browser or both and M$ certainly does not know what it will be yet denies use of the PC for a long time. A GNU/Linux thick client can also be configured to preload stuff but in Debian GNU/Linux anyway, it is not a default. With a GNU/Linux terminal server, all the commonly used apps are already in RAM for every session so preloading is totally unnecessary, hence login takes a few seconds.

    We also have M$’s internal correspondence that they were seeking “MIPS wasting” applications to squander the resources for hardware makers’ benefits.

    M$ has created a world of sick IT. IT should be for the benefit of the users, not the suppliers.

  17. Adam King says:

    GNU/Linux uses much less of the CPU than winders so you don’t have to upgrade your hardware as much. M$ has partnered with the hardware companies to make sure that you’re coming back for more every few years.

  18. That could be reasonable if there were 2.5X users/services/configurations to do on those servers. I have never seen a server running that other OS doing much of anything. One place I worked had 7 servers running 2000/2003 where a single GNU/Linux server would have been fine. My predecessor had added 1 server for each function: 2 tiny databases, WSUS, AD, file sharing, printing, and backup. So, the work of maintaining network services was huge compared to apt-get update on one server with GNU/Linux. Further, the service on that LAN sucked even while the servers were idling. We could see 30s pauses randomly at any time of the day or night.

  19. Alex says:

    10 Linux System Administrators @ ca. $120K+ per year.

    Oldman,
    Why would you need 2.5x the number of linux admins? Something smells fishy about this. Care to explain?

  20. That sound like clutching for straws to support an argument. I am sure Largo being an open government knows what it is doing and can best judge what is best for them. I am sure Dave Richards and others have documented everything well. I turned a big system over to a guy who had only used that other OS and he had few problems in the last 5 years keeping it running. In many ways GNU/Linux is simpler to operate than that other OS. Everything is modular for one thing. If you figure out how to get feature X you usually keep all the other features you are running.

  21. I see Largo as being in the real world and that other OS is not cheaper. Dave Richards’ book breaks down the costs including admin and support and they save a huge bundle which they invest wisely in other things. The fact that he and others in Largo can optimize the system and spend time on it is due to the lower costs of operation which he documents. Basically, if you spend less on maintenance they can spend more on hardware and system design and get better performance/$. It’s not magic.

    My schools are an example of taking mostly existing hardware, installing GNU/Linux and getting a huge increase in productivity for $0 CAPEX. The time I spend on IT is so little with GNU/Linux that it is not a budget item.

    The thing about Largo is that it was not a top-down decision to go with GNU/Linux. They were using UNIX and it was too expensive. It was an obvious savings to go to GNU/Linux and they are laughing at the stuff M$’s victims have to go through to keep systems running.

  22. oldman says:

    Let me try that again…

    The more common type of Linux/FOSS proponent is closer the bozo running the Munich effort who referred to their users as “happy slaves”. For this reason alone Largo Florida is at best, a corner case whose longevity beyond the team that makes it work is in my eyes suspect.

  23. oldman says:

    Pog:

    The $440. for the windows license is a one time charge. If we assume that the OS changes with the hardware ( we are on 3 year leases) that brings the raw costs per windows server to

    147 + 250 + 2400 = $2797.00 per year, versus $7300.00 per year. No matter how you cook it Pog, windows IS cheaper.

    “In my world the system admin is $0 so the numbers would be $1300 v $690. and most schools would typically have one server just authenticating, in other words, idling. With Debian the numbers would be near $0 v $690.00”

    But the article is not talking about your world Pog, it is talking about a world in which people accept a certain cost for IT. Your world as you have described it does not seem to consider desktop IT as anything other than an afterthought, to be funded willy nilly. Your world is perfectly willing to stick those who need IT with refurbished castoffs from business that cant use them anymore and who are looking for a tax writeoff. Your world is irrelevant Pog to the subject of this article.

    “GNU/Linux works for me. With so many system admins, I doubt you need those support contracts or the per year per server charges.”

    Coming from someone who has acknowledged that Linux has its flaws that can be worked around, forgive me if I view you judgement in this matter as somewhat compromised. At any rate, how Linux works for you is again, irrelevant. You and your practices are not what is being discussed here, it is the general practice of the business world, and the purchase of service contracts as part of system support, even when a staff exists to support, is part of a total cost of insuring business continuity.

    “No mention of CALs.”

    Most of our servers are not connected to directly, so CAL’s are not an issue. but since you bring it up , we have about $55,000 a year for CAL’s (11000 seats x $5 @seat.) For simplicity’s sake I will divide this number by 200 we get $257 per server which brings us back up to $3072 per year which is still quite a bit less than the Linux costs.

    “I have read many reports that people do 3X as much with the same hardware running GNU/Linux as that other OS. ”

    And I can show you workloads that drive Linux into the ground. So What. Our servers are running specific application that we are mandated to support, and those costs are accepted as the cost of doing business. But that does not change the fact that Linux based applications cost mush more to run than windows based applications.

    ““Largo …..Windows.”

    I’m not surprised that you trot out Largo florida Pog. Its one of the few situations where a pure linux environment has been successful. However, it doesn’t take much research to realize that Largo florida is where it is because of a dedicated team who expend an enormous ongoing effort making linux and FOSS work for its users. They understand that if the users are not kept happy, then eventually their solution will be called into doubt. The more common type of Linux/FOSS proponent is closer the bozo running the Music effort who referred to their users as “happy slaves”. For this reason alone Largo Florida is at best, a corner case whose longevity beyond the team that makes it work is in my eyes suspect.

  24. @oldman:
    your numbers come to $1300 + $6000 = $7300 per year per Linux server and $440 + $250 +$2400 = $3090 per year per server with that other OS.

    No mention of CALs. No mention of the workload/throughput. I have read many reports that people do 3X as much with the same hardware running GNU/Linux as that other OS. That would be enough to show an advantage for GNU/Linux. Does the fact that you claim more admins per Linux server reflect the workload or the ease of use?

    In my world the system admin is $0 so the numbers would be $1300 v $690. and most schools would typically have one server just authenticating, in other words, idling. With Debian the numbers would be near $0 v $690.

    GNU/Linux works for me. With so many system admins, I doubt you need those support contracts or the per year per server charges.
    “Largo’s IT budget is less than half of what similarly-sized municipalities typically spend for similar (and often less reliable) computing power.”

    see Largo still loves Linux
    By Robin ‘Roblimo’ Miller on September 21, 2007

    Largo uses heavy duty thin clients and expensive terminal servers ($40K or so each) with as many as 400 desktop sessions on each and they have servers dedicated to particular applications to optimize everything. Their system admin has the time to write a blog…
    “The amount of time that Richards and fellow Largo sysadmin Mike Pearlman and CIO Harold A. Schomaker spend on things like 3-D desktops, which some IT departments might cavalierly dismiss as “eye candy,” is the secret to their success at getting users to not only accept, but embrace and even love a Linux-based client-server system, instead of whining about how they’d really rather have Windows.”

  25. oldman says:

    Pog:

    Lets go over some facts from my environment.

    200 Linux servers running RHEL @ $1300.00 per year
    10 Linux System Administrators @ ca. $120K+ per year.

    200 Windows Servers @ $440.0 per server + $50K per year support contract.
    4 Windows System Administrators @ $120K per year.

    I would say that the article is Spot on.

Leave a Reply