There may not be fireworks. CIOs and IT directors may not be heaping their proprietary software on bonfires and dancing. Even so, the open source revolution is happening right now and will carry the day, said Bill Weinberg, open source architecture specialist and Linux evangelist for OSDL.
IMO anyone who posts a story on a news site, should at least have searched the archives a bit before posting. If only to check if it’s interesting in context to recently covered items.
3 articles down on the main page… Any editor/poster should have spotted that. We’re not going the Slashdot way here, are we?
Erm, these two articles aren’t the same.
@Thom: you’re right. Same site, date posted, interviewed person, even editor, and similar subject, I jumped to conclusion too quickly. My bad, I stand corrected. Face red with shame over here ;-}
What are the three top things that CIOs should know about open source software and the open source movement today?
Weinberg: One — Linux and OSS are enterprise-ready: Leading systems and platform providers, thousands of independent software vendors and a global community of developers and integrators place Linux and OSS on a par and ahead of legacy proprietary platforms and applications.
This is complete mythology. Linux is not “enterprise ready” unless you are only talking “deep server room & lots of admins”. Outside of “deep server” there is nothing in the enterprise that can bear the high cost and low configuration flexibility of Linux.
In most situations the operational complexity of Linux far outweighs any benefits.
Two — The combination of open source and COTS [commercial-off-the-shelf] hardware offers lower cost-of-ownership and return-on-investment that proprietary s/w and hardware don’t offer.
This may be true if only because the alternative is the Microsoft monopoly (or Oracle’s database mafia) with its outrageous pricing.
However the infrastructure exists today to support closed source apps while it largely does not exist to support OSS apps. So cost savings will vary greatly and may not be available for years while staff ramps up on new OSS technologies/apps/platforms/etc.
It should not be construed as any real victory for Linux to be the smaller (but still very large) evil of two great evils.
Linux’s overall cost structure is also insane. Just less insane than the entrenched monopolies like Microsoft.
Three — Linux and open source offer more than lower cost; they free companies from dependence on single suppliers and expand the ecosystem of solutions.
This is somewhat true as long as integrations between multi-vendor systems occur in ways that do not depend on the nuances of the many incompatible Linux variants.
If integrations are closer to the actual OS, there is likely no savings involved other than basic OS licensing. This still may be significant, but it is not something that will make or break a large enterprise.
On the balance, FOSS is a small reduction in cost coupled with a large increase in complexity. Whether or not this is a true payoff for an enterprise is not an easy question.
Hence OSS is more evolution than revolution. The players who fund OSS are big money players. They do not believe in revolutions.
http://distrocenter.linux.com/article.pl?sid=04/09/17/2019255
Not entriprise ready you say? Requires lots of admins? Remember Linux isn’t one product, it’s an ecosystem and philosophy of software.
Not entriprise ready you say? Requires lots of admins? Remember Linux isn’t one product, it’s an ecosystem and philosophy of software.
To their credit, Nitix looks like it has made some progress tackling the complexity of Linux.
However, when you go to install the software you want on the Nitix OS, you will like all other Linux distros it is not compatible with any other distro.
So you are stuck if you want to run anything that does not come with Nitix.
This is the only way complexity is currently handled in the Linux world today — building lots of islands with simplified rule systems. The rule systems do make life on that island easier to bear. But the rules are not the same from island to island.
So looking at the Nitix catalog, there are 8 apps which are made made for/ready for Nitix. That’s it. 8. A few more apps were “tested” but some of these apps are not even apps or don’t even run on Linux. Many of the “compatible” apps are old versions with security holes and other issues.
http://www.nitix.com/isv/catalogue/browse.php?app_name=&sv_company=…
This brings me back to the core issue I brought up. Until Linux is redesigned, all you will have is hundreds of incompatible islands each with a small core of apps that run on that island.
And you know, compared to Windows or even Mac, that is not compelling. For a small business or enterprise to rely on a bunch of small islands for their IT infrastructure is just plain stupid.
It will take a redesign of Linux to allow the specialization while maintaining compatibility. It can be done but it takes invention and hard work. Today I don’t see anything interesting happening on Linux because most Linux efforts are busy copying Windows or Mac rather than building a new innovative platform for the future that provides new benefits while reducing both complexity and cost, not just one of the two.
Does it cause you pain when someone says nice things about Linux? Are you on a crusade? Do you go around in life haranguing your acquaintances about their shortcomings?
We get it. You don’t like Linux. What do you want everyone else to do, say ‘you’re right pravda, we’ll never use or consider using Linux again’?
Not going to happen. Obviously many people disagree with you about what makes a good OS/desktop/gui/mouse. Yet you just keep on ad nauseam. Why don’t you accept that different people like different things?
And don’t pretend like it’s constructive critisism. Saying it needs to be rebuilt from the ground up is not constructive.
I am strong enough to talk about reality. Most people are not. They will stick in a bad relationship for years. And for many situations, Linux is that bad relationship. Maybe it will get better. But usually relationships get better when the people get better themselves. Ultimately this is all I’m saying.
If you are taking it personally, then you are probably not yet strong enough to talk about reality.
I am strong enough to talk about reality. Most people are not. They will stick in a bad relationship for years. And for many situations, Linux is that bad relationship. Maybe it will get better. But usually relationships get better when the people get better themselves. Ultimately this is all I’m saying.
If you are taking it personally, then you are probably not yet strong enough to talk about reality.
Like your namesake Soviet newspaper, you spout emotions and ra-ra opinions claimed as facts, yet do not offer actual facts. It’s propoganda, heat, and no light.
You put your finger on it. Only paid shills have the energy and motivation to church out FUD day and day out. In fact, as they do it so often, some of them eventually finesse their raving lunacy into what looks like a semblance of an argument or enough of an argument that it may fool some bystander or put enough fear in him not to try linux and that is all that people of this guy’s ilk are afterwards.
What a sad way to live, but there you go. To anyone who hasn’t tried Linux, my answer to you is to ask yourself what it is that you want to accomplish. Then, establish objective milestones and criteria and let yourself be guided by that. Sometimes, ClarkConnect with its server-via-a-web-browser config may be the answer, sometimes Novell Suse Server might be, but I assure that those of us that are running Linux do so because it is cheaper and more reliable than anything else we have tried.
A lot of people have tried Linux and found that it absolutely sucks for usability.
That is why there are seemingly hundreds of articles every week regarding Linux usability.
It is well established that “poor usability” = “higher costs”. It is also common sense.
Beyond poor usability, every flavor of Linux is mostly incompatible with every other flavor of Linux. Install one new package and your entire Linux system can go down in flames due to Linux’s fucked up dependency model. Now this is only true for 99% of the Linux variants out there. Some few distros do not follow the normal Linux path of maximum retardation.
So rather than attack people (cheap shill tactic), why not give some constructive information on how Linux really costs less to an enterprise. You will find that there are a few examples in the server room. But outside of that domain, “Linux costs less” is folklore mythology.
“You will find that there are a few examples in the server room. But outside of that domain, “Linux costs less” is folklore mythology.”
Right you are conceding the server room – and I guess you have to as the latest IDC survey shows Linux rapidly increasing its share of server sales (for systems sold with an already installed OS) compared to MS (who I guess you work for directly or for one of its business partners) Windows both in money and unit sales terms.
This means in the usability and cost comparison you must be talking about the desktop. For the home user there is no advantage except freedom and a more interesting and powerful technical environment. You have to pay the Microsoft tax when you purchase a home computer from any large manufacturer.
For the corporate user the situation is somewhat different. You can make bulk purchases of OSless systems and save the MS tax. However often it is held back by the need to run legacy software, lack of computer skills in MS shops etc. However where there is ether a limited set of specific apps required and either a web or Java based set of specific local corporate tools (as is typical in any corporation using a real database rather than MS SQL) such as a call center or alternately where the end user is an engineer or developer whose program needs are satisfied by those tools that run on Linux then the deployment of Linux desktops becomes more attractive form an ROI point of view, especially if you have in-house Linux/Unix admin skills available.
Finally Linux is superlative in a thin client situation where large savings can be made in hardware costs as well as savings on licenses and reduced administration overheads (yeah! not that I recommend firing anyone – but you could save a lot by letting go those surplus MCSE’s). Recent advances in thin client protocols had made this option even more desirable.
I actually work for a small ISV. Our products run on Windows today. We’d like to move them to Linux, but the software tools are either too expensive (Qt) or too crappy (everything else). The maturity of the development tools is also not good on Linux — a lot of bugs and a lot of missing polish.
But the killer for Linux is having to test on endless configurations even for a single vendor’s Linux distro. Even if we could build the software, it is cost prohibitive to test it and support it.
Linux for good or bad is architected mainly for GPL programs that companies themselves have to get working based on having the source code and making it work. Maybe this is better for the big picture, but there is a giant gap between where enterprise/corporate IT is today and this GPL world.
Also outside of the skill required in corporations to support GPL software, you also have the issue that many government agencies and large corporations have COTS requirements — Commercial Off The Shelf. This GPL software will likely not be certified as COTS for some time, if ever.
Thus the Linux cost picture is hazy. I believe the entry costs for Linux are lower, but the complexity is certainly higher. Usability for Linux end-user machines is also not good today.
How the Linux software market evolves and unfolds will be interesting. So far, there are few Linux ISVs. Maybe there never will be many other than big giant companies offering low(er) quality Linux software and lots and lots of services to install/config/update/support the software. Time will tell.
“Also outside of the skill required in corporations to support GPL software, you also have the issue that many government agencies and large corporations have COTS requirements — Commercial Off The Shelf. This GPL software will likely not be certified as COTS for some time, if ever.”
Large government agencies and corporations in the US and world wide buy RH and Novell enterprise systems these systems are virtually all GPL and other OSS licensed and of course they contain a huge range of FOSS application software. So clearly they regard buying from RH or Novell as being COTS or else COTS requirements are not widespread in this area.
You are right that there are problems for a small ISV moving into cross platform applications that run on Linux. However there are some for example Softmaker in Germany that uses QT for its Office suite. There are of course companies producing cross platform apps using Java. The native platform widgets of SWT enables the production of very nice totally native looking apps, good examples of these (admittedly CPL licensed) are Azureus and RSSOwl. I believe that using the ecj compiler in Eclipse enables the production of application binaries that do not require a JVM on the target platform.
Mitch Kapoor and the OSAF are developing their cross platform advanced PIM software Chandler primarily using Python. I recently installed there pre-production 0.5 release using a zipped distro independent binary. It just worked. Adobe has produced a very nice port of Acrobat Reader 7.0 to Linux GTK2 and it installed nicely on my Mandriva box without requiring a distro specific rpm.
Finally for VB shops RealBasic is a nice equivalent. Like QT you have to pay for it but its a lot cheaper.
> …the high cost and low configuration flexibility of Linux.
I thought you said OSS is lower offers cost.
The “low configuration flexibility” sounds wierd too. Comparing to Microsoft products, Linux has a great configuration flexibility. comparing to Unix, it has the standard configuration flexibility.
That covers most servers today. What is your system with high configuration flexibility?
I thought you said OSS is lower offers cost.
The “low configuration flexibility” sounds wierd too. Comparing to Microsoft products, Linux has a great configuration flexibility. comparing to Unix, it has the standard configuration flexibility.
That covers most servers today. What is your system with high configuration flexibility?
My English broke down a bit there.
The aim of introducing “flexibility” into the equation is to bring focus to the fact that the greater complexity of Linux actually leads to lower flexibility. Linux is more of a “configure it once and don’t muck with it” operating system. Upgrading just one component on the OS can break the entire OS. So you don’t want to change it alot.
Now your point of Linux being configurable is exactly right. Linux comes in many flavors as starting points and offers source code. It is a toolkit in many ways.
So you have the ability to configure Linux — configurability — and this is strong. However once it is configured, the cost of changing the configuration — the configuration flexibility — is high.
I hope this makes more sense.
>> Linux is more of a “configure it once and don’t muck with it” operating system. Upgrading just one component on the OS can break the entire OS. So you don’t want to change it alot.
>> In most situations the operational complexity of Linux far outweighs any benefits.
comments like this show that you really don’t know what you are talking about.
linux/unix is incredibly simple – providing you have been *trained*
if you are untrained you can flounder around in the windows GUI – but anyone can do that. untrained people are called *users* and shouldn’t be posting on techie forums like this one.
again – to *trained* techies linux/unix is easily more simple and configurable than windows which is overly complex and bloated.
IKEA Russia have been using it since about 2000. They have Linux installed on every PC in the store. What they did was customizing the system to be as easy and foolproof as possible. On some PCs they have touchscreens and just one app launched, the database client.
If they used Windows, they’d have to pay more to get the same functionality.