Opening Up for Open Source

jondaw writes "Businesses want to save money and boost IT efficiency. Can open-source software do the trick? Cnet attempts to answer this open ended question and provides a number of good case studies and examples."

Comments

Open source

[Anonymous Coward]

Only if open source is actually better can it save them money and efficiency, they'll still have to pay for it because they want support. If they have open source for cheaper and working more efficiently than closed source software, good for them, but if they pay cheaper, but need more support and things work less fluidly then overall, the efficiency is down.

Re:Open source

[Bloke down the pub]

Very insightful. Are you an MBA or an MBO (Master of the Blooody Obvious)?

Re:Open source

[Anonymous Coward]

Based on what I see daily on slashdot, I would say the Bloody Obvious is probably too often forgotten. Far too many, especially in the OS movement take for granted that the OS approach is always better and cheaper, sometimes forcing another OS tool that does half the job in twice the time as a one-time cost of $50 would solve.

I myself make a living of using open source, free tools because my manager thinks it's cheaper. In reality, I do the maintainance, and did we run other systems, I would not have as much work as I do. I have a on-demand contract so more work means less noodles and more beef along with my studies.

Of course it can't.

[Anonymous Coward]

Geez. Do they even need to ask? Noobs.

Re:Of course it can't.

[Billly Gates]

Sadlt yes

Most in IT today agree that Linux is great on a server in some circumstances but the Microsoft Salesmen come into the picture to our bosses with glossy brochures about TCO studies of costs being lower in Windows.

They also count in retraining costs and the fact that an MCSE is cheaper than a unix admin.

Many in IT are convinced that Windows is cheaper as well since its an integrated platform with VS and all the windows desktops.

Its a tough sell these days and now the MS salesmen are trained to scare CIO's about liability and lawsuits and lvoe to cite SCO.

Re:Of course it can't.

[Anonymous Coward]

Even so, if there is a $15,000 to $30,000 disparity between the cost of a paper MCSE and a *nix admin, that cost is easily recouped by not having to pay for CALs (Windows PLUS Terminal Services PLUS Exchange), SQL Server licenses, plus premiere support contracts for a small handful of servers for just one year.

MySQL is free, unless you need to run tasks that shell scripts and crond cannot manage, or unless you want to bundle MySQL with a closed-source product, or unless you want to pay for a support contract and full-color retail box, and even then you don't pay through the nose like you do for SQL Server, or, even worse, Oracle (Oracle's pricing is obscene. Even if you want to just run a development workstation you have to pay per mhz, per cpu, and factor that by the amount of RAM - not on a per-user or per-connection basis, which is more reasonable. I refuse to work with Oracle unless a client absolutely demands it).

Apache? Free.

Linux? Free, unless you want to pay for support from Novell, Redhat, etc. - and even if you want to, unless you're developing a product that speaks to the system at low-level, you will never need the support because usually Yahoo! or Google will return the solution for your problem within an average of .015 seconds of clicking submit - and if you didn't find the answer on the first two or three pages, you asked the wrong question ;).

*BSD? Free. Support? Like Linux, support is a Yahoo! or Google query away.

Sendmail? Free. No it doesn't do group scheduling, but true, free Exchange replacements will arrive soon.

Postfix? Free. No it doesn't do group scheduling, but true, free Exchange replacements will arrive soon.

Backup packages? bash, crond, and tar or dar are all free, and what's more, even a live backup of a database or email file/filesystem can be mounted in a pinch in the event that you didn't grab an offline backup or a proper "dump" - and it's less than two minutes' work. Try that with SQL Server or Oracle. Disaster recovery is MUCH easier on *nix than on Windows.

Sure, Windows gives you a slick GUI but automating routine administrative tasks is a royal pain in the ass. You need to hop on one foot and chant to various gods and hope and pray that your vbscript will work (oh, and don't forget to leave the console unlocked so vbscript will be able to access the GUI), and while praying, pray that the applications you need to interact with make all of their controls accessible by VB so you don't have to resort to display-and-font dependent (X,Y) coordinates.

For years M$ has insisted that GUI is the one true way to administer a system but at long last they have finally admitted they're wrong and are introducing bash^H^H^H^HMonad with the server operating system. It'll be 15 years later than everyone said they needed it, but hey, cut them a break. Sometimes it only takes them 8 years to admit they're wrong and bundle a disk defrag program. Eventually Windows will be just as easy to administer as *nix is - remotely, even! Problem though: in addition to having to have antivirus programs, antispyware programs, and kludgy software firewalls installed (in addition to "internet security" software to block vbs, activex, and js exploits), Windows will now be in danger of being rooted and will require rkhunter and chkrootkit to be installed as well. It'll be the most vulnerable OS on the face of the planet at that point.

What sucks about Monad? Microsoft can take open source work and thanks to the BSD license charge money for it all while continuing to spread anti-open source FUD. That's my one problem with the BSD license: Microsoft has on many occasions "borrowed" BSD code for key components of Win9x and NT, but at the same time stated that open source is the root of all evil, that it's unstable code (gee, no wonder their IP stack wasn't all that back then!), and insecure (sure, blame it on *BSD sockets).

But yeah, one day Windows will be as cheap and easy to roll out in a corporate environment as Linux or BSD are. :D

Yes, but...

[Anonymous Coward]

Only if it gets the issue of security right. The thing is, the whole claim that OSS has inherently better security has been exposed as hype for a long time now.

Some OSS projects have excellent security, because the project leaders place sufficient emphasis on it, and the coders code with that emphasis in mind.

Other OSS projects do not have good security, sometimes not even as good as Microsoft and co.

Consider this: I have downloaded patches for more security flaws in Firefox than for IE in recent weeks. Moreover, the IE patches were offered to me via automatic updates within minutes of being available on Windows Update, while the Firefox patches did not show up as automatic updates for several days after they were available from the project web site in some cases. They even had a whole version missed out of the automatic updates, because somehow a release was made that contained serious bugs of its own, and had to be withdrawn.

This is not intended to be a slam against Firefox; it's great software and the project seems to be run well, the vast majority of the time. Rather, this is intended to demonstrate that nothing's perfect.

Re:Yes, but...

[MightyMartian]

You are aware, I trust, that Microsoft frequently sits on vulnerabilities for some time before offering patches. Your metric for security appears to have nothing at all to do with security.

Re:Yes, but...

[rbarreira]

Consider this: I have downloaded patches for more security flaws in Firefox than for IE in recent weeks.

You say that as if you wanted to imply that Firefox has more security holes, but that's not a certain conclusion! Couldn't it be the case that Firefox just gets more attention from its developers?

Signed,
Captain Obvious

Re:Yes, but...

[Anonymous Coward]

Sits on them? You will of course have evidence for that. Certainly they wait until they regression test through the supported versions of the OS, as opposed to rush out a half assed fix which no-one has really tested, but sits on them aside from that? Prove it.

Re:Yes, but...

[Michalson]

You are aware, I trust, that the Mozilla foundation frequently sits on vulnerabilities for some time before offering patches.

As an example, rather then just making an unsubstantiated allegation, the most recent patch, 1.0.5, fixed a critical vulnerability ("Code execution through shared function objects") that Mozilla had been sitting on for 2 months, and a high vulnerability ("Content-generated event vulnerabilities") that Mozilla had been sitting on for 3 months.

There where also additional vulnerabilities ranging from High to Low patched in that update that had been known to Mozilla for 2 or more months.

And this is only recent. Before FireFox 1.1, Mozilla was far less forth coming about vulnerabities, often patching them at their leisure and then silently introducing them into builds without any advisory to let people protect themselves; go look at the disclosure list - you'll find pages of dangerous vulnerabilities you where never told existed and for which you remained unprotected against unless you where downloading builds on a nightly basis (and reading the list wouldn't help you - Mozilla used to intentionally keep it 2 major versions behind).

Mozilla built its reputation for security (a reputation that is dimishing as each new FireFox vulnerability is announced) by hiding its flaws and promoting fanboys (like the parent). Now that it has broken into the mainstream, it has to play like everyone else, without the special treatment and fanboy reality distortion fields to protect it.

Re:Yes, but...

[Master of Transhuman]

"Moreover, the IE patches were offered to me via automatic updates within minutes of being available on Windows Update"

Uhm, that's WHY they call it "Windows Update".

Moron. Microsoft takes longer to patch, their patches break more things, and the vulnerabilities they patch are more serious than OSS ones in most cases. Just because Firefox, and indeed, other OSS products such as Apache or Sendmail, have had a number of security issues doesn't justify tarring the entire OSS field for bad security in comparison to Microsoft.

And comparing all of OSS to Windows in comparing security is just braindead. A more appropriate comparison would be either Linux/BSD vrs any version of Windows OS, or ALL Windows apps against ALL OS apps.

As quality of OSS code has been demonstrated to be better than commercial code in several studies, it is likely that security would be at least equal, if not better. As security-concious coding practices are relatively new, both OSS and commercial code obviously need more work.

And finally, nobody ever said OSS software is perfect.

They said it was as good and cheaper than commercial software in many cases. And it is.

Re:Yes, but...

[Master of Transhuman]

Check http://eeye.com/html/research/upcoming/index.html site for a short list of overdue fixes.

And that's just the vulnerabilities THEY reported.

Re:Yes, but...

[Master of Transhuman]

And here's a quote from another article on the subject back in 2004:

http://www.techweb.com/wire/26803909

"The vulnerability in question is one of two noted as "critical" by Microsoft on Tuesday, when it released February's monthly fixes. Hackers could exploit flaws in Windows's usage of Abstract Syntax Notation (ASN), a language for defining the syntax of data messages shared between applications and computers. If attackers successfully created exploits, they could clandestinely destroy data, steal information, or compromise network security.

The bug has been characterized as one of the most serious ever due to its widespread use in many of the Windows operating system's security subsystems, including Kerberos and NTLM authentication, and in numerous server and desktop programs, such as Exchange and Internet Explorer.

The ANS vulnerability was first identified on July 25, 2003, by eEye Digital Security, but not fixed until more than seven months later.

And there's the rub.

'Microsoft had 200 days to fix this,' Mark Maiffret said in a teleconference. Maiffret is the chief hacking officer and a co-founder of eEye Digital Security, and the discoverer of the ANS vulnerability. "That's a ridiculous amount of time.'

To his credit, Maiffret kept quiet about the vulnerability while Microsoft worked on and tested a patch. Currently, there are no exploits circulating or pending.

In its defense, Microsoft said that the company needed the time to assemble,but more importantly, test, the fix. "This investigation required us to evaluate several aspects and instances of this pervasive functionality in order for our engineers to create a comprehensive and high quality fix," a company spokesperson said. 'This was an instance in which due diligence required us to very carefully evaluate the broadest possible implications of the anomaly.'

But that's no excuse, another analyst said Thursday.

'I recognize that Microsoft has thrown an incredible amount of money and resources at security issues,' said Laura DiDio, a senior analyst for the Yankee Group who has been tracking security for over 17 years. 'The company is under siege, no question. They're the number one target, like a policeman in Baghdad.

'Where I fault them -- even if you give them the benefit of the doubt -- is that you can't take seven months to patch a problem of this magnitude.'"

Did you catch that last? LAURA DIDIO, Microsoft shill par excellence, can't even justify that!

Re:Yes, but...

[superpulpsicle]

You are more than welcome to go back to IE and download spyware, adware, activeX scripts of all kinds. Guess we won't be seeing you on slashdot for awhile....

Re:Yes, but...

[Anonymous Coward]

>You are more than welcome to go back to IE and download spyware, adware, activeX scripts of all kinds. Guess we won't be seeing you on slashdot for awhile....

This reminds me of a joke:

A Harvard man and a Yale man are at the urinal. They finish and zip up. The Harvard man proceeds to the sink to wash his hands, while the Yale man immediately makes for the exit.
The Harvard man says, "At Hah-vahd they teach us to wash our hands after we urinate."

The Yale man replies, "At Yale they teach us not to piss on our hands."

My advice to Pulpsicle would be: Stop pissing on your hands, stupid.

(IE user that's spyware free.)

Re:Yes, but...

[Anonymous Coward]

Not to mention the fact that what good are windows updates available in minutes that take minutes to install, when blaster (et al.) packets were sent to you seconds after booting into windows and taking seconds to infect?

Re:Yes, but...

[Anonymous Coward]

People usually don't wash their hands because they pissed on them. The point of washing your hands is that you have just had them touching an area of your body where sweat, etc. collects.

IE has a broken security model. The biggest flaw in Firefox is that it hasn't done more to reject that model. Instead, it has tried to offer feature parity even when the feature is risky. I still feel that it offers a better security model, just not as good as it could be.

Just what are they asking?

[Anonymous Coward]

> Can open-source software do the trick?

For money, or for candy?

Re:Just what are they asking?

[ScrewMaster]

As Lazarus Long once said, "Money is a powerful aphrodisiac. But flowers work almost as well."

Re:Just what are they asking?

[jcwinnie]

Candy? You get CANDY?

Free software pays for better support

[mpoli]

I am an IT consultant and I get a lot of clients who ask about the real cost of free software. Most business here are very cautious to choose switching for open source mostly because support for this solution is still somewhat more expensive than for the old paid solutions.

In the few companies I consult that are currently switching or have switched in the past, the Total Cost of Ownership of their computer infrastructured has lowered significantly, even though the cost of the support staff is truly higher.

But, anyway, support here is somewhat cheap, as I am in a developing country that pays a lot more for software than for the people running then in a number of times.

Re:Free software pays for better support

[Glonoinha]

You have pretty much hit on the key metric that is most often overlooked - the cost of the people running it.

Honestly most of the time the cost of the actual package (database engine, operating system, office suite) is inconsequential when compared to the cost of the IT staff required to support it. The minute you need to hire a new guy (or worse yet, a $160 / hour consultant or contractor) to support the environment - you can throw the cost of the package ($100 - $1,000 - even $25,000) right out the window because compared to $100k ~ $300k per year for an additional single person to keep it all running, the cost of the warez is inconsequential.

In the long run you save the most money by standardizing on a single platform - not for cost savings at the software license level, but because a single IT staffer can support it and support even more of them (by himself) down the road. Same thing applies to hardware - shave $100 per machine by going with home-built hardware, a different configuration for every single machine, and the minute you need to add a $50k / year (fully burdened salary) to the payroll all of your savings are not only gone, but blown completely out of the water.

The only way OSS is going to save a company money is if it lets fewer people do the same stuff, or lets the same number of people do more stuff - regardless of licensing costs. Most companies spend more money each year on executive perks and bonuses than software licensing, so you are pretty much on the money when you say focus on TCO.

Re:Free software pays for better support

[mretallack]

You have hit the nail on the head. In most medium to large companies, the cost of equipment and tools is nothing compared to the cost of wages. I remember seeing a graph of a generic company and at least 40% of the money through-put was wages. This means that if you can cut the complexity of using a tool, you can save more money then the cost of the high end tool. You only need to hire someone who is not a specialist. Not to sound like flame bait but this is the only difference between Linux and Windows when it comes to TCO. Reduce the level of experience required to support the IT infrastructure and you save a fortune. The only problem is that if a real problem occurs, no one knows how to fix it if there is no button on the GUI that says "fix problem".

Re:Free software pays for better support

[Daniel Dvorkin]

You have pretty much hit on the key metric that is most often overlooked - the cost of the people running it.

Are you kidding? This is never overlooked, because the anti-F/OSS crowd keeps harping on it. "Sure, you'll save $x,000 on software," they wail, "but what about the cost of wages? That will go way up, because open source is haaaard!"

Which, of course, is bullshit. The fact is, F/OSS IT solutions cost no more to administer than comparable proprietary ones do, and often cost less, because Oracle DBA's and the like make businesses pay through the nose. I fought a long and mostly successful battle to move my employer away from proprietary to F/OSS for our IT needs, and I built the infrastructure mostly from scratch, myself. Wages for proprietary software: one employee. Wages for F/OSS: one employee, who was a hell of a lot happier working with his choice of tools than with whatever crap a "solutions vendor" wanted to foist on us.

The upshot? We have a stable, working IT infrastructure, and because of the money we saved, the department was able to grow in recent years from one employee (me) to four, keeping pace with the company's growth from a four-person shop in a single office to a $30 million / year multinational. Granted, this may not be all that impressive by MegaConglomerCo standards, but we make a good product and a lot of people, including me, are pretty damn happy about how things worked out.

Re:Free software pays for better support

[Master of Transhuman]

And this is exactly why open source is beating commercial software.

Because open source is OPEN SOURCE - you can see the software and you can tweak it. Which saves money on contracting with a closed source company to do that, as you can likely find somebody to tweak it for less money.

Closed source companies won't even tweak it in many cases, for exactly the reasons you state - it then becomes a nightmare for them to support a hundred different versions of their software.

Also, a closed source company is going to charge mucho dollars per hour to do customization. They have fixed costs for such a project. An OSS guy can charge anything that gets him the contract, which is usually less, thus saving his client money.

City College of San Francisco library requested their vendor to add custom functionality to the new ILS system. The vendor agreed, then reneged later. Now an in-house person has to mess with it - probably unsuccessfully because the software is closed.

A second reason OSS beats commercial software in this respect is that it has been proven that a Linux sysadmin can maintain more servers than a Windows sysadmin. As more servers are brought on line, fewer sys admins - less employee cost to the company.

Also most of the COMPONENTS of TCO are the same for OSS and commercial software - except the licensing cost. Therefore, most of the time, TCO for OSS is less simply because the cost of consultants remains the same in many cases but the licensing and upgrade costs are less. Especially since much commercial licensing is per server, and OSS can frequently be used on ALL the servers in the corporation (unless you're buying Red Hat, of course.)

Finally, this argument has been done to death - the FACT of the matter is that TCO must be determined on an individual corporate basis, and EVERY single company that has moved to OSS that I have read about has decided that the TCO was better - frequently DRAMATICALLY better.

Re:Free software pays for better support

[kayak334]

And this is exactly why open source is beating commercial software.

Huh? In which universe?

Re:Free software pays for better support

[misleb]

Reduce the level of experience required to support the IT infrastructure and you save a fortune. The only problem is that if a real problem occurs, no one knows how to fix it if there is no button on the GUI that says "fix problem".

Right, so you hire a consultant and blow your wage saving out of the water. Or worse, double up on your IT staff to try to compensate for their incompetence. As a consultant, I can't tell youhow many companies I have gone to which had 5 IT people doing the job of 2 or 3 and they STILL needed to bring in consultants.

-matthew

Re:Free software pays for better support

[Master of Transhuman]

This one, moron.

Did you expect it to happen overnight?

You can't read the percentage uptake of OSS products over commercial? The double-digit percentage growth of Linux and other OSS products year after year? The trade journal articles that cite double-digit percentages of companies implementing open source?

Can you read at all?

Re:Free software pays for better support

[Glonoinha]

Wow - are you and HangingChad in cahoots?
Read my reply to him here

I never said I was pro-MSFT, anti-F/OSS, or even pro-F/OSS. I didn't name any technologies, I just said 'pick the one that lets one guy get the most stuff done.'

To apply it to what you said, if there was a technology (I didn't say MSFT, I didn't say F/OSS - I just said 'a technology') that let the company grow to be as large as it is and you alone could still support it all by yourself (without adding three other IT techs at $60k to $80k / year (fully burdened salary, not just take home : benefits, all the other stuff accountants load into 'salary') that technology would be the best bet REGARDLESS of per-seat cost, and REGARDLESS of whether or not you had access to the source code.

The other issue, of course, is that it doesn't matter what it costs the company for you while you hand build this thing using all kinds of little free tools that you personally love to use. When you get promoted or move on to another job - what is it going to cost them to find another guy (or guys) to support the environment in a manner necessary to keep their $30M / year business running? That is the true long term cost of (whatever technology) and it isn't something where either camp has a winning hand.

I ran an IT shop years ago, company went from $6M / year gross rev to $30M per year gross rev (it took about 5 years) and we had to add two IT staffers to support all the one-off home built computers in the company since we hadn't standardized on a single platform. Back then desktops cost $1,500 apiece, or we could build them ourselves for about $1,250 apiece - but the minute we needed to add two more people all our savings went out the window, ended up costing us $80k / year to support the hardware we only spent like $100k on in the first place. Talk about your lousy ROI. That was then, this is now.

For the record, it costs me more in software licensing (maintenance, actually) to run SuSE 9.0 Enterprise Server (a supported platform for WSAD 5.x) than it would cost for Windows 2000 / XP - but I develop on Linux anyways because I deploy to AIX, and thus am more productive since my dev envorinment is so much more like my production environment. The software costs more, but I am more productive, using OSS rather than MSFT. Luckily my company loves me and lets me do this - probably why I am working today without worrying about 'free overtime' :-)
How's that for funny?

Re:Free software pays for better support

[Anonymous Coward]

in the current one.
My company, because of hardware solution choice made 10 years ago, is now forced to change all the hardware because the old provider does not more support nor pieces replacement.
And we have to change ALL the system just because the software and protocols between hardware are closed and not standard. of course we can't (too onerous) and we must keep the majority of our old stuff for now. ...painfully.

With open source, this case will not appear.

Re:Free software pays for better support

[kayak334]

Wow. The fact that you're so upset about it kind of shows your bias.

So growth of linux and OSS products is now equal to "OSS is beating commercial software"?

Can you read at all?

Are you 12?

Re:Free software pays for better support

[shywolf9982]

1) You need experienced people to run closed source solutions aswell, cause we're talking about enterprise solutions, not notepad.

2) The open source solution is usually harder to set up and implement. But this is due to its mayor flexibility (ergo, it can be twisted to match your needs).

3) Closed source solutions have a oh-so-better-support is a myth. You probably didn't read the licenses you purchased

4) Once you've set up an open source solution, you don't have to put someone there to watch the app running. Really... the app won't try to escape outta the window. Believe me

5) It all comes down to what you want. You can hire a semi-alphabetized babboon to run your IT department. And he might be able to set up a windows network, but if shit happens (shit that the programmers didn't imagine could happen, and I saw a few of those things) he might not be able to save your beloved business.

What I and my boss always liked in open source solutions is the fact that we were given the opportunity to understand exactly what went wrong, and be able to understand if it was our mistake or the programmer's (all observed cases had been of the first class). And to me, that's a very valuable aspect because it allows me to work (I'm not paid to watch the apps running)

Re:Free software pays for better support

[DaveHowe]

I am occasionally considered good at supporting systems - crossplatform dba and sysadmin, cisco networking, checkpoint firewalls - but as a support tech, I didn't get anywhere near $100K or even half that - and I was the highest paid support tech there.
Training is another matter, but tbh again, that's a monopoly backlash - you say that people can be expected to know the monopoly package, and to a certain extent that's true - but working for a company with almost entirely an MS monoculture, we got literally dozens of calls that were due to users not having ever been formally trained in ms office (never mind any other apps they might use) - right down to having unmaximised a mdi child window in excel, and claiming thirty minutes downtime because of the helpdesk not recognising that as a possible cause of "my scrollbars are missing"
Personally, I think any company would welcome a chance to train their staff from a clean sheet - nobody convinced they don't have to listen because they already know "all about" microsoft office then having to burn support hours (and if you think onsite is bad, wait until you have to support the road warriors who want nothing better than to blaim missing their targets on poor IT support)

Re:Free software pays for better support

[Master of Transhuman]

Look, stupid, I don't have time to respond to juvenile taunts from idiots who have no clue what OSS is about or how well it's being taken up by corporate America.

Do everyone a favor - walk in front of a bus.

Re:Free software pays for better support

[kayak334]

You really do a great job proving your point. I know very well what OSS is about and have a good idea of how well it's being taken up by corporate America. You seem to be lacking in that regard with statements like "OSS is beating commercial software." OSS is surely gaining strength and doing very well, but I don't think "beating" is a valid word there. Commercial software is still, in fact, winning. That was my entire point.

I'm sorry if my questioning of your comment offends you as a person in some way. Maybe you should relax a little bit and realize that OSS isn't tied to your soul. You're probably a Linux zealot too, right? Linux is better than Windows and OSX and every other OS in the world no matter what, in every situation, right? ::rolls eyes::

Re:Free software pays for better support

[Glonoinha]

With open source, this case will not appear.

As someone trying to get Sybase ASE 12.5.3, WSAD 5.1.2, and a few other packages all running on a current distro of SuSE (either 9.0 Enterprise server, which Sybase doesn't play well with, or 9.3 Professional, which doesn't play well with Sybase and isn't a 'certified' platform for WSAD 5.1.2, or ... well you get the idea)
I assure you that this case will appear.

All the source code in the world isn't helping me get Sybase 12.5.3 working on SuSE 9.0 Enterprise Server.

Re:Free software pays for better support

[Glonoinha]

shit that the programmers didn't imagine could happen, and I saw a few of those things

Just finished a class that went through all the advanced stuff Oracle can do (most databases, come to think of it,) including pre/post triggers. If you want to throw even your most seasoned programmers for a loop, throw a few rude triggers (like stuff that just ignores an update, but only between 9:00am and 9:10am on every other Tuesday, and only if the updated field is at least twice the pre-update value) in there.

Just playing with stupid fun examples in class, we did things that would have non-DBA programmers questioning their sanity and ready to drive off a cliff. It was fun, to tell the truth, and quite informative in a manner that will have me appreciating triggers for quite some time.

Re:Free software pays for better support

[rtb61]

Now lets get to the real cost of crap software. It is not the licences and it is not the tech support. The real cost is having yoour staff sitting there idle because that windows network has crapped out again, it is about having to repeat the same work over and over and over again because windows crashed, it is about having to buy the same software over and over and over again, it is about retraining your staff again and again and again to use the same software but now different, it is about becoming security paranoid and spending increasing amounts of money ot IT security because of crap software design.

Microsoft might think it is big stuff, but for 99.999999% of businesses out there, it and it's software are just an out of control overhead with delusions of grandeur.

For a typical small to mid size busines swapping from a windows network to a Linux network means the difference between having on site support staff to calling a consultant about once a month or so, to make changes (after having the system in place for a year, you will swear in has only been in place for a few months because you will have payed so little attention to it).

Linux consultants cost more because they are in far greater demand. They are in greater demand because companies have learnt about the true costs of using M$=B$ crapware (chances are for each $1 of profit microsoft sucked out of you, it has cost you a $1,000.00 in losses).

Re:Free software pays for better support

[Master of Transhuman]

If you'd SAID that OSS software is doing well, but not "beating" commercial software, instead of mouthing off with "Huh? In what universe?", you'd have gotten a better reaction.

Particularly since I then could have explained to you that MY point was that OSS software was "beating" commercial software in the sense of being BETTER, not necessarily because of market share.

Which still includes the point that OSS is going to beat commercial software market-share wise EVENTUALLY, with Linux being taken up twice as fast as Windows for servers and the rest of OSS a no-brainer as soon as it develops enough enterprise-level apps.

In simple English, "beating" does not mean "already beaten."

Next time, try to make a point instead of being an asshole, and you might get treated better.

Naah, never happen.

Re:Free software pays for better support

[kayak334]

Sometimes it's necessary for me to be an asshole to debunk FUD that Linux/F/OSS zealots tend to spread. It's comments like "F/OSS software is beating commercial software" that make you a zealot. That statement is essentially false by every stretch of the imagination. It's the same thing as saying something like, "Well, this shouldn't be a suprise, since Macs are faster than PCs anyway." It's a nonsensical statement with no basis in reality. I'm not saying that F/OSS software is bad, or that Macs aren't fast. It's the assumption that one is correct without a doubt that creates the problem. So in conclusion, try not to be such a zealot, and you might not get so many replies like "Huh? In what universe?" In the meantime, I'll try and not be an asshole.

Yes

[Ckwop]

Absolutely. Two cases in point:

Case One: We were looking for a bug tracking solution and we had short-listed the contenders to a choice between Bugzilla, BugTracker and FogBugz. Although FogBugz was a superior product BugTracker won because we could modify it to suit our needs. We didn't like Bugzilla because of it's clumsy interface and the fact we'd need an extra machine to run it.

We saved money on the licenses and we got something we could modify and maintain ourselves. Free software at it's best.

Case Two: We were paying through the nose for anti-virus subscription and software. We all know that anti-virus software takes a lot of real estate. Most have *HORRIBLE* splash screens that no-one is interested in seeing and they tend to slow the machine considerably.

Our solution to the problem to the anti-virus problem was the Windows version of ClamAV. It has a nice outlook plugin that protects from e-mail based virus and we set a schedule to scan the disk every night. There is no "resident shield" in ClamAV but to be honest they rarely do any good anyway.

My former boss works at a much larger company (we're still good friends) and he's deployed the strategy across a company with around thirty machines and saved a fortune.

So yes, companies can save money using Open source. The hard part is convincing them that a not-for-profit organisation can deliver quality products. I find ten minutes with Firefox usually does the trick.

Simon

Re:Yes

[bburton]

Well, it's not always just about saving money. There's a lot of open source projects out there that are much less painful to work with.

Not having to worry about CD keys, crazy EULAs, spy/adware, and vendor lock-in are big pluses of most FOSS.

Re:Yes

[Anonymous Coward]

Looks like you forgot your possessives, mate. (Woops, would John Howard allow me to call you "mate"?)

Re:Yes

[killjoe]

People overlook this benefit. The administrative cost of keeping up with licenses can be huge. In some companies there are people whose full time job is to keep track of licenses and to make sure nobody is pirating.

I also love the fact I don't have to type in those stupid keys when I install software. What use do they serve anyway?

Re:Yes

[GotenXiao]

" Our solution to the problem to the anti-virus problem was the Windows version of ClamAV. It has a nice outlook plugin that protects from e-mail based virus and we set a schedule to scan the disk every night. There is no "resident shield" in ClamAV but to be honest they rarely do any good anyway."

Oh, the number of times my AVG res shield has saved my computer from Skynet, MyDoom, other assorted Trojans... AVG's a joy to use. It's light, it's fast, it's free (for personal use). And you get free updates. Usually at least one update a day.

Open Source and Money? Are you nuts?

[4v4l0n42]

Of course, it will never work. It never did.

That is why Novell and red Hat are making millions of Euros and Dollars even though they are OSS

.
XD

Re:Open Source and Money? Are you nuts?

[DaHat]

Oh yes... because Novell and Red Hat are such great examples of making money hand over fist.

Let us also not forget VA Software, one of the original poster children for making money through Linux

Re:Open Source and Money? Are you nuts?

[Seumas]

So let's see...

Companies want free software and outsourced labor for $6/hr developers.

So essentially the new American business plan is this:

1) Get free stuff
2) Get free labor
3) ???
4) PROFIT!!!

Re:Open Source and Money? Are you nuts?

[Anonymous Coward]

You just had to come along and pop everyone's happy balloon with the awful truth didn't you?

You should be ashamed!

Re:Open Source and Money? Are you nuts?

[Anonymous Coward]

>>Companies want free software and outsourced labor for $6/hr developers.

$6/hr! Are you crazy?

Why should I even pay one cent/hr when I can get geeks who work at McDonalds all day (or are supported by Mom and Dad) and then do OSS dev/support at night for free because "they do it for the love".

Now we're talking PROFIT!!!

Go Open Source!!! Fill my pockets!!!

Re:Open Source and Money? Are you nuts?

[cnerd2025]

What we are going to see is a move toware more Free as in Fredom software and OSS. Just because it's OSS doesn't mean it's free (rights or price), just that the source code is open. I think that more developers will be contracted to do work and will release their code under more OSS/FSF licenses. I have a feeling that there will be some Intel. Prop. case that comes up and challenges the validity of the US Copyright Law (DMCA et cetera) as well as WIPO. Supreme Court Justice Sandra Day O'Connor expressed her opinion following:

"The primary objective of copyright is not to reward the labor of authors, but [t]o promote the Progress of Science and useful Arts." "To this end, copyright assures authors the right to their original expression, but encourages others to build freely upon the ideas and information conveyed by a work. This result is neither unfair nor unfortunate. It is the means by which copyright advances the progress of science and art." -- US Supreme Court Justice Sandra Day O'Connor

. I think that this is a very forward-looking opinion, as well as backward-looking (to the Framers' intent). Because of the Internet, the largest open community in the world, ideas such as OSS will be able to progress and evolve. I am optimistic about such opportunities. I also believe that there will be money involved in OSS/FSF software for a long time to come. People still have to feed their families. You can't write software very well as a bum without a computer. I think that the Microshafts of the world will feel the impact of OSS and will be forced to adapt. That is capitalism.

Re:Open Source and Money? Are you nuts?

[Coryoth]

So share price is the ultimate measure of profitability? Those charts and data tell you very little beyond the fact that around 2000 there were some suckers who were stupid enough to pay exorbitant prices for shares in... well lets' be honest, any tech company.

What you might want to look at are Novell and Red Hat, and the statistics like "profit margin" and "gross profit". Are they raking in money hand over fist? No. Are they making a healthy profit, particularly for companies of their respective sizes? Certainly. Contrary to what you seem to want to imply, they are doing quite well.

VA Software? Yeah, well they're pretty fucked right now.

Jedidiah.

Re:Open Source and Money? Are you nuts?

[g2devi]

Novell and Red Hat are bad businesses to own?

Tell you what. I'll pay for any property or computer equipment they have and assume any debts. (I'll need an equity loan, but there would be no shortage of people ponying up the money.) If what you're implying is true, the owners of Novell and RedHat would gladly accept my offer to get out of their sink hole companies and move onto something that would make them more money. *snicker*

If you want to make money hand over fist, you more than likely need a monopoly. Competitive markets tend to give more reasonable profits. Novell and RedHat have competitive profit margins for such markets. If I were purchasing a product, I'd likely purchase one from a competitive market rather than a monopoly. Vendor lock-in is not a pretty thing to have to deal with.

As for VA Software, they failed because Linux got too popular (so it's being preinstalled by several vendors or consultants) and too easy to install (so sysadmins do it themselves). They didn't try to differentiate themselves from the cheaper alternatives, so they got burned. It has nothing to do with open source. It has everything to do with proper business planning.

Do we really even have to ask?

[yfkar]

If commercial closed software can do it, why couldn't open source software?

Re:Do we really even have to ask?

[CyricZ]

The main limiting factor is, like usual, time and resources. A product like Oracle, for instance, has had years upon years of time and millions upon millions of dollars poured into it. While the open source community can produce the mighty fine PostgreSQL, they just don't have the time nor resources to produce a product equivalent to Oracle.

Like it or not, open source projects are constrained by the same factors of production that any other good is constrained by. They can't be avoided, be it an open source project or a commercial, closed-source project.

Re:Do we really even have to ask?

[Master of Transhuman]

"A product like Oracle, for instance, has had years upon years of time and millions upon millions of dollars poured into it."

While PostgreSQL hasn't had scores of millions of dollars poured into it, they also haven't had the "years upon years" - although they ARE one of the older OSS products around.

Nonetheless, their achievements are impressive.

Most of Oracle's "features" beyond PostgreSQL are stuff involving applications development, tuning, and other stuff that most smaller companies don't particularly need or which are so complicated to use that most DBA's probably don't even understand them. Oracle is one hellaciously complicated product.

Oracle has more "feature-itis" than even Microsoft.

A better comparison would be MySQL which is younger and doesn't have all the features a good database should have - but it's getting them over time.

Given that most open source is less than ten years old, and open source project methods vary across the board from one-man projects to corporate-sponsored projects with hundreds of people, I think this form of comparison to closed-source software as to end results is a bit premature.

Open source is division of labor at its best.
As the open source methodology matures, I think we'll see no real limits on what it can achieve - short of putting a man on the moon in ten years.

Re:Do we really even have to ask?

[CyricZ]

But MySQL is heavily funded and developed by MySQL AB. It's far more commercial, even if it is open source, than PostgreSQL.

Re:Do we really even have to ask?

[Master of Transhuman]

MySQL is still younger than PostgreSQL.

And being funded better by providing a double license is why MySQL is improving quickly. The more OSS programmers that can afford to work on a project, the better the project is, usually.

But PostgreSQL is older and had more time to develop, so it's still the more fully developed product.

If PostgreSQL had the kind of money MySQL has, let alone Oracle, it probably would be better than Oracle by now. But it's pointless to discuss it, because that doesn't happen in OSS.

It doesn't mean PostgreSQL isn't a perfectly good database for the people that can use it. There's no need to have every "feature" - including the most obscure ones - of Oracle to be a completely adequare replacement for Oracle for those users who don't need Oracle's extra features.

All the trade journal articles have pointed out that up to now, you had to pay Oracle's price for Oracle's features - including the ones you'll never need. Now you can get MySQL and PostgreSQL and Firebird and others and get all the features you WANT for no or less money.

This is one place where direct comparison between OSS and commercial software breaks down. Commercial software HAS to be all things to all people. OSS doesn't. It can fork, be customized, do anything it needs to be useful to varied groups of people without being a bloated POS like Oracle or Microsoft Office.

The net effect, however, on commercial software is bad for them - people stop paying for stuff they don't need. Which means the commercial company can't charge the same anymore. Which means their profitability goes down. Which also means their customers save money. So in that sense OSS software reduces the TCO of even closed-source software!

Look at how many times recently Microsoft had to reduce its license fees for various governments that were considering open source.

Did Microsoft take that reduction into account when comparing their TCO against open source?

I wonder if any of the TCO studies take that into account!

Re:Do we really even have to ask?

[killjoe]

During the SQL server 6.5/7.0 timeperiod periodic flamewars would erupt on the usenet about oracle vs sql server. The softheads would always use the refrain "sure oracle has more features but sql server does everything you need for less money". Now MS is facing the same argument. Sure SQL server has more features but postgresql does everything you need for less money.

Oddly enough oracle and sql server are the same price for the same feature set now.

Why is this on slashdot?

[schngrg]

I find it strange why almost everything which says "Open source is good" or says "google is god" gets a slashdot mention.

I think slashdot moderators *seriously* need to be more unbiased.

Re:Why is this on slashdot?

[Anonymous Coward]

of coucse although india has some support for OSS the vast majority of it's bangalorebandits of the new wild east of software do not wna topen source. as they would have to show u[ their either a/ poor coding b/amazing code which owuld put everyone out of business and themselves out of a job.

i was just kidding there.

the real reason why OSS and ant-NDA culture is bad as that they can't *enslave* indian developers into monkey-we're so grateful for the pittance wages. if they came to you with already existing open sourced code or they jsut open sourced it themselves.. there's no reason why you cannot sell the finished product yoursleves locally cutting out the middle man...
fairs fair.

the indians seem very short sghted for all their talents they fear not being able to be favourable to the waves of offshore insourcing

well as wages go up and your expectations rise.. they will jsut go to chine or russia or south america anyway.

and you know ? if you indian programmers are so hot.. why arent you starting your own amazing game houses and what not? with free software and so much out ther i DO think you can compete.
after all you seem to have the talent in coding maths [and the imagination ..once upon a time.]

you seem pretty talented...
i am a computer science student in the west at a fairly advanced institution . my colleagues are nowhere near as gifted as my average old colleagues in maths or computer science. but you know what? they understand the value of acadmeically shared knowledge form whihc you benefitted. my country did not invent any computer langauges or any transistors yet here i am...

i see you dont release any of your code.. tahts pretty sad.

[except the one pice of code thats based around an academic proof of concept open to you published in Dr. Dobbs. .. if they ahd not published that code you would not have learned so quickly... and if you had figutred it out on your own like thise fellows did .. no doubt you would not have released it publically so readers of Dr Dobbs could get at it..]

dont you see the irony?

btw im not some awful racist....
this is the same speech i trot out to my own anti-oss compatriots in the middle east, who without OSS they would never have gotten net access due to american embargoes against enemy states and exporting tech.

but our academics were able to use open standards the net was built on and examine bsd source code and the tcp/ip networking stack to create their own routers and so forth.
thus we were the first countries in asia outside of iraq to have the net.

Re:Why is this on slashdot?

[cnerd2025]

Dude, /. was created by OSS proponents. There are no pretenses about the site. It glorifies the penguin and portrays Bill Gates as a borg. It is biased, but it's not pretending to be "fair and balanced" which is more than you can say for most other media.

Not always - of course

[Elixon]

> Businesses want to save money and boost IT > efficiency. Can open-source software do the trick? It's clear that it can. But it is of course risky operation (as any other business decision) so the OSS solution must be selected with certain level of knowledge. OSS can boost efficiency but of course not always. So generalizing is not good way to ask this question. What I don't like is the SW business using OSS for faster start up. Simply pretend to be totally FREE - get fast response, fast growing community and then find ways how to generate profit from it. It is really dirty trick. Better to be honest from the very beginning and clearly declare the aims to prevent the community from being surprised.

Free software pays for cheaper labour.

[Anonymous Coward]

"But, anyway, support here is somewhat cheap, as I am in a developing country that pays a lot more for software than for the people running then in a number of times. "

India?

Re:Free software pays for cheaper labour.

[mpoli]

India?

No, Brazil.

For a counter-example

[mparaz]

Remember Project David, the WINE ripoff featured last year? They're back.

Case Study Available

[Anonymous Coward]

Made in Japan - The Teriyaki Experience - in Oakville, Ontario, is running approximately 100 restaurants, from Newfoundland to British Columbia on a Point of Sale solution which is about 99% free software. They have an IT department which consists of one person. Using rsync, HQ has its central file systems updated from each location every 10 minutes. Nagios alerts HQ of the health of dozens of system critical threshhold variables in real time. The IT staffer can open a remote display on any location from his touchscreen notebook while sailing outside Toronto Harbor, monitor and control just about any aspect of system behavior via open VPN. There's nothing in the Windows world that can come close to this kind of retail automation or information automation cost savings. Free Software is, if you know where to look, providing cost savings that simply aren't possible with the Windows way of doing things, and providing superior solutions at the same time.

Re:Case Study Available

[Master of Transhuman]

While looking for more information on this, I found this article, which is interesting, too.

http://www.linuxelectrons.com/article.php/20031110 073608244

Another good reason to use open source is that a company's infrastructure vendor can reduce its own development costs by using Linux, and pass the savings onto their customers in terms of better and lower costing products - which in turn affects their customer's TCO.

This seems to be happening in the retail POS industry.

What cost the crashing

[Anonymous Coward]

I think we should be teaching our children about Intelligently Designed Operating Systems - clearly something as complicated as an Operating System cannot have arisen by 'programming', which was clearly proved wrong in the 1800's - observe the faliure of Babbage.
Operating systems crash because of the sins of their users.
As a Pastafarian I believe that a Bistromath is the one true computer.

Quite predictable

[bkazez]

Is there really any doubt as to what Slashdot's audience thinks about this issue?

Cost of the support people

[mpoli]

You are absolutely correct, Glonoinha, when you consider costs of a developed country.

Talking once more about me experience in Brazil...

One of the companies I consult for has just renewed their campus wide anti-virus licence (about 1200 desktops running Windows). The cost of the licence alone was comparable to the salaries paid to 5 avarage-to-senior full-time supporters for that same year (if you disconsider the almost 100% government taxes on labour).

And that was for the anti-virus alone. Think about the OS, office production package, ERP, DB, server licences etc.

Those numbers are completely different when you have costy labour, but in our case here, just the licences make A LOT of the TOC.

True cost of free software

[klubar]

Don't forget the cost of training and supporting employees. It's reasonable to expect every new professional employee to have fair to good skill in MS office programs and Windows. If the office uses something else, the training costs for new employees is probable a week or two. If business transactions (documents, presentations, email) need to be compatible with those of business partner's the cost of OSS becomes very high. The $250 of so for an Office license (for small businss) is small compared to the training costs or the difficulty in trying to send an almost compatible document to a client or customer.

Here's another couple of case studies

[DSP_Geek]

If you run a Windows shop and mess up on a few licences, even by accident, the BSA will come down on you like a ton of bricks.

http://news.com.com/2008-1082_3-5065859.html

As a matter of fact, they can screw up your operations by merely conducting an audit during your busiest season:

http://news.zdnet.com/2100-3513_22-996210.html

Even if you don't use the big-buck CRM packages mentioned in the article, if you're running a business the logical choice is to avoid the risk of extortion and/or business disruption by choosing open source and telling the BSA to stick it where the moon don't shine.

Re:Here's another couple of case studies

[Anonymous Coward]

>If you run a Windows shop and mess up on a few licences, even by accident, the BSA will come down on you like a ton of bricks.

That's just plain stupid. Try these on for size:

If you run a shop and mess up a few accounting records, even by accident, the IRS will come down on you like a ton of bricks.

If you run a business and mess up your accounts receivable/payable, even by accident, your suppliers will come down on you like a ton of bricks.

If you run a OSS shop and mess up meeting the requirements of the GPL, even by accident, the FSF will come down on you like a ton of bricks. (Arguably it's worse, because the BSA just requires you to keep track of a few lousy pieces of paper, something any competent IT staff should be able to handle.)

Re:Here's another couple of case studies

[killjoe]

"If you run a OSS shop and mess up meeting the requirements of the GPL, even by accident, the FSF"

only if you are redistributing the code. IF you are only using it in house then you are fine. Nobody will come down on you.

Re:Here's another couple of case studies

[Anonymous Coward]

>only if you are redistributing the code

I think that was implied, but thanks for the clarification.

ERP systems

[theid0]

From the article: "Such a shift toward open-source software for CRM and other business software applications, such as enterprise resource planning, is now beginning at corporations across the globe."

I've got three questions about this, from my experience in a manufacturing environment.
1) Where is this open source software that so easily replaces the commercial software?
2) How can I convince a corporation that has been dealing with a vendor for a particular product for many years that it is worth the pain of switching, even if the end result is good?
3) At what point is a piece of software "safe" and when is better to be more open? Take for example OpenMFG. They could be considered more "safe" than other solutions because they seem to have everything together in a well-supported way. However, their license is unacceptable for anyone who respects the open source ideology. The product isn't free as in beer and only partially free as in speech.

There doesn't seem to be a whole lot of choice in ERP/MRP, from what I see. Unless the manufacturers start programming a lot of code themselves and don't mind giving up their work to competitors, there will always be restrictions on what open source solutions are available. I would love to see a completely free MRP that does everything needed without a lot of hassle, and still lets my company customize it to save ourselves time and money.

Re:ERP systems

[Master of Transhuman]

"Unless the manufacturers start programming a lot of code themselves and don't mind giving up their work to competitors"

Well, if you buy a closed-source product, your competitors are buying it, too. Where's your competitive advantage?

Whereas if manufacturers support open source projects run by people either directly or not directly employed by themselves, but who are familiar with MRP needs, those people will write the basics - and then the manufacturers can hire those people as consultants to customize the suites to give them competitive advantage.

As for the pain of switching, closed-source providers will never even attempt to make it easy to switch away from their own product, although they might make it easy to switch from a competitor's product. OSS providers have every incentive to make it easy to switch - provided that they've been told that's a major requirement, otherwise they're likely to ignore it in the process of making the software.

Switching is going to painful, anyway - and what do you do if your closed source vendor goes out of business? You have to switch anyway, then. OSS doesn't go "out of business" - although a given project may become moribund. As long as you've got the code, the customer is still in business.

As for your first question, ERP/MRP products are just now being worked on in OSS. It will be a while - years even - before entire suites are available to replace PeopleSoft or whoever. Doesn't mean it won't happen.

It will happen quicker if manufacturers that want MRP start expending a small fraction of their IT budgets on supporting OSS efforts. A number of European companies are doing this, and the OSS products being produced are sophisticated and powerful.

Open source doesn't mean software that springs full blown from the brow of Zeus. It still has to be developed, debugged, and tested - just like closed source software. The difference is that the quality tends to be better (if not the user interface, in some cases).

The problem with a lot of large companies is this: they want quality software NOW - but they won't pay LESS - or anything - for OSS than they do for closed source. As usual, they want SOMEONE ELSE to pay for the software, then they'll consider using it.

They handle training the same way - they only hire employees that some other company has trained.

What's wrong with this picture?

Re:ERP systems

[whitespacedout]

Try weberp, compiere or erp5.

Maybe you don't actually need erp as much as automating existing processes so that data is moved around automatically instead of being processed manually. Eg: take stuff from a database and make a spreadsheet out of it with perl dbi and perl ooolib, mailing it automatically to the right person. This is non-disruptive and frees up employee drudge time, and reduces errors.

Maybe hire a geek to figure it out for you.

Have you given away your customizations?

[workboomer]

OpenSource is great and will work in the real world only IFF you give away your code after making changes to suite your needs.

So, Have you given away your version of Bugzilla?

If everybody is a 'taker' it doesn't work. But when organizations take it upon themselves to contribute back I believe there's no stopping Open Source.

Re:Have you given away your customizations?

[nostriluu]

I think you're wrong. At its heart, free/open source has to be useful to everyone who needs it. If everyone who used it was expected to "give back" every little change, there would be no coherence as a zillion features that are useless to very few people are jammed in.

At its best, the model involves a few coders at the top who develop features based on either community need, service for a fee, or their own itch, all of which are integrated in the core offering only if it makes sense to most people. Its the same old hierarchical model of control, but there's no other realistic way to do it.

Spoken like a true MSFT rep

[HangingChad]

You have pretty much hit on the key metric that is most often overlooked - the cost of the people running it.

A MSFT rep or a politically minded CIO. Looking at examples across my customer base the cost of support is not any higher for OSS applications, or applications built on open source products, exclusive of the licensing costs. I bill the same whether I'm supporting an application built with .NET and SQL Server or PHP and MySQL.

Provided you're not trying to support open source products with MCSE's there are very few instances where the cost of support is higher in either environment.

The minute you need to hire a new guy (or worse yet, a $160 / hour consultant or contractor) to support the environment - you can throw the cost of the package ($100 - $1,000 - even $25,000) right out the window...

You're assuming that you never have to pay a 160/hour consultant (which is a cheap consultant, btw) to support a proprietary environment. I spend more time fixing applications running on MSFT than OSS, using a per application comparison. A perfect example is the last security patch that went out. A customer called in panic mode because one of their critical applications was returning 401 errors. Auto-updates to the rescue. Now some are beginning to come to grips with the fact that apps designed in the current version of .NET may need to be rebuilt or modified to run on the next version of .NET.

Additionally, if you're counting on a vendor for support, you're likely to be disappointed. One customer uses a proprietary application for mapping that was developed in NY. If their system tanked tomorrow, they could get phone support tomorrow and they pay for that. If they need a body to do something, their choices are fly in a tech, at great cost, or call me, also at great cost. But at least I'm not billing for travel on top of that.

If you break down the hourly support cost in a MSFT environment, you'll be surprised that it's not that much less. And in many cases, I would argue, more. Then add licensing costs on top of that, way more.

What I'm even more surprised about is how MSFT PR media talking points tend to get passed around like established fact by people who couldn't figure out how to access a single machine in a room full of server racks. There's a reason I run all my personal projects on LAMP servers.

But, honestly, it doesn't bother me that much. As long as the bozos in charge keep parroting those talking points, I'm going to have a lot of billable hours.

Re:Spoken like a true MSFT rep

[Glonoinha]

Wow - that was pretty harsh, considering I never said which package was more expensive to license in the first place, nor which package I personally considered more productive or easier to manage (large scale,) or under which package I feel that developers are more productive under.

For the record, I do Java and database development for deployment to an AIX environment, and for my purposes Linux (SuSE 9.1 Pro, or SuSE 9.0 Enterprise Server) gives me the most effective environment in which to do my development. A support contract for SuSE 9.0 ES cost about $250 per year, every year (annual costs exceeding those for Microsoft 2000 Pro, which was my other choice) but the additional productivity I get in my Linux dev environment is well worth the additional money. No, my company won't let me run development environments that are not supported (with support contracts) so ... those are my options.

I never said MSFT was cheaper per hour, nor do I even accept that the license cost of OSS is cheaper than MSFT (given my experience otherwise) - what I said was 'find the technology that lets your people get the most accomplished, regardless of license price, and use it.' I also said 'whatever platform (OS/database/whatever) a single tech can support the most of before needing to hire another tech - that is the cheapest (by a LONG shot) regardless of per-seat licensing cost.'

If you get more stuff done in an environment, that's what I recommend (if cost is your concern.) There's a reason I run SuSE 9.x Pro at home (composing this post in Firefox in KDE right now.)

errrm

[Anonymous Coward]

"Businesses want to save money and boost IT efficiency. Can open-source software do the trick?"

save money - maybe

efficiency - no

small biz can struggle with Windows and Office they are dead in the water with opensource apps that may or may not work, require obscure dependancies and a backroom tech just to install.

Re:errrm

[chawly]

Done any Windows + data-base + backup installations lately ? Or did you maybe just watch it being done by a crowd of truely efficient (and duly certified) experts. These experts coming from the "front room", of course. As is well known, the front room is where all the light is collected (the money too) and it is where all the light comes from in small biz.

Open ended question?

[stinky wizzleteats]

Can open-source software do the trick? Cnet attempts to answer this open ended question

Yes.

...

(Not an open ended question)

I Agree with the GP but not entirely

[einhverfr]

You are aware, I trust, that Microsoft frequently sits on vulnerabilities for some time before offering patches. Your metric for security appears to have nothing at all to do with security.

Real security is not a matter of patching vulnerabilities. It is a matter of getting design right so that those vulnerabilities are both confined and minimized. Sendmail, for example, is a textbook example of how not to design a secure program. BIND is somewhat better but historically it had many of the same sorts of issues. Apache is pretty good. Qmail is quite good design-wise (I am hoping that someone fixes the superserver though to fix that vulnerability issue with environment variables. Technically, that is not part of Qmail, though Qmail depends on it :-( ).

99% of security issues should be addressed in the design phase. FOSS is no better than commercial software in this way.

In the end, FOSS *is* a large part of the solution. With *NIX, you get a powerful RAD environment for complex automation systems. This increases productivity. With FOSS, you are able to use these powerful systems without having to worry about millions of dollars in licensing fees and free up resources that would otherwise be spent tracking the number of per-seat licenses purchased vs. used... So yes, it makes sense.

blah de blah b;ah

[Anonymous Coward]

Did you need people to support CSS stuff? Yes. Do you need it to support OSS stuff. Yes.

OSS: you get cheaper up-front. you also get to some extent to make the package follow your needs, rather than change your needs to follow the package.

If you want to blame GP for benig ni cahhots with Hanging Chad, then why don't you take to task all those sycophants going "Yes, it's so true. Very wise. Excellent point. Thought of it myself..."

Re:blah de blah b;ah

[Glonoinha]

Yea, I was just messin' with GP / Chad. I'm guessing no relation, just two guys with good ideas at the same time. Only reason I said it was that both of them read more into my original post than I had written - I wasn't saying either technology or platform (commercial vs open source) was better, I was just saying that the big cost savings aren't found in per-seat licensing, the big cost savings is in more effective utilization of your employees.

The ones that said 'very wise, etc' are the ones that didn't read into my post that I was pro-commercial or anti-f/oss (good thing, because that's not the case) - they are the ones that agree that the savings of 'more work from the same number of employees' far outweighs the savings of 'per seat license costs'.

Oh god... don't yell 'bout it...

[Anonymous Coward]

...I must sell my MSFT & Oracle shares first...

The only reason for OSS unpopularity is the lack of common sense in the management... the management has been brainwashed to think this way : "A corporation is good if it earns lot of cash, if the corp. is good then they must have a good product." Now, let's try this on politics : "a political party is good if it has many members, if the party is good then the ideology also must be good"...

I just wonder why I'm not dressed in a brown uniform and sing "Waterland, waterland..." everytime I'm drunk... or why I don't consult a little red book all the time...
anyway, both cases are called propaganda...

OT: A stupid Linux question

[Glonoinha]

Totally off topic, but odds are the people reading this thread know the answer :

I'm grew up in the Microsoft world of networking, and before that did lots of Netware - now I am dinking with Linux (SuSE 9.0 ES, and others) and one thing that I have yet to even see mentioned is how to configure 'network shares' or 'a shared file system' on Linux. How do I do the equivalent of the following in Linux:
a) set up a particular directory tree on my Linux box as shared,
b) configure rights for users that are not actually logged in to the box (not telnet'ed or SSH'ed or FTP'ed in, just 'connecting' their Linux box (or Windows box, for that matter) to my shared drive, and
c) do the equiv of "C:\net use y: \\myLinuxBox\myShare" from my Linux box to the shared folder described in a) above.

I keep hearing 'use Linux as a server' but never 'use Linux as a file server' - implying that I can use it as an application server for web services or FTP or telnet or ssh or database server (all of which I have done, for fun or profit) but ... what's the scoop on setting it up as a file server?

Yea I know, lamer - but if I don't ask, I won't know.

Re:OT: A stupid Linux question

[bburton]

Samba

Corporate CTO == Medieval Alchemist

[sinewalker]

TANSTAAFL
TANSTAASB (silver bullet)

But that won't stop Business looking for both... rather like Alchemists and the Philosopher's Stone. Maybe, one day, Corporate CTO's will go the same way as the Alchemists, hopefully by blowing themselves up... one can but hope.

Businesses aren't asking about the cost

[blackhedd]

As part of a panel discussion I moderated recently that included the participation of CIO representatives of about 30 major companies, we learned that larger businesses are:
1) VERY interested in F/OSS;
2) are NOT interested because of the potential cost savings, but rather because they believe that F/OSS can offer better technology with shorter delivery cycles;
3) are going slow because of the relative lack of enterprise-friendly support options.

If anyone is interested I can maybe write up a journal entry about this.

Firefox patches

[sr180]

Consider this: I have downloaded patches for more security flaws in Firefox than for IE in recent weeks

Your post is a blatent lie. The last Firefox update was the 25th of July. Its now the 22nd of August. There havent been ANY patches to download in the last MONTH

Omaha is a Linux-friendly place

[ubiquitin]

There's a nicely compiled list of Linux companies in Omaha. My experience is that companies in midwestern cities are slightly less eager than big cities/big co's to jump on the bandwagon when it comes to purchasing high-dollar solutions from Microsoft and the like. The local user group is pretty active too.