Buying v. Leasing

Story: Total cost of ownership: an expensive red herringTotal Replies: 4
Author Content
number6x

Jun 22, 2005
7:08 AM EDT
Maybe Microsoft's estimates of GNU/Linux's TCO are accurate. Let's just say, for arguments sake, That they are accurate. Its the estimates of Windows TCO that are way off.

After all to calculate the "Total Cost of Ownership" of a Windows based setup implies some form of ownership.

With Free and Open Source products you actually get something that approximates ownership of the code. Sure the copyright holder really "owns" the source code, but you can manipulate it as if it were your own creation. There are very few restrictions on what you can do with the code. If its GPL, you've got to re-distribute your versions, if its BSD or Apache you have to retain copyrights and give acknowledgements. But from the point of view of building products and using the code, it is very close to ownership.

Microsoft offers some "shared source" initiatives, but these are extremely restrictive. You may not even own the code you create if you sign up for these programs. That would give you less ownership.

The only way to "own" Windows would be to buy out Microsoft. This means that the first approximation for the buyout would be to calculate the market value of Microsoft. Then add any physical assets like buildings and production facilities they own. Add in all their debt, because you would have to assume that too to "own" Windows. Add in all the costs of the negotiations and writing up the contracts. Paying off a lot of executive "Golden Parachutes".

Of course Microsoft has $40B-$50B US in cash reserves. If they don't want you to "own" Windows, that could make for a pretty big poison pill stock buy-back program, pumping up the market value pretty quickly (hence the value calculated above was only a first approximation).

Yes sir, calculating the TCO of GNU/Linux is probably pretty accurate. Sure, I know, the copyright holders retain "ownership" of Linux, but the actual difference between ownership and licensing is so small with F/OSS that TCO estimates apply pretty well to Free and Open Source software use.

With Windows however, your don't really "own" it, you are just leasing it. So calculating the Total Cost of Ownership is not an accurate measure of use, unless you are considering a buyout of Microsoft that is.

:)
dinotrac

Jun 22, 2005
9:43 AM EDT
number6x -

You don't really "own" any code unless you own the copyright.

For businesses, the TCO (a mythical creation that nobody can ever properly calculate) is comprised of hardware, software licensing fees (to the extent they exist), and the cost of the people who make it all work and keep it going. At the place I've been working, we've got some fairly expensive FOSS software...expensive because it's API is a nightmare, because the developers don't adequately regression test releases and because the documentation is inadequate.

Fortunately, we've been able to put a reasonable shell around the API (that will also let us switch to another platform should the costs outweigh the benefits), and have the resources to do our own regression testing before applying new releases. Along the way we've filed a few bug reports and submitted a patch or two.

Not everybody has the resources to do that. For them, this tool might cost too much, even though the license is free.







number6x

Jun 22, 2005
10:00 AM EDT
dinotrac,

Please note the smiley. It was a feeble attempt at humor.

You are correct about some software. I think all developers should be forced to spend a year or two developing in COBOL on mainframes.

This way they would learn:

1) How much better modern languages are for developing systems.

2) What stability, regression testing, dependabilty, and uptime should be expected from applications.

(I'm partly serious here, so no smiley. You can't appreciate what we have now if you don't know what we had then!)
dinotrac

Jun 22, 2005
10:12 AM EDT
n6x -

Funny thing is, for however wonderful modern languages, tools, etc are -- and I'm with you, I'd hate to turn back the clock -- I'm reminded that much of the Y2K scare a few years back was the result of ancient code that just kept working and working and working and working...
Tsela

Jun 22, 2005
10:59 AM EDT
number6x:

How right you are. My current job (first true IT job ;) ) is to maintain and develop an in-house mainframe database application which is the backbone of the company: a healthcare insurance. It's a Unisys mainframe and the application is written in LINC, a 4GL database language derived from COBOL (but with a worse syntax, if you can believe it! :) ). It actually compiles in COBOL (and parts in ALGOL too, for coroutines for instance). You just cannot imagine the amount of self-discipline necessary to write good (well, good enough really) code. And maintaining the code of people who weren't that disciplined is a nightmare. And I'm not even talking about all the wrestling with the stupid version control system and with the structure of the code itself (the source code itself is structured as a database. It makes searching for things especially painful)... And the Dutch government plan to reform healthcare has just been adopted, so we have to make extensive changes to more than 75% of the existing code, thoroughly test it, and all this within six months. I do like a challenge, but this is ridiculous :) .

You can imagine how relieved I feel when I arrive home and I can code in my beloved Ruby ;) .

dinotrac: Funny you mention the Y2K bug. As it happens, LINC was sensitive to this bug (its built-in date primitives worked with a year coded on two digits), but our application wasn't, since those date primitives have never been used in it: somebody had the intelligence, from the beginning of the program, to write special date logics which use a date coded on 4 digits. Of course, that makes the program sensitive to the Y10K bug, but my company is planning to replace the mainframe application with something new anyway, and at the speed they're building it right now it should be ready by the year 3521. So no worries here ;)) .

At least I get to work on a project that fits my (and RMS's) software ethics ;) . And as I said, I like a challenge, and I have great colleagues ;) .

Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]

Becoming a member of LXer is easy and free. Join Us!