This may be dangerously close to a rant, but how about a set of principles to determine the actual cost of things that happen on computers?

A certain US state is currently laying charges against a former employee for installing a screen saver on a network (in the range of 100/200 computers) They are alledging that this screensaver has cost them approx $420,000 in used processor cycles.

Various financial organisations are today claiming that Code Red will cost the economy $8.7 billion.

How about adopting a formula for process usage that is based on the amount of times a (user space) process causes other processes to be waiting multiplied by a percentage of the machine's TCO representing that part of the machine life during which the processes were delayed? e.g

cost=#delayed processes * (TCO/(total machine lifetime/process delay time))

i.e. in the screen saver case the only cost incurred (presuming that the machine was otherwise unoccupied) will be during the delay between a user dismissing the screensaver and it's recognition of the interrupt.

In the worm example we could remove the cost of tech support applying the required patches (which they should have been doing anyway) and just count the time required to repair any damage in cases of actual infestation.
-- key-aero, Aug 02 2001

Aaaaaaargh! Another one. http://www.theregis...ontent/6/20773.html
And think of the cost of toilet breaks! [key-aero, Aug 02 2001, last modified Oct 04 2004]

My projects have to cost nothing and be ready yesterday.
-- angel, Aug 02 2001


I think it's a product of the same (small, but persistently annoying and dangerous) segment of corporate America that ardently believes that it costs $2,973.74 to dispose of an obsolete PC. Yeah, it might if you had your $100k+ CIO personally cut it to pieces with a chainsaw, fly it to Hawaii, and drop the pieces into the mouth of an erupting volcano... as opposed to, say, chucking it in the trash or a storage room.

Part of the problem lies with the way accountants calculate costs. An old PC occupying 3 cubic feet of space in a storage room whose FASB-defined value is $17.39/month per cubic foot isn't REALLY costing the company $52.17/month unless its presence there is displacing something else. If the storage room would otherwise be empty, it's a sunk cost, pure and simple, and any attempt to imply otherwise is intellectually and ethically dishonest. Not to mention total bulls**t.

Put another way, if the company has an empty storage room and a hundred apparently obsolete PCs, is there ANY sane person who'd argue that the company would save money by disposing of all 100 PCs immediately instead of putting them in the room in case somebody happened to find something useful to do with them someday before the space were actually needed for something? Believe it or not, I've had arguments with accountant friends of friends who seem to genuinely believe immediate disposal is preferable to incurring $2,000/month in fictional storage costs. Crazy... absolutely, positively, crazy...
-- miamicanes, May 17 2003


That is a backwards way of accounting. The only similar accounting method I've come across is used for resource management and is when you calculate the cost of a room spread over its contents and see if the cost per item is acceptable. An empty room costs infinity per item so is clearly unacceptable. One of the support issues I came across was that users expected the costs to show up for the items as well as for the building.
-- marklar, Sep 12 2007


The storage 'cost' of the redundant computers comes about as a result of an internal charging system. All teams/departments etc will be charged for the floor space they occupy (including cupboards). Presumably, in miamicanes example, the owner of the cupboard is not the owner of the computers. It is therefore entirely reasonable for the cupboard owner to pass on his floorspace costs to the computer owner. This is not, though, a 'cost' of storing the computers, it is merely a charge for doing so. The two things are not the same.
-- DrBob, Sep 12 2007


All screensavers by definition (sort of) only run when the computer is idle. SETI@home uses only spare cycles so even if you are rendering some 3D when the screensaver comes on it will not take processing power away from other tasks.

The only additional cost is for extra electricity, office cooling and bandwidth, which is minimal. I had to work this out once before I could install Folding@home on office computers. From memory it was about $10 per year per machine.
-- marklar, Sep 12 2007


//I've had arguments with accountant friends of friends who seem to genuinely believe immediate disposal is preferable to incurring $2,000/month in fictional storage costs.//

That reminds me of how corporate executives become near suicidal any time their profit percentage number goes down. In 2005 they might have made a 28% profit on $4 million in sales but in 2006 they made only 23% profit on $18 million in sales. And when I hear stories like this, I wanna pull my hair out wondering if they realize that their profit is $3 million higher than it was in the year before.
-- Jscotty, Sep 12 2007


Actually, [Jscotty], it's not such a bad thing for executives to sweat over that; it means they're worried about being less *efficient* custodians of other people's money. If I were one of their shareholders, I'd be pleasantly surprised that they cared.

In other words, though I'm not capitalism's biggest fan, the attitude you describe is a sign of capitalism working as it's supposed to, and not crazy at all.
-- pertinax, Sep 13 2007



random, halfbakery