30th Anniversary of the (No Good) Spreadsheet 407
theodp writes "PC Magazine's John C. Dvorak offers his curmudgeonly take on the 30th anniversary of the spreadsheet, which Dvorak blames for elevating once lowly bean counters to the executive suite and enabling them to make some truly horrible decisions. But even if you believe that VisiCalc was the root-of-all-evil, as Dvorak claims, your geek side still has to admire it for the programming tour-de-force that it was, implemented in 32KB memory using the look-Ma-no-multiply-or-divide instruction set of the 1MHz 8-bit 6502 processor that powered the Apple II." On the brighter side, one of my favorite things about Visicalc is the widely repeated story that it was snuck into businesses on Apple machines bought under the guise of word processors, but covertly used for accounting instead.
Re:Why use MUL/DIV (Score:5, Interesting)
Why use MUL/DIV --When you have shifts?
Well, shifts and adds. For multiply.
Shift, subtract, jle for divide. :-)
Also, remember that when multiplying 8 bit numbers with 8 bit registers results in a 16 bit result. Its not as easy.
I wrote a whole 32 bit math package for the Z80 "back in the day."
Re:Why use MUL/DIV (Score:5, Interesting)
You can do floating point in software, it was done all the time in the 8 bit days, in fact it was done all the time right the way into the 486 days (the 486sx, IIRC, lacked an FPU). It's just not all that fast. But for the size of spreadsheet you could make on a 32K RAM system, the speed of the floating point calculator probably wasn't much of a factor. It wouldn't surprise me if the spreadsheet authors used the BASIC ROM's floating point routine, if it has one (I have no experience with the 8 bit Apple machines. The 8 bit stuff I do play with, like the Sinclair Spectrum and BBC Micro, have floating point calculator routines you can call).
Only left and right arrows no up and down (Score:3, Interesting)
I've used visicalc. I was just a kid, but my friends dad had it for the apple ][+.
It was weird, because it had no up down arrows on the keyboard, you had to toggle up/down left/right mode by hitting the spacebar.
Love it or hate it, visicalc made computers way more useful. I don't think it was a bad thing
Re:Instruction set. (Score:3, Interesting)
Tooting my own horn... (Score:5, Interesting)
Speaking of 6502 programming feats, back in 1982 I worked for Acorn computers in the UK, writing software for the 2MHz 6502 based BBC microcomputer, which incidently we also used as our development machines. The BBC micro had 16K of ROM for the built-in BASIC interpreter and low level "OS", another 16K of address space into which you could map any one (at a time) of the other 16K software ROMs in that machine, and 16K or 32K of RAM depending on the model. Much software was sold on ROM - there were four sockets built-in or you could get expansion boards to allow more - but only one at a time could be selected since there was only 16K of address space for these.
One project I did at Acorn (with another guy) was to implement a Pascal development system for the BBC micro that we crammed into two of these add-in 16K ROMs. This was no cut down version - is was a full-blown ISO certified version of Pascal, the first ever implementation for a Microcomputer to implement the standard and achieve ISO certification (ISO Pascal is different from P-system Pascal which had preceded it).
So, what we fitted into 32K was:
- An ISO Pascal compiler, which compiled programs down to a P-code like stack-based virtual instruction set
- A virtual machine/interpreter for the instruction set
- A 6502 machine code relocator
- The complete Pascal run-time library (full floating point, IO library, heap, etc)
- A full-featured full-screen editor with regex find/replace (with as-you-type syntax parsing and highlighting), block copy/move/delete, etc (in only 4K of code)
- Command line interpreter
Now bear in mind that only 16K of this could actually be in the address space at one time...
The way we managed to squeeze all this in was to have the compiler in one 16K ROM, and the rest in the other. The compiler was written in ISO Pascal and self-compiled to our virtual instruction set. We had to add a few "macro" instructions especially for the compiler in order to get it under the 16K limit. The rest of the software (which I wrote) was all in 6502 assembler. Now consider that to run the compiler you also needed the virtual machine, but that was in a different ROM which could only be mapped into the same address space as the compiler (hence replacing it)... What I did was organize the VM/interpreter into pure code, pure data, and relocatable data (address tables), and implement a 6502 machine code relocator (recognize each instruction type, and know how many byyes they were, and whether they had an address component that needed relocating) which copied the VM out into RAM therefore allowing it to co-reside in the address space with the compiler.
It was a very fun project, not only because of the technical challenge (this was my first job out of college), but also very much because of the memory constraint. I had to use every 6502 trick in the book to eliminate every spare byte to squeeze the assember half of it into it's 16K ROM. Those from this generation may remember things like using XOR A, A as an alternative to LD A, 0 to save a byte, changing tail recursion/calls to jumps (JSR subroutine, RET -> JMP subroutine), taking advantage f known processor flag state to use 2 byte "conditional" (but not if you know the state) branches in place of 3 byte absolute jumps, etc, etc.
Toot toot!
Re:Wow (Score:5, Interesting)
Dvorak is an idiot. To use the old adage: "Guns don't kill people. People kill people."
If a bank trusts a spreadsheet based on a bad formula that is provided by the bank itself, is it the spreadsheet's fault? If the CEO chooses that saving 1 cent a year by outsourcing the call center to India, is that the spreadsheet's fault? Please.
There's a lot worser things than people using spreadsheet formulas. For instance, people not using them. Have you ever watched someone with no accounting or technical knowledge enter a bunch of figures in a spreadsheet then turn to the desk calculator to sum them up, turn back to the computer and key in the result? That's almost as bad as how somehow "worser" got into my spell-check dictionary and now I can type it without complaint!
Borland Did... (Score:3, Interesting)
Back in the late 80's, Borland wrote all of their compilers in Assembly. That's how they were able to compile 27,000 lines of Pascal code per minute on a '286 machine.
I shudder to think of the difficulty of that endeavor.
Re:What if... (Score:3, Interesting)
How would I play Tetris? [excelgames.org]
Re:Why use MUL/DIV (Score:3, Interesting)
the 486sx, IIRC, lacked an FPU
You are indeed correct. I still have one running Linux and acting as a router. Works just fine and consumes so little power (compared to modern systems) that I haven't been inclined to replace it. Has an uptime of >1,000 days too. I was shooting for 2,000 until we had a 71 hour long power outage that exhausted my UPS.
MBA mentality (Score:5, Interesting)
I think that Dvorak is putting too much blame on the spreadsheet: it was just an accelerant on an already-burning fire. As Frank Zappa said when asked, "What do you think happened in this country?"
It's the user not the tools (Score:5, Interesting)
Dvorak doesn't even know the difference between finance and accounting. Accountants don't really spend a lot of time with spreadsheets. I know because I am a certified accountant myself. Spreadsheets are used far more by finance analysts. Accountants track what happened in the past much like a secretary keeps minutes of a meeting. Finance analysts try to predict what will happen in the future and their main tool is the spreadsheet. These are not normally the same person though there obviously is overlap. Anyone who actually knows anything about business understands the difference but apparently Dvorak is not among them.
Dvorak blames for elevating once lowly bean counters to the executive suite and enabling them to make some truly horrible decisions
Right, because no one ever made horrible business decisions before the spreadsheet. Sigh... Used properly, spreadsheets let us make more informed, rational decisions instead of shooting from the hip. Modern finance would literally be impossible without spreadsheets or something very much like them. Ever tried to manage a company's books? Without spreadsheets and accounting software you need an army of workers to track the paperwork and calculate the numbers. Furthermore hand calculating results in errors and lots of them. Sure spreadsheets can be used badly like any other tool. They certainly are no magic cure-all for bad analysis and decision making. But that's the user not the tool.
Dvorak asks in the article:
How often in years pastâ"the pre-spreadsheet era, that isâ"did an accountant take over a company?
Frequently. John D Rockefeller was an accountant before he was a titan of industry. There are countless other examples. Accounting is what allows managers of businesses to understand what is going on. Every business manager is by necessity an accountant to some degree. Without accounting they are no different than an airplane pilot without any instruments. It should surprise no one that the people who understand the cash flows best often rise to positions of control, including the role of CEO. A spreadsheet and other computerized tools simply make the job easier and more productive. Apparently Dvorak thinks we should rely on slide rules and multiplication tables and ledger books instead.
Dvorak further asserts:
Cars are shoddy, consumer goods are junk. Toxic substances are in the food supply. Lead is in toys. Most of what we buy is made cheaply elsewhere.
Further evidence of Dvorak's stupidity.
Re:Cannot believe I am saying this... (Score:5, Interesting)
I was once promoted to manager. And for about 3 months I loved it. And then I didn't.
My wife who is an engineer like me is a VP/Director. She makes quite a bit more than I do. But guess what, I see the stress my wife has. Me I have stress, but not that type of stress.
You see with engineering I am in control. With management, you are herding cats!
Re:bad analogy - think crank (Score:3, Interesting)
I really wish to know what is the real reason for "non IT folks" to start messing up with "the data"
I can think of a few, but I cannot decide on the main one
In any case, there is often a disaster at the end of the line, a pity that the original employee has already being promoted :-)
Re:bad analogy - think crank (Score:3, Interesting)
This fills me with dread. Hitherto, when you reached the 65k row limit, you knew you were doing something wrong. You'd normally rethink what you were doing, reorganise your data in a sane way. Maybe you'd even work out what that 'Access' program was for.
What horrors await when clueless users carry on doing their business all the way down to a million rows, I fear to imagine. I've already got a 120MB spreadsheet someone made which actually contained about 300K of data, a masterpiece of duplication upon duplication. Now Microsoft offer these people a million rows.
People use Spreadsheets for general-purpose things (Score:4, Interesting)
Spreadsheets are often used for purposes which go above and beyond their intention, acting in some cases as almost a general-purpose programming environment.
Since this abuse is so common, why not take it to the next level and make a programming language which acts like a spreadsheet?
http://www.subtextual.org/ [subtextual.org]
Re:Why use MUL/DIV (Score:3, Interesting)
up to and including the 80386 the Floating Point unit was a separate chip with the number 7 at the end instead of a 6 or 8.
the 8086 and 8088 had the 8087 FPU.
The 8088 which was a version of the 8086 with a 8 bit external data bus, instead of 16 bit, and was the actual processor the Original IBM PC and the XT was based on.. NOT the 8086
http://en.wikipedia.org/wiki/IBM_PC [wikipedia.org]
the 80186/80188 had the 80187 FPU. the 80187 was a rarely used for PCs (IBM didnt use it) as it lacked full compatibility with the 8086 real mode addressing, but was used for some computers such as the RM Nimbus
The 80286 had the 80287 FPU or the 80287XL FPU (which was really a 80387sx with its extra instruction set).
the 80386sx (basically an 80386dx in a 80286 bus/packaging, to fit in 80286 motherboards) had the 80387sx.
the 80386 (full 32bit internal and external) had the 80387dx FPU coprocessor.
the 80486 was a pipelined evolution of the 80386 with 8kb of cache, and the 387 included. The 486sx was initially a normal 486 with the FPU (387) disabled. the 80487sx was actually a full 486, with a slightly different packaging (to prevent it being used instead of the slightly more expensive 486dx). when the 487sx was inserted into the socket of the motherboard, the existing 486sx chip was disabled, and the new chip "took over". Yes it was a bit of a con job. They did it again with the so called "Overdrive" processor.
http://en.wikipedia.org/wiki/Intel_80486 [wikipedia.org]
The Pentium was effectively two 486dx chips on the same die with hardware to automatically balance the instructions across the two cores.
Re:VisiCalc binary is still available (Score:4, Interesting)
I'm with AC, why shouldn't you run it on your main computer? I don't have dosbox here, so I can't check :-)
I remember reading somewhere that the Visicalc executable is used as part of Windows testing, to make sure that (really) old DOS programs still run without a problem. Can't find a citation for it at the moment, though.
Re:Why use MUL/DIV (Score:4, Interesting)
What's your throughput like on that beast?
It's sufficient to handle my 8mbit/s cable connection and the iptables/traffic shaping rules I have setup that go with it. I've never tested it with a local connection to see just how much throughput it can handle but I can easily peg my internet connection without any problems. It has 128 megs of ram and a 8 gig hard drive.
I've thought about replacing it many times but why bother? It consumes around 20 watts and doesn't even require any fans besides the one in the power supply. Any newer computer is going to consume at least two or three times as much juice and be a lot louder.
Re:What if... (Score:3, Interesting)
Re:Cannot believe I am saying this... (Score:3, Interesting)
Plus we finance guys generally know how to drive a hard bargain.
Re:What if... (Score:3, Interesting)
Or, instead of an RDBM to manage all of this info, how about a meta-database that allows you to take all the outdated spreadsheed processes of today and put them in a managable solution that allows everything to relate together.
Most of the products that do this today are focused on specific and well financed areas of the economy. Of course, if the easy money is in regulatory compliance and risk management, it would make sense to focus there. On the plus side, once companies buy products like this, often times it becomes viral and eventually takes over those non-wealthy group's processes, improving visibility across the whole company. Almost like one giant intelligent spreadsheet to run the company :)
The two that come to mind for me are Archer Technologies (http://www.archer-tech.com/) and SalesForce's new Force.com stuff (http://www.force.com/)
There are others out there, but not nearly as mature as those two.
Spreadsheets hide the Black Swans (Score:4, Interesting)
The article makes me think of an idea that I believe is found in the book, The Black Swan, by Nassim Taleb [wikipedia.org]. If it's not there, it's in his earlier book.
The idea is that human thinking is prone to several faults. It's a flaw in the construction of our brain. I believe this is something evolutionary psychology talks about. Anyway, Taleb says that human beings are woefully prone to looking at the past and convincing themselves that past data is a sure guide to what the future will bring. His idea is not just that we are prone to this mistake, but that in effect we love making this mistake -- or perhaps closer to his point, we feel great when we are making this mistake.
Putting Dvorak's article in this context, people look at spreadsheets -- and at all the wonderful graphs and charts you can make from the data contained therein -- and are lured, cognitively, into painting a particular picture of the future. To the natural inclination of our minds, this picture is so beautifully convincing that we have to actively work to resist its charms. It's almost as if we can't help buying into the future our inclinations, with the help of our spreadsheets, sell to us.
This is a small example from what is a larger problem in economics, which only some schools of economics recognize: namely, economic history is a poor guide to the future. (By contrast, the entire field of econometrics [wikipedia.org] posits itself on making "mathematical predictions" based on economic history.)
Dvorak's an ass, in my opinion; but I think he may have stumbled onto something here.
Re:Cannot believe I am saying this... (Score:1, Interesting)
Re:Loooooong time (Score:5, Interesting)
It's amusing that you don't know how wrong you are.
One of the very first uses of visual representations of data was by Florence Nightingale to Queen Victoria, a chart showing the relative causes of British military deaths in the Crimean war. The whole reason for the chart being that yes, Nightingale thought Queen Victoria's eyes would glaze over at a table of numbers and she wouldn't be able to comprehend it.
Graphs of functions may have been created for mathematicians. Charts of data were invented for people whose brains ran out their ears at the sight of numbers.
Re:Unfortunately, you're a commodity (Score:3, Interesting)
What you need to compare is not salaries, but the value each individual brings to the company.
Not true. It is patently impossible to measure the 'value each individual brings to the company'. Even if it were possible, salaries could still not be based on this. Many positions are secondary or tertiary to the real money-making operations of a company, and do not generate any money directly, wouldn't do well by your measure; but they are required for the company to operate. In truth, a person's salary is determined by many factors: the minimum wage, the amount of training and experience need for the position, the amount of time worked and the responsability of the position are a few.
As regards the pay ratio, US company presidents have gone from earning 40x more than what factory-floor workers did ~50 years ago to ~200x more today. Europe and Japan are still at the 40x level. Why? US business executives are corrupt and self serving.
Yes, I actually read it (Score:4, Interesting)
And no, the spreadsheet is not responsible for all moral decay and infamy in our society.
Cars are shoddy, consumer goods are junk. Toxic substances are in the food supply. Lead is in toys. Most of what we buy is made cheaply elsewhere. At every level of the business scene today, some bean counter does a what-if calculation before making the decisions. The spineless CEO worries about what the shareholders would think if he disagreed with what the spreadsheet and the CFO told him to do. To make him feel better, the board will give the CEO a fat bonus for saving money.
Back in the day, before spreadsheets, the US Military secretly gave the children of US servicemen whooping cough, prisoners were secretly injected with syphillus, and a deal was cut with lead paint manufacturers to leave their remaining inventory on the market rather than recalling it after everyone had given up denying it was harmful. Don't get me started on lead gasoline or cigarettes.
Amoral behavior like this is a property of all *secretive and powerful institutions*. Since the spreadsheet has been invented, private corporations have become more secretive and more powerful, and their directors have become more dedicated to institutional goals as a cultural shift. There is no causation here.
There is a lot of criticism of administrators on this thread - and it is certainly true that the administrators of powerful, secretive institutions tend to personify both the destructive social impact and caustically short-sighted, self-interested purposefully ignorant culture of the institutions where they hold sway.
But as has been pointed out elsewhere, for human beings to act productively and cooperatively, administration and logistics are required. Spreadsheets help with this task immensely - as anyone who's tried to for fucksakes budget a camping trip (how much more would it cost to bring uncle David and his kids too?) can attest.
To clarify my assertion further (and I have to credit this assertion to David F. Noble, who's ideas are primarily reflected here): The technology is neutral - if you don't like what's being done with it, that's entirely the fault of the people using it. To the extent that spreadsheets have had a deleterious effect on our society, that is because powerful individuals saw an opportunity in the technology and exploited it. In a different instutitional structure or with different power relations already in place, the effects would be totally different.
In closing, if anyone actually cares about the future of engineering professions, read Forces of Production by David F. Noble. De-skilled assembly line jobs became the norm not because it was a better way of doing things and not through any inherent properties of machine tools (let alone "market pressure"), but because it served the economic and political interests of the managerical class. Spreadsheets are (relatively minor) among the many tools that the current generation of management tries to use to do the same to engineers today.
I Still Remember (Score:3, Interesting)
My life as an accountant changed forever that day. It took me three months, and many meeting to get the local division of the major oil company I worked for to buy the first Apple in the company. I seem to recall it was $3,000 or so.
After that it was models and spreadsheets galore - well, whatever was 64k or smaller. ("Oh, you want that calculation? Well, I'll have to remove this one.)
Then on to IBM and Lotus 123 and the rest is history.
Were spreadsheets abused, full of errors, and assumed gospel when they weren't? Sure, but they also made clear, in a very short time, important data used to make good decisions.
Now long retired, that moment in SF remains with me as one of those "ah yes" moments in life.