Dan Bricklin's Ted Talk on spreadsheet

[shrivallabha]

If you search following term on Google:
"Who created Excel" or "Who invented Excel" then Google promptly returns Dan Bricklin's name and couple of his younger day photos.

In reality, Dan Bricklin never worked for Microsoft. But he did create the electronic spreadsheet before Microsoft Excel came into being. He and his partner Bob Frankston created VisiCalc which used to get shipped with Apple Macs. Dan was a student then at Harvard University when he took his problem of reworking case study calculations seriously.

In following Ted Talk, he talks about it.

 

[Marc L]

3 years after, the first I ever worked with was Multiplan on PC, before Lotus 1-2-3 …​

​

 

[Somendra Misra]

I started my journey with Lotus 1-2-3 (creating a graph in it was supercool in those days). Now all that seems so easy with Power Bi and other tools.

dBase was also the same time.... it was fun to work on those 80286 system with 2Mb RAM, 40MB Harddisk.....

 

[AlanSidman]

I, too started with 1-2-3 in 1981. I was not aware of Visicalc as I was introduced to spreadsheets on an IBM DOS computer.

 

[shrivallabha]

That kind of makes me feel quite young, I began with Excel 2003 and it wasn't love at first sight. But as it often happens, it was popular choice in office computers and therefore I kept using it.

I realized Excel's power when I could write VBA program to send emails from IBM's another product, Lotus Notes in 2012 and since then I have been a steady user. This thread on VBAExpress documents those unsteady steps
http://www.vbaexpress.com/forum/sho...sing-Excel-To-Send-Emails-Through-Lotus-Notes

 

[Marc L]

The spreadsheets veterans … ​

​

 

[Peter Bartholomew]

@shrivallabha
Thanks for posting the link to Dan's TED talk. The presentation had a major impact upon my thinking when it came out, but not necessarily in a way that might have been intended. At the time, I locked in a prolonged discussion with proponents of the FAST standard (no array formulas, no names !) in which I sought to clarify why financial modelling was so different that the well-established practices of mathematical calculation and computer programming could not possibly apply.
Home | The FAST Standard Organisation (fast-standard.org)

The moment of revelation came about 7 minutes into the presentation when Dan considered the problem of identifying the operands within a calculation to the user: "It could be done the programmer's way [using named variables] but that would be too tedious". He considers the use of sequential numbering, such as is used for sheets, shapes, tables, etc. but decided it would be difficult to track the calculation, and only then settles on a town-plan style of grid system.

From that, I learnt that what many would consider the defining characteristic of spreadsheets was only adopted in support of idleness. Of course it is not necessary to define your terms, that would be tedious, [and others might be able to read the formula and understand their intent?]. That was the point at which I committed to never using a direct cell reference again and using array formulae wherever possible.

On the nostalgia bit, I largely missed the early days of the spreadsheet. Even later I tended to use DecWrite on a VAX in preference to WordPerfect on a pc for report writing and my computation was all Fortran IV.

 

[shrivallabha]

@Peter Bartholomew
Your first and last few lines have totally evaded me Have no idea what FAST is and after reading word Finance somewhere I backed off. Just about the same for "DecWrite on a VAX".

But about the named references and self referencing formulas, I will share my experience. In my office, there's a guy who is decent in Excel formulas and picks them up pretty fast. One day, I created a fairly simple table and then wrote formulas using the table structure (based on our previous discussion on this forum). The formulas by themselves were pretty simple but the user response was totally unexpected! He said he preferred cell reference formulas and the table confused him a lot. I tried to explain but he did not budge. So we re-did it the "normal" way. Individually, I do not have preference or a stated inclination. However, my default mode is no named ranges but I use them as I go along to have cross-references, data validations etc.

Testing anything on a sample population of one and drawing conclusions would possibly amount to a statistical crime but what I'd state is for any uninitiated mind (on either side of the logic that is), the other side would appear to be confusing and each side may find enough supporters with their own valid reasons.

 

[Hui]

VisiCalc (for "visible calculator") was the first spreadsheet computer program for personal computers, originally released for Apple II by VisiCorp in 1979.

Damn I remember the first time I used that on an Apple II, it was in 1980 at High School

 

[Marc L]

The same on Apple II, I did also some 'Calc' on a Commodore 64 but I can't remember what was exactly its name …​

​

So it was the second veterans wave ! ​

​

 

[Peter Bartholomew]

@shrivallabha
You may have enjoyed the FAST reference had you chosen to open it (like you I do not have a finance background so I, too, find much of the terminology pretty impenetrable). I suspect the overwhelming majority of spreadsheet users would go along with your colleague. The mind-set seems to be "I know what I want -- and I want what I know"! The catch is that the formulas created by 'click-and-pray' are no better than the VBA code generated by Macro Recorder. If you expect to find the equivalent of Class modules, event handlers and centralised error reporting, you are out of luck!

As a result, it should be a matter of little surprise that audits have found that fewer than 10% of spreadsheets are actually correct in the sense that decisions based upon them are likely to be right. That said, the traditional spreadsheet does provide communication between the author and the user, even if the language is little better than 'point and grunt'. Programmers may despise spreadsheets, but spreadsheets owe their success to the interactive style with its underpinning of 'tips and tricks'.

As for nostalgia corner, my first home computer was a BBC Model B. I believe it was RISC architecture and the whole of the OS was held on ROM (so computer viruses were not possible!) The assumption was that all users were dying to learn Basic; somewhat questionable.

 

[shrivallabha]

@Peter Bartholomew I thought FAST was purely finance related standard / document having nothing to do with spreadsheets but it is not. Will give it a read.