British spreadsheet wizard will take mad skillz to Vegas after taking national Excel crown The inaugural finals of the UK Excel Championship have come and gone, and there is now one spreadsheet wrangler to rule them all, at least in the United Kingdom. The winner is self-proclaimed Excel enthusiast Ha Dang, a qualified accountant working for a Leeds-based company. The victory means that Dang will head to Las Vegas …
It's been a personal mantra for the past decade as pretty much every time I've come to a janky Excel implementation there was a more elegant solution that was easier to do, quicker and more maintainable.
Yet, just the other week, I created a spreadsheet that fetched data from two different systems (one client, one internal) and then sent SMS reminders to employees, so, you know, don't follow my own advice much...
You may joke but I lived through the arrival of Visicalc which started the whole spreadsheet mania.
Even now spreadsheets are vital to office workers everywhere.
Misused & abused even under threats to 'Use the right thing for the job' BUT it still persists !!!
Spreadsheets as databases etc ... Arrrgh !!!
Spreadsheets will never go and will ALWAYS be used for 'everything' ... people just think in terms of tables or columns (Accounting anyone !!!)
Well done to Ha Dang ... best of luck in the finals.
:)
Have an upvote, I too remember VisiCalc, and later Multiplan and Lotus 1-2-3. Unfortunately Lotus was advertised with 3 functions (hence its name): Spreadsheet, database, and charts. Probably responsible for your 'Spreadsheets as databases etc ... Arrrgh !!!’. I had a colleague who used it to write letters and reports too (they definitely were not pretty). Around the time that Lotus was becoming popular, I was using DEC Rdb and R:Base, so I was "not impressed" by the spreadsheet-as-a-database; but almost everyone where I worked had one...
The cool thing about Wingz is that it was written in itself.
As in, its own macro language is what they used to write the spreadsheet. And it was all Live code: you could modify it on the fly.
It was kinda like a spreadsheet version of Emacs or a LISP Machine.
I'm pretty sure emacs has a core wrtten in C. Writing a spreadsheet in a macro language in the 90's would have been a brave decision.
The one I was exposed to was TIBWingz which billed itself as a "real-time" spreadsheet. Not really in the comp-sci sense but in the sense that events on the "Teknekron Information Bus", i.e. stock prices flying across the network, could be inputs to formulae and would trigger recalculation. I recall it not being terribly performant...
As a Chemistry undergraduate in 1992 we were told to use QuattroPro because it had a much much better graphing and statistical functions (which it did - including various standard deviation calculation methods).
By the time I was doing my Ph.D. QuattroPro was gone, but people still hated Excel's graphs, so used it as a spreadsheet in tandem with SigmaPlot for graphing.
We had an artist family friend who used to write letters in Adobe Illustrator.
Definitely not a dullard or entirely computer illiterate1 but he had is quirks.
_________________
1 When he was in his 70s he taught himself Adobe Photoshop and Illustrator well enough to produce some pretty impressive art which, because of his reputation as a painter and (small 'i') illustrator, sold for considerable sums.
This proves the point that the 'tool' is not important the 'Artisan' IS !!!
A true 'Artisan' can make 'Works of Art' out of what ever is too hand !!!
This applies to any creative person including Devs etc who work in IT !!!
Creative is Creative ... period
Typical example of missing this point is in Photography where too many 'newbies' think they need to buy the latest/greatest and most expensive kit !!!
The true artist can produce 'Better' Photographs with any decent kit that meets his needs including 20-30-40 year old non-digital cameras or even 10 years old digital that you can get for a few hundred Pounds/Dollars.
The art/skill is within you and the kit is just a tool ... like a pencil !!!
:)
I started messing with spreadsheets with Visicalc on a TRS80. Moved through various other products. The ultimate spreadsheet was Lotus Improv. Was so impressed with it I bought my own copy. Unfortunately, it was a bit too different and Lotus didn't want it to cannibalise 123 sales. If they had the faith to stick with it, they might not have ended up losing out to Excel.
My first full encounter with a spreadsheet was with my first work PC, Lotus Symphony on a Dec Rainbow, back in 1985. Lotus took the three-function model further with Symphony than 1-2-3.
I ran a multi-million dollar inspection programme on a single table - different forms for entering tasks, results, outstanding work and, even, pro-forma letters populated with the necessary instructions for sub-contractors. I hadn't set it up but, in order to take on the responsibility for using (and maintaining) it my employer sent me on the only formal computer course I've had since my first stint at university (where the was a one-semester module teaching Fortran-IV) - three days on Lotus Symphony. Those two courses gave me the foundation I needed to learn what I needed as my career progressed, on-the-job.
I should add that I never held a job with any formal IT role, but the ability to use the tools IT made available was invaluable (and I was often drafted in for intra-departmental support, as it kept the dreaded hell-desk folk at arms length)!
I wrote a couple of systems that were deployed on multiple LANs. Typically 5-30 concurrent users. They started to run out of steam at about 30,000 rows in a table with lots of inserts/updates - my solution was to archive data off to other tables/dbs and just have recent/important stuff active in the main tables. As expected, the performance improved if the client PCs had their own copies of the engine, and shared just the data files on the server.
In the UK/Europe it was sold under licence by Microsoft and was, I believe, the 2nd most popular db after dBase - which I never liked much, possibly because I was used to SQL. I think it stopped being a major player after Wayne Erickson put a lot of company resources into the DEC version (Vanguard?) just as DEC had their own problems. I heard that a number of the coders then went to work for Microsoft to build MS Access. I still have the manuals and floppies for R:Base 4.5, but never went to the Windows version, as I had moved on. Current versions are available for Windows (and DOS!).
"Have an upvote, I too remember VisiCalc, and later Multiplan and Lotus 1-2-3"
I couldn't afford to buy 1-2-3 when I was a student at Uni...but at the time, long before the internet, there were lots of palces one could get hold of shareware and amongst all the various programs once could get hold of, was "As-Easy-As" which was essentially a 1-2-3-clone, that read and wrote both WKS and WK1 files.
And later on (late 1980s), there was SmartWare, which was an "Office" type DOS program, consisting of word processor, database and spreadsheet, and the latter had some functions that even the latest version of Excel does not have !!
I remember trying to Launch VisiCalc on an Apricot Portable, in a very busy exhibition centre a LOT of years ago. Saying "VisiCalc" into a microphone that was part of the Apricot was a nightmare.
Fast Forward to 2025, and trying to get Siri to do what it's asked in a car is very similar......
When the very first spreadsheet came out, traders in the big banks were buying Apple II's (the only thing it ran on) just to be able to run it. And would pick the Apple II up and carry it around under their arm so no one could cadge their work.
I was called in to troubleshoot a problem with an Excel application a project team had put together. I thought "Hmm, Excel application", now there's a clue. The problem was excessive time for calculation. They had some bright people on the team and some good suggestions on causes but lacked the depth to solve the issue.
The Excel instance they were using was part of an image I had built for these project teams I supported. It was built on the "spare" machine I had available. This was in the WinXP epoch around the single/dual core era. Of course, my "spare" was a single core, whilst their production machine was dual core. They also had single core machines on their isolated network for light weight supporting work so they demonstrated that the time difference to complete the calculations hardly varied between the dual and a single core machine machine otherwise similar in specs. I dug around and found an option in Excel (or was it XP) which referenced single vs. dual core enable/disable (mostly paraphrased due to old non-ECC bio-memory). I enabled the use of both cores and voila calculations were being completed in nearly half the time! Brilliant! or not . . .
The team was not anywhere complete in building their Excel application. These relatively bright people realized that they needed to solve the calculation time problem before moving on to scale. So, after beginning to scale, they stated the calculations were still taking too long. Now, their proposed solution was an uberworkstation 64-bit multi-core (quad maybe) XEON running XP-64. The hardware provided to these teams were supplied out of my department's budget, unless the required hardware was specialized. I had to explain to both department heads that in-spite of all the similar sounding names that their request was specialized hardware with the final fact being the price of only 25 times the top machine on our pre-approved list. I had to also explain that even the proposed workstation would not solve the issue because Excel was unable to take full advantage of all the extras the hardware had to offer. Then I had to explain why Excel was not the tool they needed for the task.
The formulas were long, complex, interdependent (but not circular) and not optimized. Any improvement in calculations were due to more cores were probably already realized at dual core with diminishing returns if Excel could handle more cores. I suggested optimizing the formulas and limiting interdependence to only what was required. I knew of another independent team's uberworkstation they might be able to borrow some time to test pre- and post- optimization (and prove my "theory"). I then suggested if their expected improvements were not realized, they should develop their application in a programming language such as FORTRAN or other mathematically oriented language.
The project wound down shortly thereafter and the grand Excel application never came to be.
True, but that's not very different from most other options that would be considered. A custom application with plenty of logging and code in version control is much better, but basically nobody is considering both options. Those who have available programmers that can be assigned to build that aren't going to decide not to bother using them, and people without them aren't going to hire that team and do the setup when Excel seems to get the job done. These are generally two different types of team or organization, and the one choosing Excel doesn't have a similarly convenient way to do it. Those who understand the problem may try to set up change tracking so that they can audit any changes made, but that isn't simple and it has some gaps.
They won by 0.293% (according to Excel), shirely they should be using that rather than "by the skin of my teeth". If this guy doesn't write emails using VLOOKUP and a dictionary then frankly, standards have slipped. Excel master indeed!
I jest of course, well done.
Sometime before spreadsheets...... As a lowly trainee on the circuit and in Sales, I was given the deadly-dull task of calculating the selling price of a large number of spares consisting of major sub-assemblies. A lot of items with different markups and numbers-off. The total, by my calculation was over £1 million (a long time ago....).
In sheer disbelief, the boss asked how I had reached that number: Paper, pencil, arithmetic and slide-rule. How they laughed: "A slide rule!". Not accurate enough..... They expected the sum to be about £10 k. After a few hours of checking, using their state-of-the-art electronic desk-calculator, they confirmed my calculation and took the job off my hands. I wasn't going to get credit for £1 million spares order.
I remember a half-page article in the FT about a study done by one of the big consulting agencies (Arthur Anderson?) that reckoned about 50% of spreadsheets in corporate use contained errors / bugs.
A problem with spreadsheets that is different from programming languages is that a cell or region can contain architecturally incorrect formulae but still display 'correct' results. E.g. incorrectly anchored formulae. Such are invisible future errors or bugs waiting to happen, perhaps when cells are copied or moved or rows / columns inserted. People without sound programming backgrounds are easily fooled into believing that if the result looks correct the underlying formulae are correct. With programming there would be many compilation or execution errors / error messages in many such situations.
> That requires an even newer version than XLOOKUP.
Within the rules [ https://excel-esports.com/rules/ ] is this important detail:
”The Cases will be structured in such a way that the users of older versions of MS Excel (2010 and up) should be able to complete them. However, newer versions of MS Excel might provide faster ways to solve the Cases.”
It would thus be interesting to know which versions of Excel were used by Lorenzo Foti and Ha Dang.
WRT XLookup, I prefer VLookup and HLookup, as the “V” and “H” visually flag to the reader what the writer intended. “X” is also confusing given the XYZ co-ordinates (although even the latest version of Excel 365 is still p*ss poor at anything in the Z dimension.
WRT XLookup, I prefer VLookup and HLookup, as the “V” and “H” visually flag to the reader what the writer intended. “X” is also confusing given the XYZ co-ordinates (although even the latest version of Excel 365 is still p*ss poor at anything in the Z dimension.
For anything with more than two dimensions or more than a few tens of rows and columns, I start reaching for Numpy. If things then start getting out of hand you can add Pandas and Dask and then Matplotlib to visualize. Onc you're in deep, Scipy & Sympy let you go completely Ramanujan. :-D YMMV
... that looks at the complexity of an Excel workbook in terms of average formula complexity and gives it a toxicity rating on a scale of “nauseating” to “nuke every city containing a datacenter where this abombination has been stored in or replicated to?’
" XLOOKUP is Microsoft's preferred option, although a recent version of Excel is required." It is not just a newer version of Excel that is required, with some devices Microsoft actually disable XLOOKUP depending on license.
Last time I looked XLOOKUP is available on all devices LibreOffice supports and also Collabora Online, I believe this includes all mobile devices and the Chromebook offline app as well irrespective of licensing.
It gets worse…
This article gives an overview of which versions ship with XLoopup support:
https://www.thespreadsheetguru.com/xlookup-excel-version/
However, buried in this article (https://exceljet.net/functions/xlookup-function - see example 10), is a note that post December 2024 the Excel 365 function gained additional functionality; which I expect is missing from Excel 2024…
This single function giving a very good example of just how much MS is unable to maintain product compatibility and likes to mess with their users. It would be nice if you could set Excel (or LibreOffice) to a specific compatibility mode so that only functions (and their parameters) supported by that version were available to be used. Microsoft’s Save As/Export and its “possible loss of formats and functionality” message is not really good enough as it doesn’t prevent the unwitting production of product version specific spreadsheets.
Aside: This website has VBA code for an XLoopup function and 7 other Excel 365 functions for Excel 2007 and later.
https://hermann-baum.de/excel/hbSort/en/xlookup.php
None of it is adding up. It’s still a mystery
Disappointingly the website doesn’t just give you a spreadsheet to download or view. I’d have thought that if they were taking it seriously they would have rendered the site’s content only as spreadsheets, instead of this html malarkey… As in you know those websites that are nowt but PDFs? Could do the same with Excel…
This actually happened to me in a job where I lasted a whole 3 months.
I was asked to carry out some relatively complex calculations on some (five in total) high vacuum distillation units which the company used. As the calculation was the same with only the numbers different, I did it on a spreadsheet (Borland QuattroPro IIRC). Having got the first calculation set up and checked that it was correct I then used the same spreadsheet for the next four units. Needless to say that it saved me a great deal of time and effort.
My line manager, who I learned had been off work for several months with stress, then checked all of my calculations separately by printing them out and going through all five individually with his pocket calculator. Needless to say that all were correct, but what had taken me a day and a half to do took him a full three days to check.
And that's one of the main reasons that I left.
On a slightly different note, one of my later companies used Lotus Symphony for DOS which was a combined word processor / spreadsheet / database, which I found very easy to use (within the limitations of DOS). The database was more of a card index than a proper relational database, but it has it's uses.
Foti got to within seven points of Dang before the latter pulled away, ending 11 points ahead. The final score was Ha Dang with 2,474 points and Lorenzo Foti with 2,463.
And they probably kept track of the scores in an ODS file as well as an XLSX one just in case the Excel guy had cocked up the SUM() ranges.
Rule 0: Solutions must be compatible with, generate the correct results and produce identical printout to a Hewlett Packard Color LaserJet MFP M283fdw on all in-support releases of Excel for Windows & Excel for Mac.
That'll larn 'em!
Anyway, where is the AcornSoft ViewSheet World Cup Final being held this year. (See icon.)
This is exactly a KRAZAM sketch from 6 years ago, down to the vlookup vs xlookup. Is someone pulling my leg? Is this an out of season april fools joke? "MAKRO | Microsoft excel stream highlights 3/19" & "MAKRO | XLOOKUP: META BREAKER? | Microsoft excel stream highlights 10/19" are the sketches in question.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
