An Ode to Excel: 34 Years of Magic | Hacker News Reader

An Ode to Excel: 34 Years of Magic

blog.stephsmith.io

49 pointszdw21 comments

Excel is great, it is still Microsoft’s best product, (in contrast to Word which is an abomination). However I am very disappointed with Microsoft’s lack of investment in Excel, over the past 10 years it has hardly improved. eg the VBA code editor is appalling, like a relic from the 1980’s (constrast VS Code). I use tables extensively (they are quite nice) problem is they require extensive use of Index/Match which becomes tedious to set up and is very slow, so manual update soon becomes obligatory. MS should incorporate some of the Power BI linking (think Foreign Keys) capability to overcome this big problem. Would be nice if they added SQL,instead of always making up their own query language language (eg M). Sumifs Countifs functions offer powerful and essential capability that could also be better done with SQL syntax. Excel is very error prone so Unique row constraints on tables or sheets would be handy. The formula editor is also archaic and could be easily improved dramatically. Yes - Excel is Fun - what an awesome YouTube channel, that guy’s enthusiasm for Excel is extraordinary.

Yes, agreed that there's lots of room for improvement. It's actually quite amazing that the core functionality has enabled it to stick for so long.

Re: Excel is Fun, found the channel while researching for the piece and it is truly a hidden Internet gem. We can only hope more people find their passion as this guy has.

It's not as poweful as SQL, but the new array functions bring some fresh air

https://blog-insider.office.com/2019/06/13/dynamic-arrays-an...

MS is conservative in general; double so with Office; triple so with their best-ever app; quadruple-so at a time when they were basically busy rewriting it to be browser-based.

Why do you think Word is an abomination? Among all the word processors, I find it to be the best.

I had so much fun writing this so it was super nice to run into it while browsing HN.

I think there's a lot we can all learn from the history of Excel (arguably the most bulletproof software of the past few decades), particularly in the age of crazy valuations and unicorns.

Curious to know what more recent software people think will hold up to the test of time, like Excel has?

Good question!

I'd nominate: Postgres

Postgres isn't recent at all.

I would say HashiCorp Terraform and Vault

Fair enough. I didn't catch the recent requirement. Well Postgres keeps on evolving with recent features, every major version brings something new and great.

No disagreementa with nominating Terraform :).

Excel is very much a consumer product. In that category, I think the big one is Firefox: it’s the oldest browser out there. Explorer is being phased out, Chrome is newer, and everything else is a microscopic niche. The web needs a Firefox to balance out the giant of the day, and Mozilla will likely cover that role for the foreseeable future.

A lot of the backwards compatability is jaw-dropping. Some things that were intended to be deprecated what feels like forever ago still work. [1] I also regularly uncover brilliant bits of spreadsheet design looking at documents made one or more decades ago. Even some old school tricks in those sheets still help me do new things. For example the (internet explorer exclusive) right click to export to excel is still among the most reliable ways to extract data that otherwise won't play nice.

However I wish speed was higher on the priority list. For big sheets with lots of calculations Excel 2016 was SLOW. Slower than 2013 by a fair margin, and actually slower than binary based sheets on 2003 by A LOT. [2] [3] [4] Excel was my default calculator on my computer for years. But it just starts too slow now. Foreign language support is weird in that regard. [5]. And on a personal note, I'm still bitter about the ribbon.

[1] Though some dissapear, you will be kinda missed, default Powerview

[2] https://social.technet.microsoft.com/Forums/en-US/8b2a498c-a...

[3] http://dailydoseofexcel.com/archives/2015/04/07/excel-2013-v...

[4] My own experience, for what it's worth.

[5] https://professor-excel.com/performance-excel-study/

Microsoft did 3 very smart things that elevated Excel from 0% to 100% market share:

1) Early versions were (or felt) Windows-native. Due to the Mac roots, Excel felt integrated with Windows 3 and GUI right when people and companies were switching from MS-DOS to Windows. Lotus 123 never really got that feel right, its early Windows versions just seemed inferior to the old DOS versions, it just felt out of place on Windows.

2) 1-2-3 compatible keystrokes. Made it easy for people to switch.

3) Bundling with Office. In the 80s you bought WordPerfect, and 1-2-3, and maybe FoxPro or something for simple databases. Office gave you everything in one inexpensive box. In fact, Word vs WordPerfect was even a bigger contrast than Excel vs 1-2-3, since Word had a WYSIWYG interface that simply blew WordPerfect out of the water.

I’ve always despised spreadsheets... the confusion between input data, methods of calculation, and output is absolutely foul.

I hung on to Lotus Improv as long as I could, and then opted for Quantrix Modeller, and now do most of my stuff in a combination of locally hosted SQL and Mathematica.

What kind of broken paradigm allows for an errant keystroke to break the flow of a calculation? Or allowing data to go stale, only to break the calculation when you refresh and get back more rows than you previously had?

Right tool for the job.

Sometimes it absolutely makes sense to have input/calc/output on the same "sheet" or set of "sheets."

Unfortunately, most people don't know how to set up a spreadsheet in a way that makes this work (clearly delineating the input/calc/output sections with data validation, formatting, protection etc).

Of course, if you were to make a solution that "forced" these choices on the user they'll cry about losing their freedom.

So we are stuck with spreadsheets in many cases.

And then there are the cases where using a spreadsheet makes no sense. I'd say it's about 50/50: Half the stuff currently done in spreadsheets makes sense, half of it would be better done elsewhere.

I've managed to move more than half of our previous spreadsheet workload out of spreadsheets and into more appropriate software systems.

One question - wasn't Lotus-123 the first spreadsheet application or was it made at same time?

Edit: I should the article more carefully.

The chronological order was VisiCalc, then Lotus123, then Excel. Microsoft had a spreadsheet tool called Multiplan during the era of Lotus123, but it didn't get much traction.

An excellent article you wrote. Great with all the details.

Thank you!

Lotus123, Calcstar then Multiplan then Excel..

Only for the record, in the "passage" between Lotus 1-2-3 and Excel there was an excellent spreadsheet by Borland (that never got many traction) called "Quattro" (the Italian for 4, as if it was a continuation of the 1-2-3 series): https://en.wikipedia.org/wiki/Quattro_Pro later it became Quattro Pro and then was acquired by Corel and I believe it is very uncommon nowadays.

As a side note, VBA has been added to Excel only after version 4, before that there was only a Macro language.

If you want to feel how an ol' version of Excel behaved (without many bells and whistles and of course without the ribbon) there is an exceptionally good little program here:

http://www.byedesign.co.uk/

that has most functions and fits (the 32 bit Windows version) in a single (portable) file hust below 1,500,000 bytes.

And, JFYI, the sad part is that I wrote my fifth "I love you":

>One “I love you” for every year since I’ve fallen in love with...Microsoft Excel?

in 1994 or 1995.

Honorable mention for Logistix on Amiga, which I think came out around the same time as Excel or shortly after.

Reply on news.ycombinator.com