His dock in this video is a trip down memory lane: Sublime Text, Textwrangler, Skype, Pixelmator... takes me back.
Also I didn't watch the whole video but if he made it through 54 minutes of excel for mac without it randomly crashing, that alone would be an impressive feat.
While the ad isn’t that impressive, I’m mainly just impressed by how much of the important functionality of Excel was already figured out by 1992. I expected formulas and at least some decent formatting and editing tools, but I was not expecting interpolation, what appears to be the most important custom number formatting features still today, and the automatic table styling. That all already existed and was in an advertisement before I was even born.
Pretty cool.
(I know a lot of the innovation came earlier with software like VisiCalc, but that’s distant enough to be hard for me to appreciate.)
I still didn't know you could select a bunch of cells and drag and drop them as a single unit. I've always just used cut and paste. 40 years later, better late than never!
Now, if someone can explain to me why, in 2022, the Online/365 version of Microsoft Word cannot insert a simple horizontal rule...
The generation before yours is lacking neither intellect nor creativity. We had Calculus in school and the *real* LSD back then.
It's also worth noting that the desired feature set of a spreadsheet is rather universal. Like a paint program, or a music player. Spreadsheets go thousands of years back in human history, and are very well understood as a means and notation for expressing data and expressing calculations over collections. It's not at all a mystery what features will do well in a spreadsheet program.
I've never worked on a spreadsheet program from that side but I have to imagine that the operational abstract of the team centering around how to educate users about how to use the various features they painstakingly develop. It's well known that spreadsheets are difficult to program, because users will inevitable collect larger and larger amounts of data in them, and it needs to do large amounts of calculations very quickly. So right off the bat you already know that a ton of man-hours are sunk into each feature overall, but then you need to really nail down a rock-solid visual grammar and teach the user some sort of intuition for finding these functions as quickly and consistently as possible, lest they get frustrated and leave.
Anyhow. I always find it laughably rude, yet predictable, reading these sorts of comments. As if anybody born before 1980 could have possibly understood the deep enigma of other human beings using computers.
I’ve tried writing code on a 486dx, which I believe is a contemporary CPU for that era. It’s not necessarily that nobody could come up with anything like this, it’s more impressive to me that it was basically already there for a lot of these things, in an environment that is, well, to be honest, quite difficult to eek good performance out of.
(And of course, heavily limited in memory. With toolchains and development environments that are constrained similarly.)
I don’t think you’re giving a charitable interpretation of the parent - I interpreted the comment as saying they’re impressed that the technology was impressive for the time rather than being critical of the generation
I know it’s not spreadsheets and even more distant, but I think it’s worth it for you to check [1] to see the origin of much of HID to this date. Best in mind that all of what you see there was done by 12 people from scratch (no OS, not even assembled hardware) in roughly 2 years until 1968.
The advantage Excel had in 1992 was it had already existed on the Mac for 7 years and had been highly successful as a Mac application. Conversely the early Windows versions had, like Windows before 3.0, not really got any traction. This meant that when Windows really took off in the early 90s, Excel was already a quite mature GUI app.
Did anybody else have an issue with "She wants to see a 10% growth?", and then Excel autofills the following quarters linearly (+$100 / $200 / $300) not exponentially (+10%)?
Never extrapolate your cells from just 2 fields folks!
You'd be surprised how often finance people make this first order Taylor series approximation.
And to be honest, the inputs they have to deal with are uncertain enough that in the end, pretending small-percent geometric growth is linear is not going to be the dominating term of the error.
In this case... meh. Actual geometric model gives a sum of 5.11 for the year, whereas linear approximation is 5.00 -- an underestimation of just over 2 %.
Do you really think the 10 % assumption we went in with is accurate to ± 0.2 percentage points?
The premise of the ad seems quite funny to me. It’s like they wanted to include the thing figuring out the patterns (I was a little impressed to see Q1, Q2, etc, but it would be nice to see it roll over to the next year as Q1 not Q5) but they somehow ended up with the excel-using businessman seeming to make up(?) some numbers for an important report and then project them linearly which doesn’t feel like the best message ever. Though they probably didn’t have enough time for someone to do a better job of modelling (if someone were doing it very quickly they probably couldn’t explain or show off drag and drop or toolbar buttons).
Back then their competition was Lotus 1-2-3. What they were showing was that they had some basic automation for some common tasks (autofill, auto calculations, auto formatting) while everyone who was coming from Lotus, their target customer, knew how manual and tedious this kind of thing was to do. Plus the UI in Lotus was only slightly less convoluted than Emacs.
The part they should have shown was how you could have large, complex (for the time) spreadsheets and not run out of memory which was a significant advantage they had back then as DOS applications had serious memory constraints even with extended memory. A lot of heavy spreadsheet users switched to Excel for that feature alone.
I was referring to the DOS version which is what the majority of 1-2-3 users ever experienced. While the Windows version existed, it took far too long[1] to be released and was quite buggy. By the time Lotus had even a remotely viable Windows product, they were in 3rd place and had mostly bled out their users.
[1] Windows 3.0 came out in mid-1990 and 1-2-3 for Windows came out at the end of 1991. The options to run 1-2-3 for DOS under Windows until then were comically non-viable. When 1-2-3 for Windows came out, it was practically unusable for those who waited and no longer mattered for those who had already switched away.
I coincidentally watched this video yesterday for the first time because it showed up in a YouTube recommendation. It got me thinking about how old videos like this go viral on YouTube. Maybe a lot of people started commenting on it, which boosts its ranking and visibility in recommendations, and it snowballs from there?
Bill Gates gave the world Excel, Word, Powerpoint, etc, so the humanity could become creative, educated and thriving. Zuckerberg gave the world social media and the world became ...
> Bill Gates gave the world Excel, Word, Powerpoint, etc, so the humanity could become creative, educated and thriving. Zuckerberg gave the world social media and the world became ...
I know it's a joke, but positioning Bill Gates, especially early-'90s-era Bill Gates, as some sort of savior doesn't work for me.
Now, Knuth gave the world TeX, and I'm all on board for praising that contribution to humanity.
Hmm I get the point, still PowerPointing is always use as a term for “not really doing any work”. I guess it is because it always to try to explain to our businesses what could work and you have to do a lot before something catches on. Yeah I know, little vision in our business.
Powerpoint was actually unique from the others in that it was an acquisition from a company called Forethought. It had a single release in 1987 before MSFT bought it and was for a very long time the only part of Office not built in Redmond (the purchase promised to keep the eng team in California within the then-newly created Graphics Business Unit).
> this might be as good as business projections get
It almost seems like weather reports. Anyone can make a good short term forecast by looking out the window, but the amount of complexity, uncertainty and variables going into a long term forecast makes them almost pointless. Which is where Excel comes in I guess :-)
It quietly became a thing you could open a spreadsheet from a network share, and so could someone else, and both live-edit it in realtime with markers for where each other is. Not just two people, either. Then gained that with SharePoint cloud backing which means the people don't need to be on the same LAN. And gained a web version which can be scripted in JavaScript instead of Visual Basic for Applications, and so desktop and web users can open, view, edit the same spreadsheet at the same time. With the Microsoft365 backing, someone can @ you in an Excel cell and M365 will email you about it to alert you. Spreadsheets are accessible from the MS Graph API and with a REST interface[1] which means they can be generated and updated from code while on M365, it means Excel can be used as a backend calculation engine in a supported way now (instead of running it headless on a server in an unsupported way), and spreadsheets integrate with PowerAutomate[2].
As I've said several times on HN, Microsoft has quietly made Office Cloud a company wide multiplayer gameworld thing, and hardly anyone seems to have noticed this new category appearing. It's not "store your files in the cloud" or "spreadsheets in a browser", it's something much more deeply integrated. If you thought it was hard to get away from Microsoft Office before because someone might send you a spreadsheet with weird formatting, this is the kind of moat/lockin which will carry Microsoft another decade and over Chromebooks and Android and macOS and Linux desktops.
I don't see much value in any of this, nor in any of the "improvements" occurring in the last 10 years.
Still Word is crashing regularly when it reaches 100+ pages, and bullets and alike are acting weird. And if the dreaded document corruption occurs, you are good to copy paste the whole in a new blank document.
That people want to pay 100€/yr min for this is beyond me.
Perhaps you don't but these are critical features in many businesses. The company I work for does government contracts where we have 3000 line cost models that are shared by 6+ organisations and maintained by 20+ people with daily updates. Without having them accessible by everyone at once, people would be creating daily copies which would need to get reconciled.
it still has locks thou - unless you use the less powerful online version. Would be better imho to have a lock free desktop version of Excel that operates like an IDE where you merge changes and resolve conflicts at the point of merge.
But also many of the core features are still broken. The fx formula bar button still opens the 1990s non resizable window with a search function that doesn't do anything, and you have to go through a list of thousands of forumula in a tiny box. They never fixed the bug that corrupts xlsm files (I recommend people to never use xlsm for that reason). They also managed to break things, like it has become almost impossible to copy a chart or table from excel to powerpoint without the whole formating being lost (the biggest pain being red negative numbers).
Yeah, there has been minor improvements. Ditching the 65k row limit, waterfall charts, xlookup (though 20 years too late, people had index/match all over their spreadsheets from day 1), 64bit version of excel. The more recent version automatically spills formula returning arrays over multiple cells, useful for users who can't do array formulas.
But many new features I wouldn't touch with a stick. The BI feature that downloads whole table from a database in memory even if you only need one day and there are billions of rows. I always found tables pretty much useless. In term of UI, they dramatically increased the number of clicks to get to a button (try inserting many shapes in powerpoint or excel, or changing many chart/shape formats, things are deeply nested). If you look at the default formats (tables, charts) have been designed by people who never looked at a corporate presentation.
And there are udf I created that I don't understand why aren't part of the default feature set. A formula to deduplicate a list, or split a text based on delimiters, or to execute a sql query against a database and return an array (I know there are buttons for that, but a button requires manual work every time you need to refresh the data, vs a shift-F9 when it is formula based). And why is it still so hard to refresh a ppt presentation based on numbers coming from spreadsheets. It is one of the most common and basic use.
It’s a powerful platform for sure. But having just been in a scenario where 8 people tried to work on a moderately complex sheet simultaneously, the collaboration aspects still suck - Google still owns that.
That is a fantastically perceptive and compact analysis. Do you have a blog or anything? Tried to stalk your username I think maybe you’re using it might be a play on a Doctor Who actor or something.
Thank you; I don't have a blog, just use HN and Reddit as sinks for opinions. My username is a wordplay on a radio telescope near Manchester, England, which I visited a few times as a kid when they had science exhibits on the grounds https://en.wikipedia.org/wiki/Jodrell_Bank_Observatory
Lotus tried to re-think spreadsheets with a product called Improv - it was full of good ideas but the product never took off. The product launched on NeXTSTEP in 1991 and then on Windows in 1993.
Here's a 1990 sales video (low-resolution) from Lotus explaining how it works, and how it differs from traditional spreadsheets: https://www.youtube.com/watch?v=rgGmKD87U3M
Have had the chance of working 18 month with NeXT computer - best (spirit) of machine ever... Improv was outstanding, but too much so... people couldn't easily adapt their spreadsheet work to the new tools (and on NeXT, in a time when green screen PCs here trying to adopt Windows 3.11 ;-) ) - seen quite a few applications never used to their potential benefits (and just wither away, replaced by less capable partial imitations)
gotta say I like the search thingy they put in there where you can just type what you want instead of digging through menus, especially for lesser used features.
Basically the idea of a command palette, although less prominent, I guess. And yes, it's the way to actually use the office suite in general instead of messing around with the ribbon.
Unless that was sarcasm and I'm whooshing, I really don't get the big deal about xlookup. It's like vlookup but I'm no longer forced to have the lookup column as the first column in the range? I mean, that's nice, it makes my workflows one or steps shorter, but it's hardly groundbreaking.
There have been a fair few changes that make a large difference if you care about that thing, but nothing earth shattering for all. Being able too define an arbitrary number of dynamic forgetting rules (where there was a fairly strict limit previously) is one change that I really appreciate. I'm sure there are many other such changes that appeal to other people's user of the thing.
