I just can't get me enough of Raymond Chen and his wonderful walks down the dustier paths of memory lane. Feels like a more innocent time where I didn't feel like I was imminently going to be turned into paperclips.
It’s probably partially an illusion, but while everything wasn’t fine back then, it seemed that there was at least a vision of a positive self-determined computing future that could be achieved and that we were roughly on-track on. Nowadays it feels more of a fight to keep things not getting worse, most of the time.
> ...that could be achieved and that we were roughly on-track on.
I think there's a strong lateral connection to this quote:
“I've come up with a set of rules that describe our reactions to technologies:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.”
― Douglas Adams, The Salmon of Doubt: Hitchhiking the Galaxy One Last Time
I think this has interesting implications wrt the perception of nostalgia, because nostalgia seems to be able to apply at any age to any event that happened far enough back in time; while the above theoretical model maps roughly to specific ages.
So I wonder what things are actually a partially overlapped Venn diagram of the above maxim and nostalgia.
In this case I think it's possible the idea that we were "roughly on-track on" with certain technologies - the notion of an emergent sense of structure that was certain to unfold - could map to some point in between points (1) and (2) in the maxim above. An objective analyses would instead recognize "success" as the survivorship bias that it is; but we're not objective :) - and I find that endlessly fascinating!
I once met this older gentleman at a wedding and when I asked about his life he said he used to be really into computers but the personal computing revolution made him sour. It turned professional field step by step into junk until there was only amateurs with their toy computers left.
I agree, but I also think it goes farther than computers and technology. Back then the overall zeitgeist was just a whole lot more positive. These days people are all about one-upping on the dark grittiness of music, movies, etc. But back then things were more upbeat. Yeah, we were worried about the cold war and nuclear armageddon, etc. But we still felt hopeful that we were working toward something better. I miss that.
I'd say, go back to the hardware. Pick a computer hobby that involves tinkering the hardware, not just the software. Holding something on one's hand means a lot.
Yeah, the sense at that time was that you master the machine. Now, increasingly, the machines will master you if you aren't careful. Of course, the machines haven't really done anything to us. They've just been locked down and remotely controlled to deliver ads and misinformation.
That's just past with its rose-tinted glasses. It was easy for someone to master the machine when that someone was a university researcher or a lone gamer, the most precious resource stored in the machine was saved term projects, and either it was not connected to anything else, or connected to fellow university researchers.
The stake was low, because nobody could use your computer to drain your bank account. And someone who would "prank" your computer beyond the social norm would get a stern talking to.
Computers these days have to support your grandma making hotel reservations online without her entire financial information being sent to hackers in Eastern Europe. They're doing jobs that 70s OS designers never thought about. It's a different world.
> ...you master the machine. Now, increasingly, the machines will master you...
Today I bump into limitations of machines that were put there by manufacturers who are trying to assert ownership of the device after the purchase. In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc). Today you're waging a nearly unwinnable battle against architectures of control baked-in to the hardware at a level below a level that the average end user has any hope of usurping.
The machine isn't trying to master me. The people who made the machine are. I wish people in the tech industry wouldn't be party to taking away computing freedom. It pays well, though, and they can console themselves with "It's not a computer, it's a phone"-type delusions (at least until the day "the man" comes for their PCs).
Our civilization desperately needs a way to modify modern microelectronics at home or at least in a well-equipped repair shop.
Regular people being able to commit contempt of companies' business models en masse seems to work well to keep them in check, but it's becoming ever harder with so much of everything becoming mobile-centric. And with all smartphones being locked down at the level of someone else's public keys being burned into the SoC at the factory, you can't do shit. They literally have technological supremacy over the rest of the humanity. And we're somehow okay with that.
> Our civilization desperately needs a way to modify modern microelectronics at home or at least in a well-equipped repair shop.
I'll take consumer protection regulation, at least in the short term.
I wish manufacturers were required to clearly inform consumers which products are sold versus rented, self-hostable versus tied to hosted services, or crippled from running Free software by firmware locks. That would allow a market for freedom-respecting products to actually develop to a reasonable size, and not just to be a fringe thing.
It used to be the case that people valued freedom and the lack of it was something blatantly apparent.
When somebody was a slave, it was a very explicit interpersonal relationship which was very obviously abusive. Even today, some cultures such as Americans are so ashamed of their slaver past that they censor the word on YouTube.
When somebody worked for a company which compensated him not with money but company script which could only be exchanged for goods in company stores, it obviously created a relationship of unequal power which over time put the weaker side at an even bigger and bigger disadvantage. People were able to see and understand this and it was outlawed.
But these days, the power dynamics are so complex and have so many steps and intermediaries, people don't even know what is being taken away from them. It's a salami slicing attack too. There are minor outrages here and there but nothing even changes, two steps forward, one step back to appease them.
---
Bottom line: if a company claims it "sells" you something, the precedent is you own it fully. If you don't, that's theft. Theft, even multi step theft, should be punished in full. That means the company should pay a fine according to how much money they made from their abuse of power, multiplied by a punitive constant.
Additionally, all people involved in the decision making process should also be punished according to how much they stole.
> In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc).
Not if you’re a mainframe customer. Capacity based licensing has been a standard practice in the mainframe world for around 50 years.
On the other hand, the confusingly named DIVX (not the codec) expiring DVD-esque format (so you could play the DVD for 48 hours after purchase, and then after that you had to pay to watch it again. The format totally failed in the market and brought down Circuit City.
> Today I bump into limitations of machines that were put there by manufacturers who are trying to assert ownership of the device after the purchase. In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc). Today you're waging a nearly unwinnable battle against architectures of control baked-in to the hardware at a level below a level that the average end user has any hope of usurping.
Even in the "before times" we had such limitations: the 486 was shipped as a cheaper version with a functional but disabled math coprocessor. There are meaningful differences in practical terms, but I definitely see it as a clear predecessor of this behavior.
Binning products into segments is different than blowing e-fuses. When I bought a 486SX I got a 486SX. It remained forever what it was when I bought it.
In a hypothetical scenario where I somehow "unlocked" the FPU functionality Intel couldn't push out a mandatory firmware upgrade to blow an e-fuse in my chip, fixing the "vulnerability" that allowed me to access the FPU and simultaneously preventing me from ever loading "vulnerable" firmware again (like, say, the Nintendo Switch).
That's different. Lithography masks are expensive. It's cheaper to make them once and use them for different models of the same chip than to make a separate set for each model. My understanding is that in this particular case they would sell the FPU-less version below the cost and the full one with some markup.
In modern times they also do this because the process of semiconductor manufacturing is imperfect and sometimes some parts of the chip would come out damaged. IIRC this happens with GPUs a lot so they tend to have spare cores.
IIRC later revisions of the 486SX have a unique mask set without the FPU, presumably the higher chips per wafer made up for the extra tooling costs.
Also, as a note, unlike modern chips where they fuse off broken cores and sell them as lower specs[1] as part of the binning process, with the early 486SX the FPU was disabled before any testing / binning, so they weren’t selling broken DX dice as SXs.
[1] Or in some cases, fusing if working silicon if the supply / demand curve works that way, see the infamous 3 core AMD Phenom.
They did ship hardware and software that wasn’t fully enabled. But your post implies there was some deception or extortion. System owners were well aware of what they paid for, or that they could upgrade and unlock some things. Today in software we just call it IAP.
I’m not saying it is how we -should-. But that IBM wasn’t rug pulling.
James Yeh, Ken Griffin's "first quant" and eventual co-CIO of Citadel, used to say, when annoyed by a junior who was overfitting to the backtest, "I don't let the computer boss me around! I tell the computer what to do!"
> The greatest trick machines ever pulled was making us believe they haven’t done anything to us.
While "guns don't kill people, people kill people" is a cliché, I think there's still considerable meaning behind it, and I'd say the same holds in the "machines don't do anything to people" sense. Sure, a lot of decision-making and faceless authority is outsourced to machines, but it's still people who are doing that outsourcing, and if those people stopped deciding to put so much weight on the output of (intentionally and unintentionally) black-boxed algorithms then that power of the machines would vanish instantly.
When I was a kid, my dad upgraded our home computer from DOS 5 or 6 to Windows 3.11 for Workgroups. It was the first GUI that I ever used, and it was amazing comparitively. Every app was mysterious and innovative and wonderful.
I tried Borland C++ and it was absolutely confusing, but I was probably just too young. Even QBasic was deeply confusing for a long time, but eventually I finally made a simple, terribly written and horribly broken Bomberman clone.
Those looking to experience something similar to that feeling should buy pico8.
> Even QBasic was deeply confusing for a long time
For one whole year, I thought that Qbasic and Turbo Pascal were text editors that could also run games. I didn't understood that I had access to real compilers and that I could actually change the programs. Sometimes kids are stupid...
As for your Pico8 suggestion, you can always get the open-source equivalent https://tic80.com/ if you don't have the money.
Tic80 is great but Pico8 is better if you can afford it.
And yeah, for a while I avoided strings in QBasic because I didn't have any clue how thread or yarn or whatever had anything to do with writing programs.
I remember being confused why the Pascal/Delphi fractional numbers were called Single, Double, and Real. Like what did those words have to do with being able to use the decimal point?
I used a version of BASIC on my father's accounting computer that had an error message which included the word "ILLEGAL" (I forget what it was, exactly). I always assumed it had something to do with tax laws and the computer warning you not to break them.
In what ways? I paid for both, but I only use TIC-80 since I prefer open source and it has a FreeBSD port. I am also not sure if PICO8 supports alternative languages like Fennel or Janet as well as TIC-80 does?
There's just something more polished, aesthetic and functional about the choice of limitations pico8 offers, whereas tic80 encourages decision paralysis.
I remember being a kid and seeing BASIC in a book from the library and not understanding how to run it. I thought maybe if you saved it in a file with the right extension it would just run. Eventually I figured out how to use the interpreter.
You've reminded me of how I near bricked the family 386 because I wanted to more easily play GORILLAS.BAS.
I was quite used to loading it up in QBASIC.EXE and then executing it to play.
But I wanted to just run it by opening the file in DOSSHELL.
I knew Windows (possibly just DOSSHELL?) had the concept of file associations, so there I went reassociating things in ways I thought might get .BAS to "just run". It didn't work to get gorillas working, and in the process it seemed to mess up a bunch of other things.
This was very late for still using a 386, I think our friends had pentiums by this point.
I don't know if my Dad realised what I'd done and kept quiet about it, or just didn't realise how I'd been fiddling with those settings, but I think the extra "things seem wonky" was a nice excuse for us to finally get upgraded into the windows 95 and CD-ROM era.
This reminds me of the story of an office clerk who after using Excel for a year or two was amazed to learn that it could do calculations and wasn’t just a “tables” program.
Not with the files saved from GW-BASIC though, unless you add the extra A parameter to save as ASCII. I learned that recently when setting up a repo to experiment with GW-BASIC 1.0 (unlike the later Microsoft BASICs the original GW-BASIC is open source).
Annoying when saving the BAS files as ASCII they are still written as full blocks of some size 512 bytes?) with a ctrl-Z EOF to mark the end of the file. So there is some random binary garbage at the end of the file to strip away before committing changes. I wonder how common it was for DOS software to do that and how much sensitive information can be found at the end of released files if you look after the EOF marker?
I guess I was really lucky that I started out on a bbc micro, then got my hands on a c64, then an Amiga, before finally beholding windows 3.0.
By the time I landed in the DOS world aged 8 or so, qbasic was my playground, and was easy to understand from the get-go, and Borland was where I cut my teeth writing something other than basic. One thing it took me a while to get my head around was that a 286 was not a 6502, and practically every little hack, address, anything CPU or memory architecture dependent thing I had learned was now irrelevant.
Coming from Amiga workbench to windows actually felt like a downgrade in many ways, but it was the computer available to me at the time, and retrospectively a good move, as by 3.11 it was clear the wind was blowing to PCs.
Either way, for me, growing with the machine was absolutely formative - the abstraction grew as I did, and I had started near the bottom.
Borland was just confusing. One of the biggest strengths of Visual Basic was how intuitive it was, even for teenagers. There was a reason every AOL prog was written in VB!
The first versions of both Visual Basic and Borland C++ were released in 1991. Turbo C 1.0 was in 1987. Maybe you are thinking of Turbo Pascal 1.0, released in late 1983, but I’d argue that one wasn’t confusing, if you were in the business of using early DOS at all.
I used Turbo Pascal 2.0 in the early 1990's because that was the old version my dad happened to have. You could not just go online and find whatever latest version of software you wanted to.
Still have the tune in my head. Then years later when everyone was waxing lyrical about snake on a Nokia and I’m “isn’t this just nibbles” and nobody knew what I was on about.
Those fill me with so much nostalgia. I think I read in a magazine about moricons.dll, this lead me to opening every dll and exe on our computer to look for icons.
Same here! It was like finding hidden treasure. On a computer without internet and not getting new software often I just wanted to look in every nook and cranny for something interesting.
it is super sad, that these times lasted so little. it is perhaps also because we grew with them, but still. would've been nice if these were with us enough so that we can actually remember which was which .) i mean - they bring some vaporware nostalgia, but actually not all these icons mean at all a thing to me. i dont think they ever meant that much altogether to anyone.
I used these all the time when rolling out Windows 3.1 and 3.11 for thousands of computers back in the early 90's.
You would just pick any icon that seemed relevant, with a focus on not choosing the same icon for two different applications.
Computer GUI's were so new then that people didn't really care if the icon was 100% correct or not.
Sometimes for big applications I would draw up an icon and then use it, but mainly try to stay on moricons.dll or any dll's that came with the application that might contain icons.
I find it funny that I use adaptive icons on android (the ones that switch color according to the OS color scheme) and I override non adaptive app icons the exact same way.
Local public transport app? Choose something resembling a bus, train or similar. Banking app? Pile of money or bill or whatever. Just as long as two apps don’t end up with the same icon it’s all fine.
This. You just needed something and we were starved for icons with the default install, so I would just dig around to find an icon that half-way matched what the application (never "app" then) was.
It’s basically just configuration data for MS-DOS apps running under Windows (or some over MS-DOS multitaskers). Stuff like EMS and XMS memory amounts, various config options etc.
Fun fact: under the hood, PIF files are sent to the same ShellExecute function as EXEs, and if you have an EXE with a PIF extension, it runs the EXE code as normal.
Funnier fact: In Windows 95 and onwards, the UI presents PIFs as a special case of shortcuts, and as with LNK files, Windows always hides (hid?) the extension, even if you have “always show extensions” on. When I get home, I’ll have to check if Windows 11 still has this behaviour…
Edit: Yeah, creating a copy of calc.exe in my ~/Documents folder and renaming it calc.docx.pif does result in a working calculator file that presents gives its name as "calc.docx", albeit with a "shortcut to generic file" icon and a type of "Shortcut to MS-DOS Program" despite the fact that I can't think of any legitimate reason to do anything with a real PIF file on Windows 11 (24H2).
I just checked on a Win11 VM, with show extensions and show hidden turned on, I created a test.pif file on the desktop and the extension is NOT visible. It appears as a shortcut with the little arrow on the icon, and the properties list the file type as "Shortcut to MS-DOS Program (.pif)".
The second link is exactly what I was hoping for, but only seems to exist on the archive. I wasn't able to locate it or other detailed information with my cursory searches. I'm glad it was preserved.
Back in the day, sending the string ".pif" to any MSN Messenger group chat was enough to immediately disable it for everyone. Fun way of dealing with spam :)
This is the kind of thing you should implement in a JS library, or by preprocessing the image before you upload it to the web site, not add to the browser platform spec.
I was half-joking, and primarily meant to point out that the pixelated versions aren’t how those icons looked like to users back in the day either.
But also, for faithful CRT rendering you want to take the actual screen pixels into account, possibly even the subpixel layout, very much like for text rendering. Therefore there’s something to be said for that to be a browser-level feature.
It was kind of Microsoft to include icons for so many products that rivalled their own. It was also a good way to get the MS-DOS user onboard with Windows, even though it was not needed for these MS-DOS applications.
Most of these MS-DOS applications installed into well known folders. Was there any tool that came with Windows to find these applications and put them in program manager groups, or was this something that one had to do for oneself?
The post alludes to it, the previous post has details[1], setup did it, based on the data in APPS.INF (a massive file shipped with Windows, although you could add your own data[2]).
Max nostalgia. When I was ~12 years old I had an 386 PC we got for $5 from the thrift store across from the dump (a bargain for sure even at that time). I had to self-teach myself about DOS, BASIC, Win 3.1, etc. I somehow broke the Win 3.1 installation and all I could find was Win 3.0 on floppy disc. I got the thing reinstalled, but found that the sound drivers didn't work, random programs wouldn't open, and it kept crashing.
I learned years later that there are a huge number of changes between 3.1 and 3.0. The biggest being support for more memory and multimedia extensions. The latter was the first time I learned what a dynamic link library (DLL) was and that took me down the rabbit hole of C++.
I'm not sure about everyone else, but even though I haven't used Windows in about 23 years, every time Raymond Chen has a post on Hacker News it's always a pleasure to read.
My WP days were under DOS (I still remember the cardboard cheat-sheet you got with it that fit around the function keys). But I remember this DLL and wondering where they got the icons. Did they contact the vendors via official channels, or did someone at Microsoft just call up a buddy and ask them what they thought would look good?
One thing the article did not answer is “why?” I think I am missing something but why did Microsoft feel they needed to ship icons for other software vendors’ applications? Wouldn’t Lotus and Quicken want to ship their own icons with their software?
These were existing MS-DOS programmes that had already shipped. They wouldn't have shipped with a Windows icon as they were made before that Windows version existed (or at least shipped) and weren't even intended to run on that platform. Once Windows had shipped, and software vendors started making software for it, they will of course have included their own icons. The "why" is simply Microsoft wanting to make Windows play nice with users' existing software, and thus enhancing the user experience.
Part of the Windows 3.1 installer would search your hard drive for existing applications and add them to the Program Manager. Many of these were DOS applications that had no embedded icon resources. To keep them from all being default applications, they used this little DLL database of icons.
Microsoft wanted to make their new GUI shell for dos look good. There were no apps with icons so they put the most popular and some good stand ins so their new GUI file manager/launcher worked well.
No thanks. Then it just causes confusion when some people use the roman numerals and other people convert them to arabic numerals. Especially for cases like Applause and Crosstalk mentioned above.
You may laugh, but when I heard that Microsoft was working on code-name "Cairo" I turned that around phonetically in my head until I realized they were indeed referring to "Windows Chi-Rho".
I didn't know these were intended for anything. I thought they were provided to users as a swiss-army knife of reasonably expressive icons for use however they saw fit.
They did Turbo Pascal dirty with those icons... I get that Borland C++ also was blue/yellow for the DOS versions (looks like they stopped the DOS version of it in 1993 with version 4). And I assume these were from the early days of Windows so it was a DOS product at that point, but... Man. Poor TP, getting the shaft.
I loved TP, in 93 it was the language we learned for programming in high school. My first compiler was Borland C++ 4.5 bought used from ebay LOL.
N.B. Access for DOS was not a database. It was a communications program with preconfigured settings for popular services like CompuServe and Dow Jones. It wasn't generic like Crosstalk (also featured); to my knowledge it couldn't do anything that it didn't have a configuration for. Clearly aimed at the "on the go" business person who wanted to check stock prices or something without worrying about baud rates or stop bits.
Looking at the icons now I feel the ones where Microsoft put the icon on top of a row of horizontal lines are hard to read at a glance. The lines make the icon visually busy. Probably not so bad when you are on a 640x480 display and those icons are relatively large, but on a modern display they're a bit hard to make out.
Word Perfect!!! I'm almost positive that was the editor they taught me in the early 1990s in grammar school! (We called it grammar school back then, for it was the 1800s.) And yet I had never seen or used it since. This brings back so many memories. I was sitting next to a girl named Dana, the only Dana I ever met.
It was huge back then, but it tanked in the transition to Windows. I kept using the DOS version for years after that because I had muscle memory for WordPerfect’s shortcuts and liked the reveal codes feature.
Also I’ve met two Danas that I can remember. Both were lovely people.
Yep. It was THE program. A bit like how Office was THE program in the late 90s and in the 2000s before Google Docs really started taking off.
The kind of thing people bought computers for. You didn’t need a computer. You needed Word Perfect.
I still remember the little card you could put above the function keys on your keyboard that showed you what alt-F7 or ctrl-F9 did. Each modifier was a different color.
First program I remember seeing people really use on a computer when I was a kid.
As I understand it, WordPerfect held on in the legal professions after most others had switched to MS Office (Word). I guess there were a lot of good templates and many law firms had all their boilerplate documents in WP and didn't want to change.
Word Perfect was like LaTeX and MS Word at the same time. You could edit text or you could edit the codes and there were no nasty surprises or any random reorganization of the document because you copy pasted something.
Also, by editing the codes you could dictate the precise way the document should look.
I think it was ahead of its time.
Sadly, WP 6.0 changed the macro language too much (they made everything an object and many features were lost) and it was not as successful as WP5.1, because you just don't make all the macros of your customers obsolete overnight.
Part of the appeal of WordPerfect (and Reveal Codes) in the legal industry was the ability to create a document that look exactly what you needed it to look like. Appellate briefs, as I remember, had to be no more than a certain number of pages of text of a certain font and point size with specified margins and line spacing, and doing that in Word was doable but a nightmare.
Lotus 1-2-3 was THE program for PCs in the DOS era. WordPerfect was a close second, followed by dBase and then maybe Flight Simulator.
When magazines reported, or OEMs advertised, that a particular computer had "100% IBM compatibility", generally there wasn't like a formal benchmark for this. It basically meant that the PC versions of Lotus and Flight Simulator ran fine on the machine.
I was trying to get across it was one of the juggernauts of the day. A program normal people knew about if you knew the names of any computer programs.
Perhaps it was the most popular for home users, I don’t know. I was far too young in its heyday.
WordPerfect was the go-to in the legal profession. Among other advantages, its "Reveal Codes" feature was an enormous help in making sure documents were formatted correctly, which was essential in law, as courts often have specific formatting requirements for filed documents.
Microsoft Word not only lacked Reveal Codes, it mocked it in an Easter egg in Word for Windows 2.0. Tells you what you need to know about Word, basically.
I remember when I first heard the legal profession still used WP like 2 decades after its heyday. It was a real surprise.
But if you’re going to have a niche market, that’s a pretty good niche. They care about formatting a lot and have a lot of money to spend if it helps them enough.
Google Docs has a feature set more in line with Wordpad (the stripped down word processor that comes with Windows) than Office Word. I can't understand how anyone can use it for anything at all without running into its limitations, so I'm not surprised it's not all that popular outside of basic shared documents.
The vast majority of documents people write for within Google Docs’ limitations. I wouldn’t use it for a book or maybe a dissertation. But it works great for a lot of day to day internal business documents.
I recall hearing that WordPerfect usage survived longer in the legal field.
I also remember in the late 90s, before StarOffice/OpenOffice/LibreOffice took off, WordPerfect had a resurgence on Linux, because there was a need for a word processor. I seem to recall a distro shipping with it prominently. Was it Corel Linux?
I remember when WordPerfect vs Wordstar was the religious debate in the PC world like vi vs emacs. As Macs brought WYSIWYG and desktop publishing to consumers and small business, things got messy.
Does anybody remember AMI Pro? I think that was where I first started to learn about "paragraph styles" and the disciplined way someone could use these WYSIWYG editors a bit like one can use LaTeX styles. Assign styles abstractly, then be able to change in one place and modify all occurrences throughout a document.
However, this just set me up for decades of frustration as the vast majority of business users cannot be bothered to use any discipline at all. Documents would be littered with overlapping bits of manual formatting that would drive me insane with their impossible interactions. And if they defined paragraph styles, it was as endless litter of new styles so that they were hardly ever reused.
It may be my Wordstar indoctrination coming through, but I think these tools would all be better with explicit markup for boundaries, so you could escape the madness of never being able to predict, "which implied style(s) are going to apply to the next character I type with the current cursor position?"
I recall using the Windows version of WordPerfect 6. The performance was awful versus Word on the same hardware.
It was also very unstable. I remember going to a friend's house to use his better computer and WordPerfect 6 to dress up a paper for school. It took hours longer than it should have due to the constant crashes. It was a very solid lesson in "save often" for teenage me.
I adored WordPerfect under DOS. The experience of WP6 for Windows was so bad that I switched to Word and never tried any future WP versions. Maybe they made it better in 7, but the damage was done.
Word Perfect 6.0 (they also had a DOS version) should be taught in software engineering universities as one of the prime examples of product managers killing the company.
5.1 was perfection, but the interface was text console. 6.0 was a complete redesign of … everything, and it was more a bad clone of MS Word than an improvement over WP 5.1.
Man, I had to use Word Perfect in school in 1997. It seemed dumb at the time, but I learned some valuable lessons about markup and the importance of seeing hidden control characters to troubleshoot formatting issues.
I'm around that age. In poorer areas we just had whatever cast-off computers. Circa 1995, my grammar school featured:
- A Franklin ACE (Apple ][ Clone) with an amber screen and two floppy drives
- An Atari ST with a daisywheel printer (Letter quality!)
The newest additions to the school were:
- A Compaq 386 with an amber screen - this is where we had WordPerfect
- A Tandy 1000, 8086 with DOS and a color monitor. Not sure if it could do VGA, definitely CGA was a thing. We ran Where in The World Is Carmen Sandiego and a paint program on that. Because it was in my classroom, this was really the first computer I learned exhaustively. I read its MS-DOS manual cover to cover and enjoyed writing batch files, building launcher menus, etc.
Microsoft C 7. I started programming with this for a few months before Visual C++ came out. It had a DOS IDE that was a bit like QBasic but it's hard to remember the behavior exactly.
Holy crap, Crosstalk XVI. Almost forgot about that one. Edit: Forgot to mention it's a terminal emulator that supported a few file transfer protocols like kermit and xmodem. My father used it to send programs to his CNC machines. I remember getting it working in Windows 3.1 but he did not like or care for Windows as all of the software he ran was DOS: Design CAD 3D, Q&A for word processing and database, and a custom payroll program his friend wrote. His favorite game was gorilla.bas.
I remember clicking many of these as a kid on my friends dads computer and being disappointed because it wasn't the game I thought it was. I particularly remember the FoxPro icon.
I just can't get me enough of Raymond Chen and his wonderful walks down the dustier paths of memory lane. Feels like a more innocent time where I didn't feel like I was imminently going to be turned into paperclips.
It’s probably partially an illusion, but while everything wasn’t fine back then, it seemed that there was at least a vision of a positive self-determined computing future that could be achieved and that we were roughly on-track on. Nowadays it feels more of a fight to keep things not getting worse, most of the time.
Zeroing in on this exact sentiment:
> ...that could be achieved and that we were roughly on-track on.
I think there's a strong lateral connection to this quote:
“I've come up with a set of rules that describe our reactions to technologies:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.”
― Douglas Adams, The Salmon of Doubt: Hitchhiking the Galaxy One Last Time
I think this has interesting implications wrt the perception of nostalgia, because nostalgia seems to be able to apply at any age to any event that happened far enough back in time; while the above theoretical model maps roughly to specific ages.
So I wonder what things are actually a partially overlapped Venn diagram of the above maxim and nostalgia.
In this case I think it's possible the idea that we were "roughly on-track on" with certain technologies - the notion of an emergent sense of structure that was certain to unfold - could map to some point in between points (1) and (2) in the maxim above. An objective analyses would instead recognize "success" as the survivorship bias that it is; but we're not objective :) - and I find that endlessly fascinating!
Dude even kids recognize that TikTok is brainrot.
I once met this older gentleman at a wedding and when I asked about his life he said he used to be really into computers but the personal computing revolution made him sour. It turned professional field step by step into junk until there was only amateurs with their toy computers left.
Did you dare to mention 'vibe coding'?
I agree, but I also think it goes farther than computers and technology. Back then the overall zeitgeist was just a whole lot more positive. These days people are all about one-upping on the dark grittiness of music, movies, etc. But back then things were more upbeat. Yeah, we were worried about the cold war and nuclear armageddon, etc. But we still felt hopeful that we were working toward something better. I miss that.
Computers were so new for most people that they weren't really yet on the "critical path".
And when you got them working, they saved so much time that you had extra time laying around.
Building value vs extracting value.
I'd say, go back to the hardware. Pick a computer hobby that involves tinkering the hardware, not just the software. Holding something on one's hand means a lot.
Yeah, the sense at that time was that you master the machine. Now, increasingly, the machines will master you if you aren't careful. Of course, the machines haven't really done anything to us. They've just been locked down and remotely controlled to deliver ads and misinformation.
That's just past with its rose-tinted glasses. It was easy for someone to master the machine when that someone was a university researcher or a lone gamer, the most precious resource stored in the machine was saved term projects, and either it was not connected to anything else, or connected to fellow university researchers.
The stake was low, because nobody could use your computer to drain your bank account. And someone who would "prank" your computer beyond the social norm would get a stern talking to.
Computers these days have to support your grandma making hotel reservations online without her entire financial information being sent to hackers in Eastern Europe. They're doing jobs that 70s OS designers never thought about. It's a different world.
> ...you master the machine. Now, increasingly, the machines will master you...
Today I bump into limitations of machines that were put there by manufacturers who are trying to assert ownership of the device after the purchase. In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc). Today you're waging a nearly unwinnable battle against architectures of control baked-in to the hardware at a level below a level that the average end user has any hope of usurping.
The machine isn't trying to master me. The people who made the machine are. I wish people in the tech industry wouldn't be party to taking away computing freedom. It pays well, though, and they can console themselves with "It's not a computer, it's a phone"-type delusions (at least until the day "the man" comes for their PCs).
Our civilization desperately needs a way to modify modern microelectronics at home or at least in a well-equipped repair shop.
Regular people being able to commit contempt of companies' business models en masse seems to work well to keep them in check, but it's becoming ever harder with so much of everything becoming mobile-centric. And with all smartphones being locked down at the level of someone else's public keys being burned into the SoC at the factory, you can't do shit. They literally have technological supremacy over the rest of the humanity. And we're somehow okay with that.
> Our civilization desperately needs a way to modify modern microelectronics at home or at least in a well-equipped repair shop.
I'll take consumer protection regulation, at least in the short term.
I wish manufacturers were required to clearly inform consumers which products are sold versus rented, self-hostable versus tied to hosted services, or crippled from running Free software by firmware locks. That would allow a market for freedom-respecting products to actually develop to a reasonable size, and not just to be a fringe thing.
Not just commit contempt, we should punish them.
It used to be the case that people valued freedom and the lack of it was something blatantly apparent.
When somebody was a slave, it was a very explicit interpersonal relationship which was very obviously abusive. Even today, some cultures such as Americans are so ashamed of their slaver past that they censor the word on YouTube.
When somebody worked for a company which compensated him not with money but company script which could only be exchanged for goods in company stores, it obviously created a relationship of unequal power which over time put the weaker side at an even bigger and bigger disadvantage. People were able to see and understand this and it was outlawed.
But these days, the power dynamics are so complex and have so many steps and intermediaries, people don't even know what is being taken away from them. It's a salami slicing attack too. There are minor outrages here and there but nothing even changes, two steps forward, one step back to appease them.
---
Bottom line: if a company claims it "sells" you something, the precedent is you own it fully. If you don't, that's theft. Theft, even multi step theft, should be punished in full. That means the company should pay a fine according to how much money they made from their abuse of power, multiplied by a punitive constant.
Additionally, all people involved in the decision making process should also be punished according to how much they stole.
> In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc).
Not if you’re a mainframe customer. Capacity based licensing has been a standard practice in the mainframe world for around 50 years.
Part of the magic of the personal computing revolution was breaking away from these shackles.
As a society, we’ve since decided that we actually prefer to rent everything and to have no agency.
I don't think the marketplace does a good job of explaining the difference to people. That might change things. Maybe. Probably not. >sigh<
On the other hand, the confusingly named DIVX (not the codec) expiring DVD-esque format (so you could play the DVD for 48 hours after purchase, and then after that you had to pay to watch it again. The format totally failed in the market and brought down Circuit City.
So that suggests it's a user education issue.
> Today I bump into limitations of machines that were put there by manufacturers who are trying to assert ownership of the device after the purchase. In the "before times" limitations were either a fact of the hardware (i.e. you only have so much RAM, storage, CPU cycles, etc) or of your own ability (you don't know how to crack the protection, defeat the anti-debug tricks, etc). Today you're waging a nearly unwinnable battle against architectures of control baked-in to the hardware at a level below a level that the average end user has any hope of usurping.
Even in the "before times" we had such limitations: the 486 was shipped as a cheaper version with a functional but disabled math coprocessor. There are meaningful differences in practical terms, but I definitely see it as a clear predecessor of this behavior.
Binning products into segments is different than blowing e-fuses. When I bought a 486SX I got a 486SX. It remained forever what it was when I bought it.
In a hypothetical scenario where I somehow "unlocked" the FPU functionality Intel couldn't push out a mandatory firmware upgrade to blow an e-fuse in my chip, fixing the "vulnerability" that allowed me to access the FPU and simultaneously preventing me from ever loading "vulnerable" firmware again (like, say, the Nintendo Switch).
That's different. Lithography masks are expensive. It's cheaper to make them once and use them for different models of the same chip than to make a separate set for each model. My understanding is that in this particular case they would sell the FPU-less version below the cost and the full one with some markup.
In modern times they also do this because the process of semiconductor manufacturing is imperfect and sometimes some parts of the chip would come out damaged. IIRC this happens with GPUs a lot so they tend to have spare cores.
IIRC later revisions of the 486SX have a unique mask set without the FPU, presumably the higher chips per wafer made up for the extra tooling costs.
Also, as a note, unlike modern chips where they fuse off broken cores and sell them as lower specs[1] as part of the binning process, with the early 486SX the FPU was disabled before any testing / binning, so they weren’t selling broken DX dice as SXs.
[1] Or in some cases, fusing if working silicon if the supply / demand curve works that way, see the infamous 3 core AMD Phenom.
"The first thing we do, is we kill all the law--" Er, I mean abolish the DMCA.
Picking a lock on a device you own shouldn't be a federal crime.
Before the before times, there are claims that IBM os360 would be delivered purposely handicapped until you paid the extra fees for the upgrade
They did ship hardware and software that wasn’t fully enabled. But your post implies there was some deception or extortion. System owners were well aware of what they paid for, or that they could upgrade and unlock some things. Today in software we just call it IAP.
I’m not saying it is how we -should-. But that IBM wasn’t rug pulling.
This is “my computer” This is “my phone”
Telemetry? Subscription based software? SMS text with 2FA just to open outlook that I purchased and that’s installed on my computer?
Ok, wait, whose computer is this? Whose phone is it ?
James Yeh, Ken Griffin's "first quant" and eventual co-CIO of Citadel, used to say, when annoyed by a junior who was overfitting to the backtest, "I don't let the computer boss me around! I tell the computer what to do!"
The greatest trick machines ever pulled was making us believe they haven’t done anything to us.
> The greatest trick machines ever pulled was making us believe they haven’t done anything to us.
While "guns don't kill people, people kill people" is a cliché, I think there's still considerable meaning behind it, and I'd say the same holds in the "machines don't do anything to people" sense. Sure, a lot of decision-making and faceless authority is outsourced to machines, but it's still people who are doing that outsourcing, and if those people stopped deciding to put so much weight on the output of (intentionally and unintentionally) black-boxed algorithms then that power of the machines would vanish instantly.
I love his posts. Read every one of them.
When I was a kid, my dad upgraded our home computer from DOS 5 or 6 to Windows 3.11 for Workgroups. It was the first GUI that I ever used, and it was amazing comparitively. Every app was mysterious and innovative and wonderful.
I tried Borland C++ and it was absolutely confusing, but I was probably just too young. Even QBasic was deeply confusing for a long time, but eventually I finally made a simple, terribly written and horribly broken Bomberman clone.
Those looking to experience something similar to that feeling should buy pico8.
> Even QBasic was deeply confusing for a long time
For one whole year, I thought that Qbasic and Turbo Pascal were text editors that could also run games. I didn't understood that I had access to real compilers and that I could actually change the programs. Sometimes kids are stupid...
As for your Pico8 suggestion, you can always get the open-source equivalent https://tic80.com/ if you don't have the money.
It's a testimony of Turbo Pascal team.. the things was so lean and swift, compilation was near transparent. All this on early pentium and old cpus..
I continue to run Turbo Pascal on a Z80-based machine, with 64k of RAM. A pentium would be luxury!
Amstrad CPC or Tatung Einstein?
(Hi Steve!)
Custom hardware running CP/M :)
Oh very cool.
I liked Turbo Pascal when I was young. Debugger just works. Peeking into variables just works.
Unfortunately, now I used print to debug for other languages because I thought debugger is too hard to setup
Tic80 is great but Pico8 is better if you can afford it.
And yeah, for a while I avoided strings in QBasic because I didn't have any clue how thread or yarn or whatever had anything to do with writing programs.
I remember being confused why the Pascal/Delphi fractional numbers were called Single, Double, and Real. Like what did those words have to do with being able to use the decimal point?
Sharing fun kid computer misconceptions:
I used a version of BASIC on my father's accounting computer that had an error message which included the word "ILLEGAL" (I forget what it was, exactly). I always assumed it had something to do with tax laws and the computer warning you not to break them.
In what ways? I paid for both, but I only use TIC-80 since I prefer open source and it has a FreeBSD port. I am also not sure if PICO8 supports alternative languages like Fennel or Janet as well as TIC-80 does?
There's just something more polished, aesthetic and functional about the choice of limitations pico8 offers, whereas tic80 encourages decision paralysis.
I remember being a kid and seeing BASIC in a book from the library and not understanding how to run it. I thought maybe if you saved it in a file with the right extension it would just run. Eventually I figured out how to use the interpreter.
You've reminded me of how I near bricked the family 386 because I wanted to more easily play GORILLAS.BAS.
I was quite used to loading it up in QBASIC.EXE and then executing it to play.
But I wanted to just run it by opening the file in DOSSHELL.
I knew Windows (possibly just DOSSHELL?) had the concept of file associations, so there I went reassociating things in ways I thought might get .BAS to "just run". It didn't work to get gorillas working, and in the process it seemed to mess up a bunch of other things.
This was very late for still using a 386, I think our friends had pentiums by this point.
I don't know if my Dad realised what I'd done and kept quiet about it, or just didn't realise how I'd been fiddling with those settings, but I think the extra "things seem wonky" was a nice excuse for us to finally get upgraded into the windows 95 and CD-ROM era.
This reminds me of the story of an office clerk who after using Excel for a year or two was amazed to learn that it could do calculations and wasn’t just a “tables” program.
Then you learned your mistake and assumed that nibbles in FastTrackerII was coded into a module. Computers are hard.
That's ok, it took me like a decade to realize you could edit .bas files in any text editor.
Not with the files saved from GW-BASIC though, unless you add the extra A parameter to save as ASCII. I learned that recently when setting up a repo to experiment with GW-BASIC 1.0 (unlike the later Microsoft BASICs the original GW-BASIC is open source).
Annoying when saving the BAS files as ASCII they are still written as full blocks of some size 512 bytes?) with a ctrl-Z EOF to mark the end of the file. So there is some random binary garbage at the end of the file to strip away before committing changes. I wonder how common it was for DOS software to do that and how much sensitive information can be found at the end of released files if you look after the EOF marker?
I guess I was really lucky that I started out on a bbc micro, then got my hands on a c64, then an Amiga, before finally beholding windows 3.0.
By the time I landed in the DOS world aged 8 or so, qbasic was my playground, and was easy to understand from the get-go, and Borland was where I cut my teeth writing something other than basic. One thing it took me a while to get my head around was that a 286 was not a 6502, and practically every little hack, address, anything CPU or memory architecture dependent thing I had learned was now irrelevant.
Coming from Amiga workbench to windows actually felt like a downgrade in many ways, but it was the computer available to me at the time, and retrospectively a good move, as by 3.11 it was clear the wind was blowing to PCs.
Either way, for me, growing with the machine was absolutely formative - the abstraction grew as I did, and I had started near the bottom.
Borland was just confusing. One of the biggest strengths of Visual Basic was how intuitive it was, even for teenagers. There was a reason every AOL prog was written in VB!
VB came around 9 years later.
The first versions of both Visual Basic and Borland C++ were released in 1991. Turbo C 1.0 was in 1987. Maybe you are thinking of Turbo Pascal 1.0, released in late 1983, but I’d argue that one wasn’t confusing, if you were in the business of using early DOS at all.
I used Turbo Pascal 2.0 in the early 1990's because that was the old version my dad happened to have. You could not just go online and find whatever latest version of software you wanted to.
I perused the files section of qbasic.com back then: https://web.archive.org/web/20050804015051/http://www.qbasic...
(One of my favorites is "3D Experiment" in category "Graphics": it shows a wireframe model of a spaceship that can be manipulated with the keyboard.)
Ah QBASIC my first love
My first bug fix ever was getting rid of that stupid tail that kept growing and wouldn't go away when you moved in NIBBLES.BAS.
Still have the tune in my head. Then years later when everyone was waxing lyrical about snake on a Nokia and I’m “isn’t this just nibbles” and nobody knew what I was on about.
Man qbasic and borland C were great on DOS
My college decided C++ was the first language they’d teach us. There was definitely a curve there.
Those fill me with so much nostalgia. I think I read in a magazine about moricons.dll, this lead me to opening every dll and exe on our computer to look for icons.
Same here! It was like finding hidden treasure. On a computer without internet and not getting new software often I just wanted to look in every nook and cranny for something interesting.
it is super sad, that these times lasted so little. it is perhaps also because we grew with them, but still. would've been nice if these were with us enough so that we can actually remember which was which .) i mean - they bring some vaporware nostalgia, but actually not all these icons mean at all a thing to me. i dont think they ever meant that much altogether to anyone.
I used these all the time when rolling out Windows 3.1 and 3.11 for thousands of computers back in the early 90's.
You would just pick any icon that seemed relevant, with a focus on not choosing the same icon for two different applications.
Computer GUI's were so new then that people didn't really care if the icon was 100% correct or not.
Sometimes for big applications I would draw up an icon and then use it, but mainly try to stay on moricons.dll or any dll's that came with the application that might contain icons.
I find it funny that I use adaptive icons on android (the ones that switch color according to the OS color scheme) and I override non adaptive app icons the exact same way.
Local public transport app? Choose something resembling a bus, train or similar. Banking app? Pile of money or bill or whatever. Just as long as two apps don’t end up with the same icon it’s all fine.
What is old is new I guess.
This. You just needed something and we were starved for icons with the default install, so I would just dig around to find an icon that half-way matched what the application (never "app" then) was.
Remember the good old days of editing PIF (Program Information Files) files [⁂]? Ah yeah.
Googling a bit, it looks like a lot of this lore has just been lost. I don't know if there are modern explanations of PIF files kicking around.
⁂ I realize this is an ATM machine phrasing, but we called them PIF files in the day.
It’s basically just configuration data for MS-DOS apps running under Windows (or some over MS-DOS multitaskers). Stuff like EMS and XMS memory amounts, various config options etc.
Fun fact: under the hood, PIF files are sent to the same ShellExecute function as EXEs, and if you have an EXE with a PIF extension, it runs the EXE code as normal.
Funnier fact: In Windows 95 and onwards, the UI presents PIFs as a special case of shortcuts, and as with LNK files, Windows always hides (hid?) the extension, even if you have “always show extensions” on. When I get home, I’ll have to check if Windows 11 still has this behaviour…
Edit: Yeah, creating a copy of calc.exe in my ~/Documents folder and renaming it calc.docx.pif does result in a working calculator file that presents gives its name as "calc.docx", albeit with a "shortcut to generic file" icon and a type of "Shortcut to MS-DOS Program" despite the fact that I can't think of any legitimate reason to do anything with a real PIF file on Windows 11 (24H2).
I just checked on a Win11 VM, with show extensions and show hidden turned on, I created a test.pif file on the desktop and the extension is NOT visible. It appears as a shortcut with the little arrow on the icon, and the properties list the file type as "Shortcut to MS-DOS Program (.pif)".
Even funier fact: the fact that Windows treated pif and lnk files as executable lead to a lot of exploits.
There's still plenty of information on the file format kicking around:
https://www.fileformat.info/format/pif/corion.htm
https://web.archive.org/web/20220214185118/http://www.smsoft...
As well as a basic explanation of the file's purpose:
http://justsolve.archiveteam.org/wiki/Program_information_fi...
The second link is exactly what I was hoping for, but only seems to exist on the archive. I wasn't able to locate it or other detailed information with my cursory searches. I'm glad it was preserved.
For now, depending on how long the Archive survives.
Back in the day, sending the string ".pif" to any MSN Messenger group chat was enough to immediately disable it for everyone. Fun way of dealing with spam :)
Pretty unconventional to use an asterism as a footnote symbol.
> PIF
it's also an onomatopoeia for opening up a can of biscuits
Ah, lovely blast from the past! I remember finding this DLL and being delighted that I could now put pretty icons on more things.
Also remember taking ages to figure out that it meant "more icons" rather than just a silly made up word.
They were composed by Ennio Moricons.
I can hear this comment.
> Also remember taking ages to figure out that it meant "more icons" rather than just a silly made up word.
The good ole' days of having to figure out a meaningful name in 8 chars.
That is almost my favorite part of DOS retro computing. Every file to name is a little puzzle.
Moricons' Magical Watchdog!
Moraff's World was right there, it wasn't hard to think it was related.
Try this CSS if you want to zoom in without the icons turning into a blurry mess:
There should be an `image-rendering: crt;`.
This is the kind of thing you should implement in a JS library, or by preprocessing the image before you upload it to the web site, not add to the browser platform spec.
I was half-joking, and primarily meant to point out that the pixelated versions aren’t how those icons looked like to users back in the day either.
But also, for faithful CRT rendering you want to take the actual screen pixels into account, possibly even the subpixel layout, very much like for text rendering. Therefore there’s something to be said for that to be a browser-level feature.
Now, this is the kind of obscure hacking I'm on HN for!
Also, I wish I had known about that little trick ... years ... ago.
It was kind of Microsoft to include icons for so many products that rivalled their own. It was also a good way to get the MS-DOS user onboard with Windows, even though it was not needed for these MS-DOS applications.
Most of these MS-DOS applications installed into well known folders. Was there any tool that came with Windows to find these applications and put them in program manager groups, or was this something that one had to do for oneself?
The post alludes to it, the previous post has details[1], setup did it, based on the data in APPS.INF (a massive file shipped with Windows, although you could add your own data[2]).
[1]: https://devblogs.microsoft.com/oldnewthing/20250505-00/?p=11...
[2]: http://www.bitsavers.org/pdf/microsoft/resource_kits/0030-54... page 183
Max nostalgia. When I was ~12 years old I had an 386 PC we got for $5 from the thrift store across from the dump (a bargain for sure even at that time). I had to self-teach myself about DOS, BASIC, Win 3.1, etc. I somehow broke the Win 3.1 installation and all I could find was Win 3.0 on floppy disc. I got the thing reinstalled, but found that the sound drivers didn't work, random programs wouldn't open, and it kept crashing.
I learned years later that there are a huge number of changes between 3.1 and 3.0. The biggest being support for more memory and multimedia extensions. The latter was the first time I learned what a dynamic link library (DLL) was and that took me down the rabbit hole of C++.
I'm not sure about everyone else, but even though I haven't used Windows in about 23 years, every time Raymond Chen has a post on Hacker News it's always a pleasure to read.
How many of us saw the title and the microsoft.com label and immediately knew it was Raymond Chen? He's definitely a treasure!
Count me in!
Having worked at WordPerfect during the days of Windows 3.0/3.1, I'm surprised they accommodated WP with icons.
My WP days were under DOS (I still remember the cardboard cheat-sheet you got with it that fit around the function keys). But I remember this DLL and wondering where they got the icons. Did they contact the vendors via official channels, or did someone at Microsoft just call up a buddy and ask them what they thought would look good?
I’m really amazed by the fact how expressive and solid such low-res icons can be. Creating them is a true art
Indeed - we now routinely do so much less with so much more!
Can I quote you when I finally release my app in a week or two?
One thing the article did not answer is “why?” I think I am missing something but why did Microsoft feel they needed to ship icons for other software vendors’ applications? Wouldn’t Lotus and Quicken want to ship their own icons with their software?
These were existing MS-DOS programmes that had already shipped. They wouldn't have shipped with a Windows icon as they were made before that Windows version existed (or at least shipped) and weren't even intended to run on that platform. Once Windows had shipped, and software vendors started making software for it, they will of course have included their own icons. The "why" is simply Microsoft wanting to make Windows play nice with users' existing software, and thus enhancing the user experience.
Part of the Windows 3.1 installer would search your hard drive for existing applications and add them to the Program Manager. Many of these were DOS applications that had no embedded icon resources. To keep them from all being default applications, they used this little DLL database of icons.
The 'why' was answered in the first article of this series.
https://devblogs.microsoft.com/oldnewthing/20250505-00/?p=11...
Microsoft wanted to make their new GUI shell for dos look good. There were no apps with icons so they put the most popular and some good stand ins so their new GUI file manager/launcher worked well.
The article links to earlier blog posts that explain the “why”.
Roman numerals in software versioning should make a comeback.
The Cicero-approved desktop of MS-DOS programs from this moricons.dll set would include these:
Applause II 1.5
Framework III
Crosstalk-XVI 3.71
PC Paintbrush IV Plus
And of course you’d want dBase III+ and Deluxe Paint II Enhanced.
No thanks. Then it just causes confusion when some people use the roman numerals and other people convert them to arabic numerals. Especially for cases like Applause and Crosstalk mentioned above.
I support this. "Windows XI" looks so much more professional and enigmatic.
You may laugh, but when I heard that Microsoft was working on code-name "Cairo" I turned that around phonetically in my head until I realized they were indeed referring to "Windows Chi-Rho".
https://en.wikipedia.org/wiki/Chi_Rho
https://en.wikipedia.org/wiki/Windows_XP
They should go all the way to Windows MMXXV etc.
This website best experienced with Chrome CXXXVII+ on Windows XI Enhanced.
Build ↂↂↁↀⅭ.ↀↁⅬⅩⅠ
Lotus I-II-III?
I didn't know these were intended for anything. I thought they were provided to users as a swiss-army knife of reasonably expressive icons for use however they saw fit.
At least that's how I used them.
Man, things had character back then. Everything is so flat and boring now.
More importantly, though, Raymond Chen is an absolute treasure. I love these posts.
They did Turbo Pascal dirty with those icons... I get that Borland C++ also was blue/yellow for the DOS versions (looks like they stopped the DOS version of it in 1993 with version 4). And I assume these were from the early days of Windows so it was a DOS product at that point, but... Man. Poor TP, getting the shaft.
I loved TP, in 93 it was the language we learned for programming in high school. My first compiler was Borland C++ 4.5 bought used from ebay LOL.
N.B. Access for DOS was not a database. It was a communications program with preconfigured settings for popular services like CompuServe and Dow Jones. It wasn't generic like Crosstalk (also featured); to my knowledge it couldn't do anything that it didn't have a configuration for. Clearly aimed at the "on the go" business person who wanted to check stock prices or something without worrying about baud rates or stop bits.
Why can I hear these icons?
Seriously, all of the window .wav files just got loaded from my long term memory, along with the sound of Windows launching from a magnetic HDD.
Oh yeah, sound of rats inside of water pipe.
OR floppy sound - startup button, boop beep trrrrrrrr bomp.
Looking at the icons now I feel the ones where Microsoft put the icon on top of a row of horizontal lines are hard to read at a glance. The lines make the icon visually busy. Probably not so bad when you are on a 640x480 display and those icons are relatively large, but on a modern display they're a bit hard to make out.
Word Perfect!!! I'm almost positive that was the editor they taught me in the early 1990s in grammar school! (We called it grammar school back then, for it was the 1800s.) And yet I had never seen or used it since. This brings back so many memories. I was sitting next to a girl named Dana, the only Dana I ever met.
It was huge back then, but it tanked in the transition to Windows. I kept using the DOS version for years after that because I had muscle memory for WordPerfect’s shortcuts and liked the reveal codes feature.
Also I’ve met two Danas that I can remember. Both were lovely people.
Yep. It was THE program. A bit like how Office was THE program in the late 90s and in the 2000s before Google Docs really started taking off.
The kind of thing people bought computers for. You didn’t need a computer. You needed Word Perfect.
I still remember the little card you could put above the function keys on your keyboard that showed you what alt-F7 or ctrl-F9 did. Each modifier was a different color.
First program I remember seeing people really use on a computer when I was a kid.
As I understand it, WordPerfect held on in the legal professions after most others had switched to MS Office (Word). I guess there were a lot of good templates and many law firms had all their boilerplate documents in WP and didn't want to change.
Have you ever used LaTeX?
Word Perfect was like LaTeX and MS Word at the same time. You could edit text or you could edit the codes and there were no nasty surprises or any random reorganization of the document because you copy pasted something.
Also, by editing the codes you could dictate the precise way the document should look.
I think it was ahead of its time.
Sadly, WP 6.0 changed the macro language too much (they made everything an object and many features were lost) and it was not as successful as WP5.1, because you just don't make all the macros of your customers obsolete overnight.
The loss of `reveal codes` is still felt when Word does arbitrary craziness with documents!
Does (did) Word even have "codes" as such? I thought it was quite different from WP in that regard. Modern Word .docx is an XML format now anyway.
Part of the appeal of WordPerfect (and Reveal Codes) in the legal industry was the ability to create a document that look exactly what you needed it to look like. Appellate briefs, as I remember, had to be no more than a certain number of pages of text of a certain font and point size with specified margins and line spacing, and doing that in Word was doable but a nightmare.
Lotus 1-2-3 was THE program for PCs in the DOS era. WordPerfect was a close second, followed by dBase and then maybe Flight Simulator.
When magazines reported, or OEMs advertised, that a particular computer had "100% IBM compatibility", generally there wasn't like a formal benchmark for this. It basically meant that the PC versions of Lotus and Flight Simulator ran fine on the machine.
Right. It wasn’t the only or the biggest.
I was trying to get across it was one of the juggernauts of the day. A program normal people knew about if you knew the names of any computer programs.
Perhaps it was the most popular for home users, I don’t know. I was far too young in its heyday.
As productivity software went though it was BIG.
WordPerfect was the go-to in the legal profession. Among other advantages, its "Reveal Codes" feature was an enormous help in making sure documents were formatted correctly, which was essential in law, as courts often have specific formatting requirements for filed documents.
Microsoft Word not only lacked Reveal Codes, it mocked it in an Easter egg in Word for Windows 2.0. Tells you what you need to know about Word, basically.
I remember when I first heard the legal profession still used WP like 2 decades after its heyday. It was a real surprise.
But if you’re going to have a niche market, that’s a pretty good niche. They care about formatting a lot and have a lot of money to spend if it helps them enough.
A bit like how Office was THE program in the late 90s and in the 2000s before Google Docs really started taking off.
This is wishful thinking. Google Docs is only hovering around 10% market share.
Google Docs has a feature set more in line with Wordpad (the stripped down word processor that comes with Windows) than Office Word. I can't understand how anyone can use it for anything at all without running into its limitations, so I'm not surprised it's not all that popular outside of basic shared documents.
I would say Works, it’s much more than Wordpad.
The vast majority of documents people write for within Google Docs’ limitations. I wouldn’t use it for a book or maybe a dissertation. But it works great for a lot of day to day internal business documents.
That seems low to me, but I wouldn’t be surprised at all if it’s not dominant. However it has an awful lot of mindshare.
The only keyboard thing similar to that I ever used was the full keyboard cover that came with the Flintstones PC game.
This is either it or very very close:
http://xahlee.info/kbd/wordperfect_shortcuts_strip.html
And for context, this link has a picture of it on top of a keyboard (cropped):
https://luciamonterorodriguez.com/atajos-de-teclado-y-raton-...
I remember Janes AH-64 Longbow also had a keyboard cover.
https://archive.org/details/janes-ah-64-d-longbow-95/keyboar...
I recall hearing that WordPerfect usage survived longer in the legal field.
I also remember in the late 90s, before StarOffice/OpenOffice/LibreOffice took off, WordPerfect had a resurgence on Linux, because there was a need for a word processor. I seem to recall a distro shipping with it prominently. Was it Corel Linux?
I remember when WordPerfect vs Wordstar was the religious debate in the PC world like vi vs emacs. As Macs brought WYSIWYG and desktop publishing to consumers and small business, things got messy.
Does anybody remember AMI Pro? I think that was where I first started to learn about "paragraph styles" and the disciplined way someone could use these WYSIWYG editors a bit like one can use LaTeX styles. Assign styles abstractly, then be able to change in one place and modify all occurrences throughout a document.
However, this just set me up for decades of frustration as the vast majority of business users cannot be bothered to use any discipline at all. Documents would be littered with overlapping bits of manual formatting that would drive me insane with their impossible interactions. And if they defined paragraph styles, it was as endless litter of new styles so that they were hardly ever reused.
It may be my Wordstar indoctrination coming through, but I think these tools would all be better with explicit markup for boundaries, so you could escape the madness of never being able to predict, "which implied style(s) are going to apply to the next character I type with the current cursor position?"
You can lock down almost all the manual formatting in Word (i.e. allow styles only), and you can make a template out of it too.
I remember a version of WordPerfect on an AT&T Unix workstation someone was selling in the early/mid-90s. I wanted one so bad.
WordPerfect 7 during the win95 era was alright... but businesses went office, and compatibility was a crapshoot, so home users followed suit.
I recall using the Windows version of WordPerfect 6. The performance was awful versus Word on the same hardware.
It was also very unstable. I remember going to a friend's house to use his better computer and WordPerfect 6 to dress up a paper for school. It took hours longer than it should have due to the constant crashes. It was a very solid lesson in "save often" for teenage me.
I adored WordPerfect under DOS. The experience of WP6 for Windows was so bad that I switched to Word and never tried any future WP versions. Maybe they made it better in 7, but the damage was done.
Word Perfect 6.0 (they also had a DOS version) should be taught in software engineering universities as one of the prime examples of product managers killing the company.
5.1 was perfection, but the interface was text console. 6.0 was a complete redesign of … everything, and it was more a bad clone of MS Word than an improvement over WP 5.1.
WordPad ! shipped with windows for 20 years after that. for free. .RTF! :)
When I first read the Hitchhiker’s Guide, I thought “Ford Prefect” was a pun on “Word Perfect”, not knowing anything about the car model at the time.
Man, I had to use Word Perfect in school in 1997. It seemed dumb at the time, but I learned some valuable lessons about markup and the importance of seeing hidden control characters to troubleshoot formatting issues.
Maybe her parents liked Ghostbusters?
While it’s hard to imagine someone using Word Perfect in grammar school born during or after 1984, I hold out hope that her middle initial was “Z”.
That said, I believe I learned spreadsheets in high school using Lotus for DOS in 1996, but can’t imagine kids 3 years later still doing the same.
I'm around that age. In poorer areas we just had whatever cast-off computers. Circa 1995, my grammar school featured:
- A Franklin ACE (Apple ][ Clone) with an amber screen and two floppy drives
- An Atari ST with a daisywheel printer (Letter quality!)
The newest additions to the school were:
- A Compaq 386 with an amber screen - this is where we had WordPerfect
- A Tandy 1000, 8086 with DOS and a color monitor. Not sure if it could do VGA, definitely CGA was a thing. We ran Where in The World Is Carmen Sandiego and a paint program on that. Because it was in my classroom, this was really the first computer I learned exhaustively. I read its MS-DOS manual cover to cover and enjoyed writing batch files, building launcher menus, etc.
I always thought the fancy DOS icon meant the program was broken - it looked sort of like a sinking ship to me.
Microsoft C 7. I started programming with this for a few months before Visual C++ came out. It had a DOS IDE that was a bit like QBasic but it's hard to remember the behavior exactly.
Another delightful The New Old Thing post, ever living up to the name.
I used many of those icons...
I wonder how many people out there actually created an icon for Kid Pix and used the icon from Moricons.dll? I was one of those people.
Oh man that brings back memories of Windows 3.1 and makes me feel oh so old.
Harvard Graphics! That's where it all started for me.
Did they store the icons as byte arrays?
As it happens, Raymond Chen has a blog post on that, too.
https://devblogs.microsoft.com/oldnewthing/20231025-00/?p=10...
Thank you!
Holy crap, Crosstalk XVI. Almost forgot about that one. Edit: Forgot to mention it's a terminal emulator that supported a few file transfer protocols like kermit and xmodem. My father used it to send programs to his CNC machines. I remember getting it working in Windows 3.1 but he did not like or care for Windows as all of the software he ran was DOS: Design CAD 3D, Q&A for word processing and database, and a custom payroll program his friend wrote. His favorite game was gorilla.bas.
I remember clicking many of these as a kid on my friends dads computer and being disappointed because it wasn't the game I thought it was. I particularly remember the FoxPro icon.
[dead]
[dead]