Updating the most influential book of the BASIC era

> Every time a user does a repetitive task that should have been a loop they created themself, that waste of life is a small death, and the blood is on our industry.

The majority of people have no interest in programming, no matter how humane a language were handed to them.

Actually there's a reason beyond stupidity and conspiracy for things going the way they did. It is very easy to make simple things simple, but by doing that you will likely turn complex things even more complex.

The first example shows a video-poker-like game implemented in 98 lines. You can get close to that with processing or lazarus-ide. Now try to do that so you can play on-line, using the browser, with multiple people at the same time, storing score, records, with login, a network of friends, authentication... You'll end up seeing that node developers and users are neither stupid nor conspiracists.

At least VB.NET is part of the list.

I guess the only modern environments where we could replicate the experience would be something like running MicroPython on an Raspberry Pi 400.

Or the Swift Playgrounds (assuming an external keyboard).

> I’m pretty disenchanted by the choice of languages. None of them embody the human-first optimism of early BASICs.

Nothing yet exists that has the same properties: simplicity and automatic availability. JS wins on availability as it is present on almost every general purpose device via a web browser, but the syntax is not as human friendly as

    10 PRINT "Hello"
    20 GOTO 10

or better, lets tweak the language a bit and get rid of the line numbers:

    This is the start
    Print "Hello"
    Goto the start

Though as much as we love to hate them, I'd say spreadsheets come closest to That What Was because the people who use them are automatically there for work or to do some basic calcs for themselves, and the “programming” side of it is just there for them to discover as they need more complexity, they don't need to decide “I want to program”, they have a problem, they look up a solution, find one online, and modify it for their own needs, much like me in the 80s typing in a 10-line graphics demo or basic game and learning more by tweaking it. To a lesser extent things like minecraft count too.

> I don’t have a strong proposition for a specific language.

> People try to tell me that dream is dead. God I hope not,

There is no such language for current times, so the dream of reproducing the feeling BBC BASIC¹ in the days of yore because not only has the tech mover on but so have people's expectations. No one is going to be satisfied with those simple introductions as we were back then, so the barrier for entry is higher - the language and wider environment need to provide much more and make all that feel like the heady days of drawing random coloured dots in graphics mode 2 while, hiding all the complexity long enough to not put people off, while not limiting itself by hiding that complexity (making the complex more so, or even impossible) so that people give up when they discover they need to move to something else to get beyond a certain point⁴. And it needs to somehow compete with the rest of the world, gone are the days that you are sat in front of a disconnected computer with not much else to do but experiment and learn - it can't wait to be discovered it needs to drag people away from consuming and actively attract them to creating⁴. And we aren't just dealing with a bunch of nerds/geeks who speak at least a little English, anything like this these days needs to be embarrassingly easy to perfectly internationalise⁵, without breaking anything⁴. And it then has to get commonly accepted to become popular and therefore survive³. If you create a language/environment that manages all that⁴, then why not finish off your morning with breakfast at MILLYWAYS?!

[1] The best 8-bit basic, bar non. No silly-short naming limits, proper procedures and functions² rather than just GOTO & GOSUB, a built-in multi-pass assembler for if you got really adventurous, ...

[2] Well, there were significant limits compared to what we have now, but understandably so in such small power & memory.

[3] Not unlike how gods and belief work in Pratchett's Discworld.

[4] This is, of course, impossible.

[5] ref: “reflected and it’s very English syntax”, and it isn't nearly as simple as just swapping keywords out for local ones either.

> BASIC was designed for normal humans to actually get work done,

I hate to break it to you but we were, at best, a subset of normal humans, in the eyes of the rest of the humans!

> Every time a user does a repetitive task that should have been a loop they created themself,

That is rather too dogmatic. I'm a dev-cum-DBA with decades of inexperience behind me, and I do many repetitive things that I've not yet automated. If you automate everything you do more than twice you end up automating things that you'll never use again after next week, and never actually do much beyond making and testing automations.

> that dream is dead

Going back to that: the general dream is dead IMO. We need to stop hoping to find the One True Path to lead people along. The dream can still exist in smaller parts though, if we accept multiple very different paths without disparaging any particular ones of them: people starting by automating Minecraft stuff, people starting by trying to make DayJob easier in Excel, people in learning settings using interative notebooks to explore in more detail, people who want to make Alexa tell a fart gag at an inappropriate time to embarrass gran, people adding interaction to a simple web page, ... No one language or environment can do all, or even a fraction of, the everythings that might work for some people.

I, like many others, unfortunately have no idea what to do about it, beyond pontificating as above and hoping that someone else comes up with some bright new ideas (or resurrects some forgotten old ones that were not fully realised/appreciated in their original frame).

With that in mind, it might be more effective to write a BASIC REPL in Javascript (I’m sure one already exists) and then run the original programs.

Why not QuickBASIC or (for modern systems) FreeBASIC[0]? It can be used both like old-school BASIC with line numbers and such, or as a modern style with named functions and subroutines, and FreeBASIC is FOSS.

[0] https://www.freebasic.net/

> People try to tell me that dream is dead. God I hope not

I make no claims, but I think you might find this interesting...

https://asciinema.org/a/SM9tZ2ZiJofikPRxHCQyqDvRY

Just published this in Hackage (Haskell), it is a very early version,so probably contain a lot of rough edges..

https://hackage.haskell.org/package/spade

Give PureBasic a try

https://www.purebasic.com/

Commercial, but not that expensive. There is a demo you can try.

Free introduction book

http://purearea.net/pb/download/PureBasicBook.pdf

> Now, kids feel like they've accomplished very little

Are you observing kids feeling like this frequently? Where/how are they developing? I think there's a lot of enthusiasm in for example Roblox.

> Now, kids feel like they've accomplished very little, even though they're doing more with less code than we did during our generation.

I haven’t observed that. My nephew has great fun building things in Scratch and is really happy with what he has made even though it’s not some triple A game. I had the exact same experience writing stuff in Basic and knowing it could never be like the Buck Rogers arcade machine I’d played.

Personally I preferred the basic listings to the machine language ones, with the asm listings somewhere in the middle. With the basic listings you could understand what you were typing in as you did it. With ml you’re just a monkey typing in numbers. Sure you could disassemble it at that point (if you had a disassembler), but the barrier to reaching that understanding was so much higher than with basic listings.

My abiding memory of typing in listings on my TI-99/4a is how often I intended to type shift-equals to get '+' but instead hit function-equals, which reset the machine and lost all my work since I last saved (and since it was saving to tape, I didn't save all that frequently...)

I tried but quickly realized that teaching is a different skill than coding itself. i also got overwhelmed with the choices and did not know how to start. Scratch did not really grab his interest. And "dad, teach me how to make a minecraft mod" was not possible. I also did not see the patience necessary to learn it.

I bought a programming magazine in the 1980s and copied a game to my Commodore 128. It didn't work, and I failed to debug it, so I set the magazine aside and starting writing my own games. In retrospect I think it was a great educational lesson. My first attempt to copy/paste code was a complete failure, so I was forced to think up stuff on my own. Man, youngsters nowadays can copy/paste in milliseconds, whereas back in the day it took us hours!

I think the key is not to go too fast. Simply writing simple statements and seeing it work and do stuff is enough. So I found it best to start simple and work your way up. These little games look like great examples to work from.

To be influential, it would help if it were sold with one of the most popular systems. There were only around 1.7 million Apple IIgs systems sold.

The Commodore 64 sold the most systems but so many owners only played games on it that the manual probably wasn't the most influential book on BASIC.

Exempting system manuals, the BASIC Computer Games books were probably it.

1. Port them all to Unreal Engine 5 [0]

2. Something something Hacker News

3. Profit?

[0] https://www.youtube.com/watch?v=WU0gvPcc3jQ

They had some notes around the graphics examples saying that differences in syntax may prevent the code from being parsed successfully, and sure enough, none of those commands were available in the V2 BASIC dialect the C64 had.

Even the manual of the machine only included a single hi-res mode example in an appendix, and dedicated one chapter to sprite handling, but that was it. Yes, I am still angry.

Same here, though my copy is in abysmal condition. Between this book and the 80s era enthusiast magazines that had delivered their code directly on the printed page I learned how to program and not be fearful of programming. I have to say that calling Ahl's book the most influential BASIC book sounds like a very feasible claim.