Hacker News Reader

Comment by amadeuspagel | Hacker News Reader

I'm confused by this insistence on visual programming. Why not teach students javascript and SQL?

Context: A history professor wants her (presumably history) students to build web sites to explain history research to the public. If good-enough-for-that-job visual programming tools are available, then teaching them javascript and SQL would, at best, be a time-sucking side quest.

Put another way, this is an example of people in tech seeing computers as the end to a means while everyone else rightfully sees computers as a means to an end (aka a tool).

It's natural for the CS professor to instinctively teach how computers work first, because the computer is the goal. The inner workings of computers are the means to an end: "How do I make a computer say 'hello world'?"

But for everyone else like the history professor, they want to know how and why making a computer say "hello world" can be useful to them and their goals. The computer is simply just another tool that might help them achieve something, and nobody cares about how any of it specifically works.

Yeah, I'm the one who sees tech an "end to a means" for thinking students should learn boring technology[1], while CS professors who use students to try out their custom-made visual programming kit are the ones who "rightfully see computers as a means to an end".

[1]: https://mcfunley.com/choose-boring-technology

The history professor doesn't want to teach what a website is, she wants to teach why you might want a website; in this case to present information in a database. Therefore, how you make that website is completely irrelevant; the website is not the end here, it is a means.

You, on the other hand, are asking why she isn't teaching what a website is. So yes, you're seeing computers as the end to a means while the history professor is asking how computers could be a means to an end.

Does "boring technology" in a university context mean Sqlite and JS, which the students have never seen before and which the university doesn't use for teaching introductory courses? Or does it mean the existing education tool Snap which is already in use and familiar to students?

I bet more Freshmen already know basic SQL and JS than Snap. I know some primary/highschools teach Snap, but a lot more must be teaching JS. SQL as well, I know at least some highschools teach the basics and tbqh basic SQL queries are basically requests phrased in English. And it should be easier than ever now that LLMs are pretty good at taking requests actually phrased in English and turning them into valid queries. You don't need to learn relational algebra to learn 'select name from patients where age > 50;'

Did you know in the 70s and 80s, common undergrads (not CS students) used to program Fortran on punchcards? My mother was an undergrad biology student, when she learned I was interested in programming she pulled a dusty box of punchcards out of the attic to show me programs she had written. If colleges are using Snap instead of asking students to write some basic SQL queries, I think they're damning those students with faint expectations.

Not only that, but the history professor is gearing the goals towards particular professional goals and using the students to further them. I.e: "What do you want your humanities students to know about computers? How to help me professionally."

If any of those students in humanities need to do data science later for their own careers how does building websites with Snap help them?

Why not use the lingua franca tools of the university website.

You're wildly mischaracterizing and misunderstanding where the people who developed and inspired Snap! are coming from, and what their background and experience in teaching computer science and other subjects to kids and college students really is.

If you'll please take the time to read some of the things they've written, especially the The Beauty and Joy of Computing, and the long history of Scratch and Logo and Scheme that led to and inspired the development of Snap!, you'd probably find that you agree with a lot of things they say.

https://en.wikipedia.org/wiki/Brian_Harvey_(lecturer)

https://bjc.berkeley.edu/team/snap/

https://snap.berkeley.edu/research

https://www.snapcon.org/users/5

https://en.wikipedia.org/wiki/Cynthia_Solomon

https://news.ycombinator.com/item?id=36755836

Seymour Papert, introducing "The Turtle":

https://www.youtube.com/watch?v=-V_OPfmbbCk

https://www.youtube.com/watch?v=fTO-Ruby-Uo

https://www.youtube.com/watch?v=CJlRGe5QGhs

Also thanks to Lars Brinkhoff's research: Here's a video uploaded by Cynthia Solomon. Seems legit.

https://www.youtube.com/watch?v=c4kMzrDr4jQ

Definitely check out the rest of Cynthia Solomon's youtube video treasure trove, with lots of great stuff by Marvin and Margaret Minsky, Seymour Papert, and others from MIT and Atari Cambridge Research:

https://www.youtube.com/user/cynthiaso/videos

Seymour Papert on Logo, Turtles and Giraffes:

https://www.youtube.com/watch?v=maDzjHIiXZc

https://www.youtube.com/watch?v=lDyym_9-E-g

https://www.youtube.com/watch?v=ha8sTgtUejM

A gestural programming system developed by Margaret Minsky, Danny Hillis, Daniel Huttenlocher, David Wallace (Gumby), and Radia Perlman at the MIT-AI Lab:

https://www.youtube.com/watch?v=-Wq6SQTVM9M

Marvin Minsky demonstrating a Logo Machine with an acoustic modem and cassette tape, talking about education theory, and showing a part of his first "thinking" machine: a simulated nerve synapse (1 of 40) with an adjustable knob that he built in 1951 out of WW-II surplus hardware, and discussing playing with Tinker Toys as a child:

https://www.youtube.com/watch?v=c4kMzrDr4jQ

https://www.youtube.com/watch?v=S72xF3gd-mI

https://www.youtube.com/watch?v=yZRQQl8mA0c

https://www.youtube.com/watch?v=dfKRNHRyD64

David Levitt demonstrating his musical improvisation software:

https://www.youtube.com/watch?v=ocwsVkqEKys

List of Logo memos from MIT AI Lab:

https://github.com/pdp11/sits/wiki/List-of-Logo-memos

Brian Harvey’s Personal Narrative on Snap!: Scheme Disguised as Scratch

https://escholarship.org/uc/item/1623m1p3

>In 2009, the University of California, Berkeley, was one of several universities developing a new kind of introductory computer science course, meant for non-CS majors, to include aspects of the social implications of computing along with the programming content. Scratch wasn’t quite expressive enough to support such a course (it lacked the ability to write recursive functions), soProf. Daniel Garcia and I thought “What’s the smallest change we could make to Scratch to make it usable in our course?” After 20 years teaching Structure and Interpretation of Computer Programs [Abelson et al.1984], the best computer science text ever written, I knew that the answer to “what’s the smallest change” is generally “add lambda.” I joined forces with German programmer Jens Mönig, who had developed BYOB (Build Your Own Blocks), an extension to Scratch with custom (user-defined) blocks, including reporters and predicates. [...]

I'm not criticising snap, I only question whether it's a good way to make websites.

What do you mean by characterizing Snap! and its HTML extensions as a "custom-made visual programming kit"?

How is it "custom-made", and what's wrong with that?

The HTML extensions to Snap! are essentially just block based macros.

Are Lisp macros a "custom-made domain specific language", therefore it's bad to make websites with Lisp if you use macros because they're "custom-made"?

Paul Graham would probably disagree. What's wrong with using "any language you want", what's so "custom-made" about Snap!, and why is that a problem?

https://news.ycombinator.com/item?id=606619

https://ep.turbifycdn.com/ty/cdn/paulgraham/bbnexcerpts.txt

Lisp in Web-Based Applications

Paul Graham

(This is an excerpt of a talk given at BBN Labs in Cambridge, MA, in April 2001.)

Any Language You Want

One of the reasons to use Lisp in writing Web-based applications is that you can use Lisp. When you're writing software that is only going to run on your own servers, you can use whatever language you want.

[...]

> What do you mean by characterizing Snap! and its HTML extensions as a "custom-made visual programming kit"?

This is what I mean by custom-made:

> Undergraduate Fuchun Wang created a great set of blocks explicitly designed to look like SQL for manipulating CSV files. We used these blocks to talk about queries and database design in the class.

To me the fact that someone had to create this for this class indicates that this is not a common and well-supported way to build database-backed websites.

> Are Lisp macros a "custom-made domain specific language", therefore it's bad to make websites with Lisp if you use macros because they're "custom-made"?

I'm not saying that making websites with any tech is bad, but I think students should learn how to make websites using standard tech. And that's how I understood the goal here: My students build websites. And maybe a freelance web developer would be a better teacher then a CS professor.

You're moving the goalposts. You said "custom-made visual programming kit" not "custom-made functions in a visual programming language".

It's strange that you would object to writing "custom made" functions in a visual programming language, while you seem just fine with writing functions in JavaScript. Why are "custom made JavaScript functions" ok but "custom made Snap! blocks" aren't?

It sounds to me like you originally misunderstood that Snap! has already existed for many years, and that you could "build your own blocks" with it (which was its original name: BYOB), and mistakenly thought the teachers wrote their own entire block based visual programming language just to generate html. And now you're trying to move the goalposts when it's clear that simply isn't true, because Snap! and the other block and text based programming languages that inspired (like Scratch, Logo, and Scheme) have decades of experience and widespread use in education.

And you're also totally missing the point that this is about a multidisciplinary university, not a trade school for pumping out code monkeys and search engine optimization "experts".

People definitely use this idea of choosing boring technology to justify the use of visual languages and/or domain-specific tools when building systems, instead of general-purpose programming. Lots of people might say go for FileMaker or Access over writing your own lower-level JS & SQL, if you want to generate a site or use a database to generate any kind of automated report or output.

To be honest, these types of posts come off with a level of smugness that I don't have words for. It's this holier-than-thou attitude that prevents us from getting people to use their computers better. Yes, there are 1,000 better options than a visual programming language. But to your point about "using boring technology" the professor chose the most boring technology available that will convey the message effectively. When considering an audience of Non-CS, possibly non-technical, people you will look like a jackass trying to tell them life will be better learning something more complicated. Not everyone wants to spend 6-12 months learning something just to start what they want.

It may humble you to go volunteer to teach adults how to program. Once you're outside the bubble of technically minded people it becomes very obvious how intractable the problem can become. If you're smart, you will arrive at the same conclusion the professor did. In some cases it's often better to start people off with PlaySkool My First Language than immediately plunging them into the depths of language hell. Especially if the goal is simply to improve their productivity.

An example from my personal life. I have taught business people how to think about programming. I generally will sit them down and teach them about Excel, Excel macros, etc. I don't use these in my daily life. I might prefer R, Pandas, or something else. If I recommended those to a non-technical person they would learn nothing. But if I teach them something they can click, play with, etc they will not only be more effective but perhaps eventually be inspired to spread their wings a little. It's not a hit to my ego and I don't look down on them.

> It may humble you to go volunteer to teach adults how to program.

Indeed I've never done that. I would have thought that college students are young and intellegent enough to learn to code, but if that's not the case, and they're supposed to learn how to make websites, then they should learn wordpress, or some other CMS that's actually used.

> An example from my personal life. I have taught business people how to think about programming. I generally will sit them down and teach them about Excel, Excel macros, etc.

Yes, Excel is a tool that's used in the real world, and that's the kind of tool people should learn.

We can reasonably presume these students are young and intelligent, but whether they have any interest in becoming computer programmers is a different matter entirely.

Remember, these are liberal arts students; their focus of learning is going down a completely different path from computer science or anything to do with engineering.

As the history professor wants to show her students, a computer can be used to construct a database of information and a website can be used to present that information for subsequent use. This is for furthering their subject of liberal arts, none of them care exactly how computers or websites work because it doesn't matter to them.

So long as they have a rudimentary understanding of what a database is, what a website is, and how they can be used for their goals of studying liberal arts, that's all that matters. They will likely just hire someone with the professional expertise to actually handle the database and website itself later in their careers, because their job is bigger than being a webmaster or a programmer.

It's like how the CEO of a company will want to know how a website could further his business endeavours, but wouldn't necessarily be interested in knowing what a website is because his job is running a company and not a web server.

Context: These visual programming tools were not in fact available but had to be custom-made. EDIT: This is what custom-made refers to:

> Undergraduate Fuchun Wang created a great set of blocks explicitly designed to look like SQL for manipulating CSV files. We used these blocks to talk about queries and database design in the class.

If you want to learn something, you need lots of examples. You have that with javascript and SQL, and today you even have ChatGPT.

If you want to deploy websites you need something well-supported.

Snap! is hardly a "custom made" visual programming language -- version 9 was just released, it's been widely available for many years, many people and schools are using it, many papers have been written about it, there are many extension packages and forks for network, game, AI, and embedded device programming, even controlling embroidery machines and robots, and much thought and experimentation and learning from experience in the real world of college and high school classrooms has gone into it:

https://forum.snap.berkeley.edu/t/whats-new-in-snap-9/14534

I've been writing about Snap! on Hacker News for years, including just recently, comparing it to Logo and Scheme and Scratch:

https://news.ycombinator.com/item?id=36755852

Here's the discussion of Version 5 from four years ago, and it's come a long way since then:

https://news.ycombinator.com/item?id=20309162

There's a whole book of computer science curriculum that's being taught at UCB and many other high schools and universities:

https://snap.berkeley.edu/bjc

>The Beauty and Joy of Computing

>The Beauty and Joy of Computing (BJC) is an introductory computer science curriculum using Snap!, developed at the University of California, Berkeley and Education Development Center, Inc., intended for non-CS majors at the high school junior through undergraduate freshman level. It is a College Board-endorsed AP CS Principles course. It is offered as CS10 at Berkeley.

Your mischaracterization of Snap! as "not in fact available but had to be custom-made" is only true in the narrowest sense that every programming language, including JavaScript and SQL, were once not in fact available but had to be custom made.

This is the custom-made part:

> Undergraduate Fuchun Wang created a great set of blocks explicitly designed to look like SQL for manipulating CSV files. We used these blocks to talk about queries and database design in the class.

So is writing a JavaScript function or library the "custom-made" part of developing a web application? What's wrong with that?

>If you want to deploy websites you need something well-supported.

That's not what Paul Graham believes.

https://ep.turbifycdn.com/ty/cdn/paulgraham/bbnexcerpts.txt

I assume because it’s faster, easier, and requires less expertise to get started with a Visual Programming Language (VPL). This is why people use Scratch to introduce kids to programming, so maybe it makes for humanities students? But people have certainly been using VPLs for a long time for web site generators, for business automation, CRUD apps, point-of-sale systems, etc. There are lots of different goals people talk about with VPLs, including simplifying maintenance, limiting technical debt and over-engineering, having the ability to hire and employ people who have non-CS expertise, or just don’t have a CS degree, reducing training time, etc.

What do you mean by “insistence”, are you referring to this article or the broader ongoing existence of visual programming languages?

The vast majority of actual software engineers can't even write JS and SQL right.

It would make no sense to send humanities students on a wild goose chase trying to learn two harder languages when visual programming delivers the message just as effectively.

We often get lost in this industry thinking anyone can do what we do. We even see it with this new "ChatGPT is going to take your job" non-sense. The truth is what we do, when we do it well, is hard. If the goal is to deliver a greater point and use some programming construct as a conduit you can't get simpler and more effective than something visual.

Because Javascript and SQL are not the basic principles needed to get started. Logic, algorithms, loops, functions, variables, and other such things are. So you learn the basics with whatever tool is easiest, and then once you have those down, apply them when learning a new language.

I think that makes even less sense the visual programming.

The final project of the course is a static web page. Databases need not be involved at all, much less JS code. Why not just HTML and CSS?

It's easier to get started with flowcharts. Sometimes the biggest obstacle is getting started itself.

Reply on news.ycombinator.com