Hacker News Reader

Comment by simias | Hacker News Reader

I've been coding for almost 20 years now, I definitely saw many changes in that time. Languages like Java and C# became huge, scripting languages like Python or JavaScript became not-just-scripting-languages. C went from "The English of programing languages" to "The Latin of programing languages". Then you have a newer generation starting to take over some niches with the likes of Go, Rust and a few others.

If anything what bothers me is not the lack of innovation in programming languages per se but the relative lack of innovation in program representation. We have high res monitors with incredible graphical capabilities but we still mostly just represent code as flat text files. There've been a few experiments to change the status quo, I remember the Lighttable editor making a bit of a buzz a few years ago, but nothing really changed overall. I wish that a big company like Google, Apple or Microsoft attempted something bold in that space.

View on HN

> We have high res monitors with incredible graphical capabilities but we still mostly just represent code as flat text files

I think the key issue here is that flat text files remain an universal standard, easily handled with commodity tools and editors, often bundled with the OS itself.

Once in a while someone wants to revolutionize code representation (like Light Table, which you mentioned). "What if we represented code as a... flowchart, a graph of nodes, a city, a pie chart, a mind map, $YOUR_NOVEL_IDEA_HERE? Wouldn't that be great and take advantage of current hardware to the fullest?"

It turns out it's really hard to beat good old flat text at convenience and interoperability with standard tools.

Well one alternative representation really took off, that one where code is is a dataflow graph, whose nodes are all one-liners on a massive grid. I wonder what led to its success, where so many other interesting ideas have failed?

Like LabVIEW the spreadsheet paradigm took off in limited domain and became progressively more maligned as it got used for everything else.

I agree completely, hence why I want a big player that can throw their weight around to try something. You'd basically have to create a whole ecosystem around your fancy new tool.

It's a very tall order for some lone coder in their basement, but it's called Monday for Google.

> I wish that a big company like Google, Apple or Microsoft attempted something bold in that space.

Me too. I tried stirring things up a bit when we were nearing the end of Eve, but nobody bit. This is work that _can_ be done, someone just has to fund it and let it grow without too much interference. The big players certainly have the ability to take care of the first part, not so sure about the second though.

> I remember the Lighttable editor making a bit of a buzz a few years ago

It's hard to believe.. but that was 8(!) years ago now. Time sure does fly.

> C went from "The English of programing languages" to "The Latin of programing languages".

There's still a huge amount of development being done in C.

> If anything what bothers me is not the lack of innovation in programming languages per se but the relative lack of innovation in program representation.

One of the benefits of using text is you get a huge number of tools that work with it, without the things that are par for the course these days like vendor lock-in. It'd be interesting to see new efforts in this area, but there's a huge barrier to this.

> One of the benefits of using text is you get a huge number of tools that work with it, without the things that are par for the course these days like vendor lock-in.

That’s true, but you can store code in flat text files even if you don’t represent or manipulate it as plain text.

>There's still a huge amount of development being done in C.

I know, I do some of it! My point was more that 20 years ago in my experience you couldn't call yourself a coder if you didn't know some C. Nowadays I find that most novel coders know about it but can't really write it "fluently".

Many people still know, study and even use it, but it's not quite as ubiquitous as it once was.

I think it's more like cursive: some people still use it, but fewer than before.

https://www.luna-lang.org/

I'm not sure that it needs to be changed. Math has been represented in a similar fashion for a very very long time.

Computers got fast enough that applications can now be usable and useful despite doing things in massively inefficient ways. That inefficiency is wasteful in terms of CPU cycles but it can also be helpful in terms of development cost as it means developers can spend less time on problems like designing efficient data structures. For example, an inefficient JS application that uses reference passing is much easier to write than an efficient application that requires the developer to manage pointers and memory addressing. That's a good thing for the 99% of software that runs on CPUs that are idle 98% of the time, and a terrible thing if you have a CPU-bound programme.

That inefficiency has a beneficial side effect too - it eliminates whole classes of bugs that come from devs doing hard stuff.

We have some experiments not so much in changing textual representation but more in visual tools to make use of said high res monitors. Nothing concrete to share publicly yet.

I also wrote recently about how GPT-3 being so good with little training at code generation might make for interesting advancement in programming environments: https://blog.repl.it/codingai

>If anything what bothers me is not the lack of innovation in programming languages per se but the relative lack of innovation in program representation.

Exactly, We don't need a new programming language every month. We need a new IDE every month. I believe there is a lot of space for improvements.

I hear what you are saying. Check out BlockStudio:

www.blockstudio.app

It uses extended graphical rewrite rules, so you define the “logic” for your program by manipulating the objects that define your program. I could write more words here, but visit the link and check out the tutorial.

The reason why I bet internet will win because adoption Of its many protocol is much easy as it is using text not any ftam.

Text win - XML win too verbose and we have Json.

Whilst explore is ok (microbit I use the graphic as it is new but switch to text once i know the command). The horrible windows use click and we still use server as it can scripted, seen, edit. Easily.

We look with 3D graphic but please do not opt for a VR based prog language. Some part of the design may be graphic. But mostly text.

Neural networks/teach by example is that program representation paradigm change.

Reply on news.ycombinator.com