Hi I thought it could be cool to create a domainspecific programming languange for manipulation of geospatial data. It would be a scripting language where
-
geometries are standard types
Polygon myPolygon = Polygon ((( .... // wkt syntax -
you could manipulate geometric data using default operators
Polygon myClippedPoly = myPoly & clipperPolygon;
Polygon myUnionPoly = myPoly | myOtherPoly;
and similarly for rasterdata:Raster[1] raster3 = raster1 + 10 * raster2; -
easily access common geometry formats (read/write) shapefiles, tabfiles, postgis, wfs, geojson, geotiff
-
easily view data on a map
obviously it would have common programming language features such as functions, loops etc. common types like integers, floats, booleans and strings.
Do you have any cool ideas what such a language should be able to do? Any other thoughts?
Note that I'm talking about the original versions of COBOL, something like COBOL-85, and not the modern versions. I think it is a DSL, for the following reasons:
-
"COBOL" literally stands for COmmon Business-Oriented Language. And something that is <a particular field>-oriented is a DSL;
-
It was designed by people who came from commerce or government (not Computer Science) for developing business applications;
-
It's less powerful than general-purpose programming languages (for example, it doesn't support recursion and function arguments); also, it doesn't have scope, so all variables are global;
-
Its syntax is English-like (it even has verbs, clauses, sentences, paragraphs, etc), designed specifically for non-programmers. It's also very verbose, which is also a consequence of the fact that it's targeting non-programmers, those who work in business;
-
It doesn't really have a type system. Instead, it uses a "declaration by example" strategy, by using literally
PICTUREs. As a consequence, it doesn't support user-defined types (except for structures, which are just a bunch of non-scoped variables anyway); -
I haven't found any mention of compilers written in COBOL. Provided that writing a compiler back then was quite an undertaking, it probably shows that it was impossible to write one in COBOL, which proves that it's not suited for a wide range of application domains;
-
This says "Different Programming Languages have constructs to support facilities needed for solving problems in specific domains. For example, COBOL has support for solving business related problems, C for systems programming problems [so is C a DSL now?!]...";
On the other hand, it probably isn't:
- This comment by an expert in programming language design says it's not (they also say this here). On the other hand, neither Java, nor C were designed for usage in a specific field, so, that COBOL is in some ways similar to them doesn't make it general-purpose, does it?
- This lists COBOL as a general-purpose programming language, along with C and C++. The commenter in the previous point works for this company, so it's probably the exact same argument;
- [This](https://www.intentsoft.com/inten
I was just thinking about how learning works, and a big part of it that isn't touched on is acquisition of skills.
The common misconception is that thought and movement are directly connected, but there's a reason why watching football doesn't make you better at it. When you catch a ball, navigate through your cluttered house, or flip a bottle (kill me), that is coming straight from your "motor neurons".
Now, this isn't a lesson on the motor cortex, especially since it's mostly theory based on what I know. The question is, how would the AI store these movement patterns? Running from text and natural language understanding is highly inefficient, and using the same programming language for everything couldn't cover every paradigm.
Another huge supporting factor for this is the way the brain stores information: in chunks. Like an infinitely compressing tree of files, and we "unpack" these in our working memory (not a capacity of 7 bytes or 7 digits; 7 chunks). Think of it like extracting a language's source code from a ZIP file and building it.
Perhaps the primary obstacle for Strong AI is our rigid psychological fixation on using a single programming language for the whole thing, but no one language can do everything. A lot of languages are logic-based and determinstic. How are you going to create emotions with that? Speaking of which, any ideas on how an emotion-oriented programming language would work?
TL;DR - Our body utilizes so many different cell types, chemicals, nutrients, molecules, etc so why shouldn't we?
Ive read the "book" and rust in action and really loved the idea of exploring rust while writing software for a certain domain. Even thought there are some amazing books that go more in depth into rust such as rust for rusteceans, I would like to learn those topics while developing a whole application. Are there any high quality books that explore a domain and build something with rust as its language of choice?
After researching a bit I'm considering these books:
- Zero to production in rust
- Black hat rust
- Hands on rust: Effective Learning through 2D Game Development and Play
From the listed books I would start with zero to prod since Ive worked with back-end development with other languages. Any book recommendations or review of the books listed above would be appreciated.
Noob question here that I’ve tried researching to no avail.
I’ve heard Solidity talked about for Solana (and Etherium) and Haskell for Cardano, but read once that Python could be used broadly…. I think that each one of them have good use cases but being a novice I’m concerned that I may pigeonhole myself into one language, unlike a developer who could more easily learn new languages.
Of course JavaScript is use too right? And there are a bunch of frameworks built on it like Mongo, Angular, Node, React ….
I have taking various short online courses for a few languages a few years ago (Python, JavaScript, PHP), but would have to redo them since it’s been a while.
Where should I start?
"Use LISP."
"What'th the eathietht way to parthe a domain thpethific language?"
(Apologies to people with speech impediments)
(A riff on Tim Vine style joke)
I believe that the intersection of python programmers and hydroponic plant growing enthusiasts is not empty. Therefore I present my hobby project with the hope that it can be useful for the community.
This simple Python package provides handy tools for managing hydroponic nutrient solutions and utilizes the power of mathematical optimization to make your hobby even more enjoyable!
https://github.com/dstrelnikov/hydrosolver
When I read books focused on diving deep into a specific programming language, I find all of them to be boring. I would rather be building something with the language and find information in the docs as needed.
I feel like reading books about a specific language is like being a carpenter and reading a book that tells you everything about a hammer; what materials it's made out of, how to use it, etc. I know some people enjoy that, and I see how it can be useful, but it is too boring for me to remain focused on that type of material.
The books that I find interesting are the ones about algorithms and how they are applied to solve problems.
What important things am I missing out on from not reading books that focus on a specific language that I would not learn from reading books about and building projects?
this is my first rust project, there is a lot of code to improve, but i hope you like it! https://github.com/Alonely0/Voila
Hey, Fish-people!
I made this little tool for myself that lets you (dangerously) load and unload functions based on the directory you are in in your shell. It's great for complicated one-liners you'd usually have to memorize on a per-project basis. You can think of it as a mini domain-specific language for your project!
I found it useful in a ton of projects I've been working on. I figured that the Fish community might be interested in it as well.
It's installable with fisher via fisher install loqwai/dangerload-fish
This is the first Fish plugin I've published, and as you might notice by the exclamation points, I'm pretty excited about it! Any feedback is very welcome.
