motet_a

I love the idea, but the repository search is actually pretty hard to use. It should be able to find quickly `org/repo` or `user/repo`. Moreover, the layout kinda breaks on mobile. Is this website open-source? Some people could (possibly) contribute.

However, I think tweets and blog posts are two serious competitors. How does thankyouopensource.com differentiate from those?

> There is one project structure to rule them all. There MUST be separate directories for source, test, libraries, compiled binaries, etc…

I don’t think so. Look at the Git sources (https://github.com/git/git). Most of the C source files are in the top directory.

Some other people like to write unit test in the same directories as the source files, by adding a `.test` suffix to each source file name.

Some other people like to embed unit tests in source files, “à la” Rust.

There is no One True rule. Not organizing things the way you do is not necessarily bad.

But why do Ethereum people maintain a custom programming language? This could easily add a huge amount of complexity (and security issues) with very few benefits.

Moreover, maintaining a large and critical project without automated tests seems impossible to me.

In France, there is a similar issue with the SNCF. More details in this (French) article:

http://transports.blog.lemonde.fr/2017/05/27/inoui-sncf-jarg...

I don't know if anything similar exists in other coutries, but in France there is a growing "workshop" trend - which are basically short web curriculums (a few weeks learning Rails, jQuery, Symfony or others) - to turn anyone into a "developer". I've heard many salesmen or others saying "I will learn it". They don't realize how hard it is. You can't simply write any program with pyramids of jQuery callbacks. You need automated testing, fuzzy string matching, automata theory, murky x86_64 assembler and much more.

Note that the title is confusing since the article is more related to react-redux than Redux itself.

Redux is too simple to suffer from performance problems.

Yes, Docker is great but learning it can definitely require more than 20 minutes. A `curl ipinfo.io` is still simpler.

IMHO, even "developer" is a stupid job title which shoud be replaced by a more descriptive one like "computer programmer" or "software engineer".

Reading automated tests and doing extensive searches in the codebase (with recursive `grep`s) is often helpful. If you use an IDE with a go-to-definition function, it can replace grep.

I have the overall feeling that most people don’t realize that parser combinators are nothing more that disguized recursive descent parsers. Personally I love both.

Personally, I don't bookmark anything anymore. I have zero bookmarks.

Chrome history and autocomplete are great, and as said before, links rots over time. My memory and Ctrl+H do the work.

I don't have a huge memory, but if I forgot something it is probably not very interesting.

Of course, sometimes, I forget some URL and I spend a little time searching in the history. But it is worth nothing compared to managing thousands of bookmarks.

Chrome looks so nice without the bookmark bar :-)

C is interesting but old and unsafe, leading to memory corruption bugs. These bugs are often hard to find and compromise security. It is not always a good choice in real-world.

Rust, C++ and Haskell are much more complicated than JavaScript IMHO. Lisp is pretty old but there is a lot of modern Lisp-like languages.

Try them.

As cnnsucks said, reading code is important no matter what language you use. You can perfecly use JavaScript, Python, Go or Ruby and learn incredible tricks everyday, and these languages are used in the real-world.

Yes. Complex languages are not required to solve complex problems. This is why Go and JavaScript are so popular.

Moreover, CPython is a interpreter without JIT while GHC compiles to machine code. This is totally different. IMHO, it is much more correct to compare a JIT-enabled interpreter to GHC.

How does the router really work? Large apps often have hundreds of paths with nested/sub/sub/sub/resources. Is it possible to declare subresource path without specifying the parent resource path?

The Wired article was probably bad, but I don't think blaming "the media" is a good idea, since "the media" is often understood as "any media". It creates confusion.

The author has written an article. Somehow the author is also a media.

The media must not manipulate our view of the media.