3 ms·

Billion dollar companies have been built on perl[0]. So let us not seek to fix the blame for the past, let us accept our own responsibility for the future. Ser

Billion dollar companies have been built on perl[0]. So let us not seek to fix the blame for the past, let us accept our own responsibility for the future. Seriously though, how large a collection of programming language ecosystems does the world want or need? How many equivalent stacks, all solving the same problems with more or less the same patterns and concepts? Just a thought: Open source communities could benefit from less dispersion of energy and talent, less endless re-invention of the wheel, less endless "not-invented here syndrome"... [0] https://www.slideshare.net/cmnunes/bookingcom-perl https://www.slideshare.net/cmnunes/bookingcom-perl

dreyfan 5y ago
There’s far more venture capital than innovative ideas and the business problems are rarely hard to solve. Engineers get bored so we have a constant flow of new languages, frameworks, and lately, an overwhelming focus on tooling and infrastructure. At the end of the day, most companies are just generating dynamic HTML after performing some rudimentary CRUD database operations. It’s disappointing so much time and money is spent only to arrive at the same place that /cgi-bin/ and a perl script solved over two decades ago.
mywittyname 5y ago
>how large a collection of programming language ecosystems does the world want or need? There are many problem domains that can be solved by computers, and having a language specific or well suited to that niche can be a productivity boon. As long as humans continue to discover new problem domains in computing, they will continue to develop new programming languages to work in them more efficiently (and existing languages will expand to try and fill that niche).
- streamofdigits 5y ago
  yes, I am not arguing for a single language. But many of the domains you mention are actually intrinsically similar once you map them to the underlying information flows. This explains for example the extreme ubiquity and longevity of things like relational databases and SQL. There are several genuine and orthogonal domain "dimensions", e.g real time systems or parallel processing etc, but I would say that is a handful, not the multiple dozens even hundreds of idioms we use to program digital devices...
Hermitian909 5y ago
The open source ecosystem is concentrated. There are far more OSS packages for javascript or C++ than, say, Rust. Less popular programming languages are talked about much more than they are contributed to, but they're talked about precisely because they're often solving problems that are major pain points in people's lives. I'm personally rooting for OCaml, assuming some current efforts bear fruit I'd love a language with the following properties: - An extremely expressive type system with type inference - Incremental builds so compiling is fast - Good compile to JS story so frontend/backend share types as well as serializers and deserializers - Only slightly slower than C (think Java perf) - Good multicore, perhaps even with typed effects - Can be run in a unikernel so I can ship my program as a simple image Any language with this feature list would solve a lot of personal and organizational problems for me and many other programmers I know. None of the really popular languages I'm familiar with can achieve these goals due to backward compatibility requirements.

dreyfan 5y ago

There’s far more venture capital than innovative ideas and the business problems are rarely hard to solve. Engineers get bored so we have a constant flow of new languages, frameworks, and lately, an overwhelming focus on tooling and infrastructure. At the end of the day, most companies are just generating dynamic HTML after performing some rudimentary CRUD database operations. It’s disappointing so much time and money is spent only to arrive at the same place that /cgi-bin/ and a perl script solved over two decades ago.

mywittyname 5y ago

>how large a collection of programming language ecosystems does the world want or need? There are many problem domains that can be solved by computers, and having a language specific or well suited to that niche can be a productivity boon. As long as humans continue to discover new problem domains in computing, they will continue to develop new programming languages to work in them more efficiently (and existing languages will expand to try and fill that niche).

streamofdigits 5y ago
yes, I am not arguing for a single language. But many of the domains you mention are actually intrinsically similar once you map them to the underlying information flows. This explains for example the extreme ubiquity and longevity of things like relational databases and SQL. There are several genuine and orthogonal domain "dimensions", e.g real time systems or parallel processing etc, but I would say that is a handful, not the multiple dozens even hundreds of idioms we use to program digital devices...

Hermitian909 5y ago

The open source ecosystem is concentrated. There are far more OSS packages for javascript or C++ than, say, Rust. Less popular programming languages are talked about much more than they are contributed to, but they're talked about precisely because they're often solving problems that are major pain points in people's lives. I'm personally rooting for OCaml, assuming some current efforts bear fruit I'd love a language with the following properties: - An extremely expressive type system with type inference - Incremental builds so compiling is fast - Good compile to JS story so frontend/backend share types as well as serializers and deserializers - Only slightly slower than C (think Java perf) - Good multicore, perhaps even with typed effects - Can be run in a unikernel so I can ship my program as a simple image Any language with this feature list would solve a lot of personal and organizational problems for me and many other programmers I know. None of the really popular languages I'm familiar with can achieve these goals due to backward compatibility requirements.