The big break in computer languages
esr.ibiblio.org
The big break in computer languages
1–6 of 6 posts
Re: The big break in computer languages
#2I was wondering why OP group Go and Rust together...
After the long post the conclusion from what I take away was OP is sick of managing memory and rather have a garbage collector to do it. API not stablized for Rust is a valid concern but overall all I can hear was wanted GC and barrow checker was too hard.
Re: The big break in computer languages
#3I was wondering why OP group Go and Rust together... After the long post the conclusion from what I take away was OP is sick of managing memory and rather have a garbage collector to do it. API not stablized for Rust is a valid concern but overall all I can hear was wanted GC and barrow checker was too hard.
Eric Raymond is not a fan of Rust. See "Rust severely disappoints me"(http://esr.ibiblio.org/?p=7294). On one level, he's clearly right. Go was designed to be an simple language that is easy to pick up. Rust had a very different design space. On the other hand, when picking a language for your long-term project, I think there are other factors beyond how proficient you can be in your first few days that should come into play.
Personally, I'm not very bullish on Golang moving into the lowest levels of the stack (the space currently dominated by C and C++). It just doesn't seem like an area where the language designers of Golang are interesting in competing. Swift, on the other hand, definitely has aspirations in this area and it is actively being developed by a company that writes a lot of low-level system software.
I think Go, Swift, and Rust each are interesting in their own way and see them jostling to become the next generation of industry standard languages in the next 10 years.
Re: The big break in computer languages
#4>The only consolation you get, if it is one, is that the C++ fans are screwed worse than you are. At least Rust has a real prospect of dramatically lowering downstream defect rates relative to C anywhere it’s not crowded out by Go; C++ doesn’t have that.
Except, C++ is updated frequently and is constantly addressing these concerns. With every update C++ is becoming safer and safer. C++ is a good bet. It's a language with an upgrade path vs throwing all your code away and rewriting it in the hottest new programming language.
If you accept the cost of a GC then C++ might look bad but there are already hundreds of alternatives to C++. If a gc is unacceptable then only C, C++ and Rust remain.
Re: The big break in computer languages
#5Are we all just pretending that Ada doesn't exist now? It's type safe, thread safe, memory safe, overrun safe, a whole bunch of other sorts of safe. It was designed to accommodate huge projects. It has been used in major projects. It has an open source toolchain, and the unusual feature of a freely available ISO standard. And it has been around as long as C++. Yet not a word about it anywhere in the article.
Re: The big break in computer languages
#6Are we all just pretending that Ada doesn't exist now? It's type safe, thread safe, memory safe, overrun safe, a whole bunch of other sorts of safe. It was designed to accommodate huge projects. It has been used in major projects. It has an open source toolchain, and the unusual feature of a freely available ISO standard. And it has been around as long as C++. Yet not a word about it anywhere in the article.
Ada is proof that language selection is political or social, not based on technical merit. My opinion is Ada is a viable candidate language for use today, if you care about what it costs to develop or how long it takes. For over 20 years, the ada compiler and run time have been minimizing defects to a greater degree than any other language / run time. (Note: In my world, C++ has banished ada to the trash heap of 'legacy code', along with ton's of fortran. The C++ replacements are rarely equal; in perfection or in supportability)