I was thinking a bit about my developer career circa 2001-2004. I was told in 2002 that code automation would replace me by 2004. It was a dubious claim made by someone who was trying to goad me; when I was more susceptible to goading than I am now. At the time, I was building a somewhat impressive automation capability with web UI in Perl. There was a fair amount of low-level work that had to be done then to create cookies, interact with the database, etc. This was all done in VIM, stored in Subversion. I might have had syntax highlighting, but the LLMs that exist today were not there then.
In 2004, DHH announced Ruby on Rails to the world. With one simple command, Ruby on Rails would generate what probably took me a few months to code. No, I would not have been replaced, but my focus would have shifted from building what should have been boilerplate functionality to building value for my customer. By 2009, Ruby on Rails was poised to dominate the web market.
By 2011, that had all fizzled. JavaScript's rise began with Angular, then React. This gave developers a new set of niche's to fill. It takes a lot of study to get good at any one framework or library. It takes a lot of developers to build and maintain a JavaScript application. (That's a dubious term, but I'm trying to stay simple.) When a framework has a major version update, every-other-year, a virtual new rewrite has to happen. Start-ups turn to JavaScript because it's viewed as the "it" language. When they're spending other-people's money, they feel a need to justify the money by hiring more developers, ignoring Brook's Law. The result are large bloated teams that slowly build value while burning through cash. This is also a bonus when you have a rich customer who doesn't feel a need to focus on ROI...hire a bunch of developers and burn their cash for them.
Developers attempt to counter automation by making things harder than it has to be. They fracture the domain. That's one reason why JavaScript is so complicated. The other is that they are attempting to do what the language was not intended to be used for. Its ergonomics are wrong.
In 2021, a liberal arts major wrote a web application, by herself, that generates serious revenue with Rails after taking a bootcamp. It generates $30K after expenses in 2024. Rail's ergonomics are that straight forward. If I were to use 2024's Rails to build the web UI I mention above, it might take a month; but probably less. It took me six months and it was more brittle than I would have preferred. There are many government agencies that have small teams delivering significant value to their citizens by using Rails.
In 2004, DHH announced Ruby on Rails to the world. With one simple command, Ruby on Rails would generate what probably took me a few months to code. No, I would not have been replaced, but my focus would have shifted from building what should have been boilerplate functionality to building value for my customer. By 2009, Ruby on Rails was poised to dominate the web market.
By 2011, that had all fizzled. JavaScript's rise began with Angular, then React. This gave developers a new set of niche's to fill. It takes a lot of study to get good at any one framework or library. It takes a lot of developers to build and maintain a JavaScript application. (That's a dubious term, but I'm trying to stay simple.) When a framework has a major version update, every-other-year, a virtual new rewrite has to happen. Start-ups turn to JavaScript because it's viewed as the "it" language. When they're spending other-people's money, they feel a need to justify the money by hiring more developers, ignoring Brook's Law. The result are large bloated teams that slowly build value while burning through cash. This is also a bonus when you have a rich customer who doesn't feel a need to focus on ROI...hire a bunch of developers and burn their cash for them.
Developers attempt to counter automation by making things harder than it has to be. They fracture the domain. That's one reason why JavaScript is so complicated. The other is that they are attempting to do what the language was not intended to be used for. Its ergonomics are wrong.
In 2021, a liberal arts major wrote a web application, by herself, that generates serious revenue with Rails after taking a bootcamp. It generates $30K after expenses in 2024. Rail's ergonomics are that straight forward. If I were to use 2024's Rails to build the web UI I mention above, it might take a month; but probably less. It took me six months and it was more brittle than I would have preferred. There are many government agencies that have small teams delivering significant value to their citizens by using Rails.
--
Ben
In tenebra solus sto