2 ms·
You know, I guess these kinds of articles are formulaic, but I enjoy seeing the rationale for switching. There definitely are reasons to sometimes migrate to a
by tombert 7y ago
You know, I guess these kinds of articles are formulaic, but I enjoy seeing the rationale for switching. There definitely are reasons to sometimes migrate to another infrastructure or language (be it cost, ease of hiring, performance, etc), and I like hearing the rationale behind it, even if I don't agree with it.
Going onto the other end of the spectrum, I had a job in 2012 doing ColdFusion, ActionScript, and DOS FoxPro. ActionScript was still semi-relevant, but ColdFusion was barely supported anywhere anymore, and FoxPro (in any version AFAIK) hadn't been updated in at least six years.
Whenever I got annoyed by this, they had this strong attitude of "if it ain't broke, don't fix it", and told me that there's no reason to rewrite anything.
Coldfusion isn't the worst language I've used, but ColdFusion code can be extremely difficult to debug, especially before they introduced CFComponents, and after about six months, I quit. I haven't talked to anyone there in a number of years now, but my understanding is that they really haven't been able to replace anyone who left, since the number of people willing to do ActionScript and Coldfusion and FoxPro is shrinking every day.
Changing to $Hyped_technology has one really good side effect: it attracts talent. Is your site getting 2000 visitors a day really going to benefit from Kubernetes? Is your Android app really going to see an appreciable difference if you rewrite the server in Rust? Is your code really going to have less bugs if your new server it in Haskell? The answer to all these questions is "maybe", but all of those technologies have the advantage of being sexy to people who absolutely love compsci, and being able to attract that kind of talent has value.
I don't know how much of a success story this really is, but I took a job at Jet.com, when that was still relevant, because it was using F#, and that seemed pretty neat. I met some of the most incredibly smart and talented engineers that I've ever worked with there, and I think that's in no small part because a lot of people who are interested in F# do programming for more than "just the money".