3 ms·

i dont think this would apply as much in the modern context of package driven development . Nobody is delivering value as a SE without some package manager toda

i dont think this would apply as much in the modern context of package driven development . Nobody is delivering value as a SE without some package manager today . we are so far beyond the discussion if its a good idea to write your own tools for doing http request . And large scale adoption of the modern open source ecosystem was only possible after 2013 . Also to keep in mind that we are in a post moore's law era where people are relearning how to make things more performant.

mrkeen 4y ago
When I first heard of the term "web application", I started trying to download/install and play with GlassFish and Tomcat and JBoss, etc. When one seemed too complicated I'd jump to the next. There were jars and wars and admin panels and endless configuration. Total mind-fuck. I never got the hang of them. The next time I visited web apps, Jetty was around and seemed way smoother in comparison. Microsoft's IIS used to have an outrageous footprint (if memory serves, you needed a whole computer to host a domain, or something like that). So Apache took off. Then virtual machines. But even they were too bloated, so Docker took off. And the BSD people Likewise with Spring. It was huge and super-powered! By the time I got the hang of it, it had already become Spring Boot, shedding a bunch of the Enterprise baggage (especially xml) and trying to become leaner and meaner. I was completely blown away when I saw https://sparkjava.com/ https://sparkjava.com/ and once again when I saw https://learnbchs.org/easy.html https://learnbchs.org/easy.html Now when I code web apps (or "microservices" these days), my mindset is that http is just a function from request to response, and I feel so much better for it! > we are so far beyond the discussion if its a good idea to write your own tools for doing http request In short, I've become skeptical of the value-adds of gigantic software artifacts, as well as existing ways of doing things. I can't wait for more shitty little side-projects to displace bloated incumbents. > And large scale adoption of the modern open source ecosystem was only possible after 2013 . What happened in 2013? It looks like CPAN's been doing its thing since 1995.
- dhab 4y ago
  How do you know there are less vulnerabilities in a smaller package than there are in a better adopted one(usu bigger)? And isn’t it more likely that big packages (they got big I feel due to expanding user-base requirements) get the patch sooner?
  - mrkeen 4y ago
    You can't have vulnerabilities in code which isn't there. Compared to what I could be using, my ideal stack does not have: class loaders, virtual machines, dependency injection containers, web application servers, Log4J, etc.
- elisharobinson 4y ago
  what PG said about time to market is most part of the article IMO. what is the time to market, how long does it take to cold start a project from zero to mvp. sadly our ideals as SE might not be able to match that reality always. As much as we like to pretend that we know what is the best stack , at the end of the day it is the market / customer / client who votes with their money bloat , lean , fast , slow , typed , etc none of it matters if nobody buys your solution. in 2013 is when there was a huge shift in enterprise to be more open OSS until then enterprise would only use solutions from the catalog of either oracle or microsoft .

mrkeen 4y ago

When I first heard of the term "web application", I started trying to download/install and play with GlassFish and Tomcat and JBoss, etc. When one seemed too complicated I'd jump to the next. There were jars and wars and admin panels and endless configuration. Total mind-fuck. I never got the hang of them. The next time I visited web apps, Jetty was around and seemed way smoother in comparison. Microsoft's IIS used to have an outrageous footprint (if memory serves, you needed a whole computer to host a domain, or something like that). So Apache took off. Then virtual machines. But even they were too bloated, so Docker took off. And the BSD people Likewise with Spring. It was huge and super-powered! By the time I got the hang of it, it had already become Spring Boot, shedding a bunch of the Enterprise baggage (especially xml) and trying to become leaner and meaner. I was completely blown away when I saw https://sparkjava.com/ https://sparkjava.com/ and once again when I saw https://learnbchs.org/easy.html https://learnbchs.org/easy.html Now when I code web apps (or "microservices" these days), my mindset is that http is just a function from request to response, and I feel so much better for it! > we are so far beyond the discussion if its a good idea to write your own tools for doing http request In short, I've become skeptical of the value-adds of gigantic software artifacts, as well as existing ways of doing things. I can't wait for more shitty little side-projects to displace bloated incumbents. > And large scale adoption of the modern open source ecosystem was only possible after 2013 . What happened in 2013? It looks like CPAN's been doing its thing since 1995.

dhab 4y ago
How do you know there are less vulnerabilities in a smaller package than there are in a better adopted one(usu bigger)? And isn’t it more likely that big packages (they got big I feel due to expanding user-base requirements) get the patch sooner?
- mrkeen 4y ago
  You can't have vulnerabilities in code which isn't there. Compared to what I could be using, my ideal stack does not have: class loaders, virtual machines, dependency injection containers, web application servers, Log4J, etc.
elisharobinson 4y ago
what PG said about time to market is most part of the article IMO. what is the time to market, how long does it take to cold start a project from zero to mvp. sadly our ideals as SE might not be able to match that reality always. As much as we like to pretend that we know what is the best stack , at the end of the day it is the market / customer / client who votes with their money bloat , lean , fast , slow , typed , etc none of it matters if nobody buys your solution. in 2013 is when there was a huge shift in enterprise to be more open OSS until then enterprise would only use solutions from the catalog of either oracle or microsoft .