I Miss Rails • Hacker News Board

Re: I Miss Rails

#451

Post by tombert » Fri, Mar 29, 2019, 3:37 PM UTC

Kalium wrote: post #448

tombert wrote: post #447

Earlier quoted context omitted.

I threw Clojure in as an example, but my point is that if I avoid the ceremony, and stick to a few slightly-lower-level libraries, I end up with a lot less code that I feel was overall, easier to write and maintain.

Again, you're completely right. You will get something easier to fit into your head. Is it perhaps possible that there might be some room to question if the goal of something that fits neatly it one person's head is really the best one? For professional applications, some might opine that there could be other considerations. Certainly most of my comments about Clojure apply to any tools that work from the same mindse…

Sure, it's possible that my mentality is totally wrong, and that having a large amount of cruft is necessary for a lot of big projects. I even built my own framework to experiment with these ideas [1][2] (forewarned that I think I swore a lot in the codebase and READMEs if that's a problem).

I can see that there's probably a case to be made for "forced-best-practices", which a lot of frameworks kind of entail, and I think that if you don't trust yourself (or coworkers) to write quality code, then having the guardrails for that Rails and ASP.NET MVC give you probably feel nice.

However, I think a case can be made that these same guardrails can feel restrictive and noisy to more experienced developers (or people with large-enough egos like me). I have no desire to ever touch J2EE ever again in my life, no matter how much I could get paid doing it, because it felt like getting anything done was about as much fun as filing my taxes.

Rails is certainly not as terrible as J2EE, so I'm not trying to draw a direct equivalence, and maybe a balance can be struck...Personally, I find that the "give me a server and I'll set up some middleware on top" approach like Express (for Node), and the aforementioned Ring/HTTP-Kit for Clojure give me the best balance.

Clearly there's a middle-ground -- I certainly wouldn't suggest that most people go and reimplement TCP and HTTP from scratch for a project, unless they're severely masochistic, or they just want to learn more about TCP or HTTP.

[1] https://gitlab.com/tombert/frameworkeyPromiseEdition [2] https://gitlab.com/tombert/Frameworkey-Erlang

Re: I Miss Rails

#452

Post by vbsteven » Fri, Mar 29, 2019, 3:39 PM UTC

ethagnawl wrote: post #433

m23khan wrote: post #409

Earlier quoted context omitted.

+1 - Kotlin/Java with Spring is a totally different beast as compared to java apps of yester years.

This is a good point. I haven't used Spring in forever, so I don't know if/how it takes advantage of any modern Java features, but I could see Java 8/9/10 and (a framework like) Spring being pleasant and productive.

The Spring and Spring Boot teams have done a lot of work on this. IIRC Spring Boot 2.1 is compatible up until Java 11, and Java 12 support should be in 2.2.

Some API's are adapted to take advantage of modern Java features like for example JPA repositories that can return Optional for single row queries and Stream for multi row queries. Some API's have been improved with functional equivalents.

If you combine this with Kotlin you get all the power of the Spring framework together with excellent nullability checks and modern language features that Kotlin brings to the table.

Re: I Miss Rails

#453

Post by dlandis » Fri, Mar 29, 2019, 3:44 PM UTC

chasd00 wrote: post #425

gambler wrote: post #398

Earlier quoted context omitted.

>Server-rendered used to mean slow and clunky I don't get why everyone keeps saying this. Is it because of WordPress? In 2005 I used my own MVC framework with ORM and stuff. It was written in PHP. My pages had the usual generation time of 0.03 seconds. And that was with no specialized caching or any crazy optimizations. Even with no AJAX, page loading seemed instant, unless you connected to the server from across the…

the lighter text means the post has been down voted correct? Why was the parent post downvoted then? They make a perfect point, there's nothing inherently slow about traditional web apps. There's definitely things regarding the UX that traditional server side web applications can't do vs browser side but they can be incredibly fast. one nit pick, what does "server-side rendering", and even "client-side rendering", ev…

> Does "rendering" really mean template parsing?

It's kind of a misnomer; it usually means generating a "renderable" thing (e.g. HTML) , not actually rendering it on the screen because yes the browser does that.

Re: I Miss Rails

#454

Post by Kalium » Fri, Mar 29, 2019, 3:47 PM UTC

tombert wrote: post #451

Kalium wrote: post #448

Earlier quoted context omitted.

Again, you're completely right. You will get something easier to fit into your head. Is it perhaps possible that there might be some room to question if the goal of something that fits neatly it one person's head is really the best one? For professional applications, some might opine that there could be other considerations. Certainly most of my comments about Clojure apply to any tools that work from the same mindse…

Sure, it's possible that my mentality is totally wrong, and that having a large amount of cruft is necessary for a lot of big projects. I even built my own framework to experiment with these ideas [1][2] (forewarned that I think I swore a lot in the codebase and READMEs if that's a problem). I can see that there's probably a case to be made for "forced-best-practices", which a lot of frameworks kind of entail, and I…

I've personally watched experienced, mature, and very senior developers build sophisticated systems with Clojure, Ring, and http-kit. And blink at me owlishly when I ask them how they sanitize their inputs to guard against SQL injection. They'd been intensely focused on the core logic, architecture, system design, and business functions. Everything else received cursury consideration.

My takeaway from that is that it's unreasonable to expect even the best of developers to remember every detail every time. Even the best forget or make mistakes. If our tools don't take this into account and protect us from ourselves, we're going to get burned.

Re: I Miss Rails

#455

Post by clairity » Fri, Mar 29, 2019, 3:53 PM UTC

tomaskafka wrote: post #395

clairity wrote: post #291

Earlier quoted context omitted.

> “...JS developed so fast and so wildly that there was never a breather to form and write down best practices or opinions.” js and ruby were released in 1995, with rails coming 10 years later. so js development wasn’t fast in the least. after the little ajax blip resulting from outlook web access’s release in 98, js stagnated until node came along and made it interesting again. prototype/jquery a little before it an…

Well of course js development is slow - you can upgrade Rails v1 to differently opinionated Rails v2 on a server overnight. Try upgrading 8 billion people to a browser that supports eg. ES6.

sure, but that wasn’t a problem 10 and 20 years ago when js development was even slower. the language just hasn’t evolved as much as others.

Re: I Miss Rails

#456

Post by clairity » Fri, Mar 29, 2019, 3:59 PM UTC

rimliu wrote: post #364

clairity wrote: post #291

Earlier quoted context omitted.

> “...JS developed so fast and so wildly that there was never a breather to form and write down best practices or opinions.” js and ruby were released in 1995, with rails coming 10 years later. so js development wasn’t fast in the least. after the little ajax blip resulting from outlook web access’s release in 98, js stagnated until node came along and made it interesting again. prototype/jquery a little before it an…

JS may be released in 1995 but it was not used much for anything except changing status bar messages, or image rollovers (remember those?) for at least ten years. Sure there also was Microsofts DHTML on IE, which failed to get much traction. Only starting with GMail and AJAX in 2004-2005 JS started to be treated seriously, it still took some years and SPA hype to become "a must". So about ten years max. Also Rails wa…

yes, agreed, ruby is a much friendlier language to start with, but javascript is more accessible since everyone has a browser or three.

while gmail brought a nicer ui/ux, i still credit ms owa for showing the world what js could be when it grew up.

Re: I Miss Rails

#457

Post by cpursley » Fri, Mar 29, 2019, 4:06 PM UTC

cpursley wrote: post #367

jph wrote: post #14

Elixir and Phoenix are excellent, and you'll recognize many of the same concepts in these. The communities are excellent as well, and welcoming to newcomers.

I've found that Elixir & Phoenix are the right abstraction of Rails-like productivity but "enterprise" reliability and scalabiliy. You get real-time functionality out of the box (dosen't feel cobbled together like Rails channels and blows anything Node out of the water). And with LiveWiew coming soon, it's going to be even quicker to build modern interactive experiences.

Speaking of LiveView, this is freaking impressive:

https://twitter.com/i/web/status/1111641314955849728

Re: I Miss Rails

#458

Post by tombert » Fri, Mar 29, 2019, 4:10 PM UTC

Kalium wrote: post #454

tombert wrote: post #451

Earlier quoted context omitted.

Sure, it's possible that my mentality is totally wrong, and that having a large amount of cruft is necessary for a lot of big projects. I even built my own framework to experiment with these ideas [1][2] (forewarned that I think I swore a lot in the codebase and READMEs if that's a problem). I can see that there's probably a case to be made for "forced-best-practices", which a lot of frameworks kind of entail, and I…

I've personally watched experienced, mature, and very senior developers build sophisticated systems with Clojure, Ring, and http-kit. And blink at me owlishly when I ask them how they sanitize their inputs to guard against SQL injection. They'd been intensely focused on the core logic, architecture, system design, and business functions. Everything else received cursury consideration. My takeaway from that is that it…

I don't disagree with what you're saying; as I've stated, I'm not opposed to having some good libraries to handle dangerous stuff. I personally try not to do SQL with direct string-concatenation, and instead opt for something to sanitize stuff before I actually run anything (at least anything that has the chance of ever touching outside my basement :) )

I just don't like having the libraries forced as part of the full structure, and instead would (typically) prefer them to be functions (or if you're in a Lisp, possibly macros).

Re: I Miss Rails

#459

Post by toasterlovin » Fri, Mar 29, 2019, 4:17 PM UTC

hombre_fatal wrote: post #328

toasterlovin wrote: post #259

Earlier quoted context omitted.

This might come as a surprise, but async everything is not a feature for a lot of people. It’s a bug.

Yes, of course. Everything is a trade-off. Why the condescension? But there are almost zero async-everything options in the space. And being confined to a second-class async subworld inside a synchronous ecosystem is a classic error-prone challenge whether you're using Twisted, Event Machine, Tokio, or Netty. It's a pretty big downside of using Event Machine which even created its own networking primitives instead of…

Sorry, I guess on a second reading you were specifically talking about situations in which you want to do things async. In that case, I can see that everything being async from the start is preferred.

I was talking about the general case of everything being async in JS. That's frequently touted as a benefit of JS, but it's utterly maddening to workaday web developers. You want your requests to be served async (which should just be handled by whatever framework you're using), but inside of a single request you mostly just want to write synchronous code, even when you're dealing with IO. It's a lot easier to reason about.

Re: I Miss Rails

#460

Post by Kalium » Fri, Mar 29, 2019, 4:20 PM UTC

tombert wrote: post #458

Kalium wrote: post #454

Earlier quoted context omitted.

I've personally watched experienced, mature, and very senior developers build sophisticated systems with Clojure, Ring, and http-kit. And blink at me owlishly when I ask them how they sanitize their inputs to guard against SQL injection. They'd been intensely focused on the core logic, architecture, system design, and business functions. Everything else received cursury consideration. My takeaway from that is that it…

I don't disagree with what you're saying; as I've stated, I'm not opposed to having some good libraries to handle dangerous stuff. I personally try not to do SQL with direct string-concatenation, and instead opt for something to sanitize stuff before I actually run anything (at least anything that has the chance of ever touching outside my basement :) ) I just don't like having the libraries forced as part of the ful…

I totally agree. I don't like the feeling that comes from being babysat by code that tells me how I need to do everything.

But I don't know a better way to guard against all the things I might not think of. I've seen entirely too many cases of people doing something wildly unsafe and reckless because it seemed the easiest to them at the time and their tools didn't handle things for them. Use serialized objects to communicate between systems? Why not? We have Avro schemas, those will keep us safe as we deserialize random data from the network, right?

As much as I don't like being treated like a child by a bucket of bits, I also recognize that Clojurian minimalism requires me to think of literally everything. I have to know everything that's dangerous, how it's dangerous, and how to guard against it. That feels like even more load on my brain than being babysat - and I'm a security specialist!