Everything is not equally good | Hacker News Reader

Everything is not equally good

37signals.com

44 pointsmh_27 comments

"FORTRAN and Python are equally good languages. YMMV"-useless and somewhat suspicious.

"Python is generally better than FORTRAN."-more accurate for most people

"FORTRAN is good for numeric computing when performance is absolutely critical. Python is better for almost anything where processing speed isn't a major concern."-most useful and most accurate.

It seems that you right with your last statement:

http://benchmarksgame.alioth.debian.org/u32/benchmark.php?te...

based on the "Computer Language Benchmarks Game".

I wish he would use real examples in these kind of posts, instead of just talking about abstract good and bad ideas.

I think pg's essay "How Art Can Be Good" has a similar thesis but is much better argued, with plenty of specific examples. http://paulgraham.com/goodart.html

All you have to do is follow him on Twitter. The blog posts are always preceded by a flurry of Twitter chatter.

So, what's better? Ruby on Rails or PHP? Probably Ruby on Rails... but then, Facebook and half the internet has made amazing stuff out of PHP in likely less time (and more bugs) than it would take to do with Ruby of Rails.

So.. Your Mileage May Vary ?

Leaving alone the idea that Facebook is written in PHP -- or any one language -- if you are evaluating the OP's idea correctly you would have to look at things being built today in PHP. In my experience there are precious few serious, large-scale, sophisticated projects being built today. (Probably because if you need to build such a thing, you need real programmers, perhaps with CS backgrounds, to do it and such people are very unlikely to use, or be convinced to use, PHP)

Since this post is about evaluating ideas, within that context we would have to look at the trend away from PHP and conclude that the community of engineers* does not endorse PHP as a good idea.

*For some definition which includes people who work on sophisticated projects and excludes people who tweak Wordpress templates

Feels like nitpicking, but alright...

What's better for making a blog? Tweaking a PHP wordpress theme or making something using Ruby on Rails? Your Mileage May Vary ?

The reason that term is used a lot is because there is a huge amount of evidence that each technology choice usually has very specific project criteria in which it shines, and if you change the criteria even a little bit, there is often a technically better choice - let alone a new technology that could be developed specifically for those new criteria.

My experience with blogs says they will quickly turn into their own application and then you either have to mess with integration between the two systems, or you build the application on top of the blog. Each come with plusses and minuses, but in the latter case, if you start with something like Rails, then at least you have Rails at your disposal for the growing application.

Though if it is really never going to become more than just a blog, I choose static files with a simple compilation procedure, personally.

I'm no ruby wizard but there are things like wordpress built on top of rails. monologue, enki, jekyll, etc so you give a false dichotomy

This seems like shifting goal posts. First we're talking about Facebook and half the internet, now we're talking about context-specific technical solutions (and I don't concede that WP is the unilaterally superior blogging option either.)

Moreover, I'm not sure where your disagreement with OP is, since you're simply saying now that for a given set of criteria, a superior choice exists. Which was the message in the first place.

The OP is saying we should not use terms like 'Your Mileage May Vary'. Instead, we need to just say which solution is better. I'm pointing out how this isn't the correct way to go about it unless you specify the exact circumstances.

And this is where I agree with you completely. You can't say that because someone is making a blog they must use WP as it is the correct choice. You can say 'I believe WP will be the correct choice in making a blog but Your Mileage May Vary depending on what you want to do with the blog and if you are more proficient in Ruby than in PHP.

This is exactly my argument, as the OP says you can't do this kind of thing and need to give a concrete solution. However, this kind of thing is incredibly necessary, and saying to use WP for every blog is just wrong.

this is really a problem in software development? a lack of opinions?

i wish i lived in the author's world. in mine, developers seem to take it as a badge of honour to have an opinion about everything. and if they can be an intolerant, insufferable arse about it: bonus points!

[edit: actually, i should clarify. my colleagues are actually pretty good about this; it's online where the volume is always turned to 11.]

Right. This post is just stupid ;)

This kind of reasoning on the part of DHH is troubling to me. He's basically saying that there is always one best way to do something. That may be correct. But what's troubling is that Rails is so focused on its "One best way" to do everything. Rails forgoes complete through-and-through consistency in favor of "magic" -- because it's quicker and easier. That's fine, but it doesn't have the ring of complete truth that I expect from a truly great idea, or framework.

To be fair, a big part of the Rails 3 rewrite was making it modular so you could have different opinions. Of course, not everybody was as equal as everybody else; a lot of Rails extensions seem to assume that you stuck with the defaults. Devise, last time I checked, was very hard to integrate if you switched out ActiveRecord for Sequel. But my information is a couple years out of date.

I think this blog post is really more about reminding people of DHH's brand than adding anything new. Rails has always been "opinionated" software.

Didn't DHH basically use "your mileage may vary" in his "The Parley Letter" essay? [1]

> So yes, the closer your application is to an application like Basecamp -- and I make that net extremely wide, I consider, say, 500px, Github, Shopify, and others to fall into "close enough" -- the closer you are to the primary use case that guides the development of Rails.

...I read that as "your mileage may vary" for rails if your application is dissimilar to basecamp, because design decisions are primarily made to support basecamp's use case (and not yours). Its more subtle than that, and you need to read the essay for the context of that statement, but that was one of the things I took away from it.

[1] http://david.heinemeierhansson.com/2012/the-parley-letter.ht...

I like this non-relativist stance. Simply stating YMMV often kills what could have led to a good conversation, thus possibly preventing us from learning more.

I'd like to couple this attitude with the great advice of having "Strong opinions, weakly held". (http://bobsutton.typepad.com/my_weblog/2006/07/strong_opinio...)

If YMMV is getting abused, that's a problem with the abusers, not YMMV. I've always taken YMMV to mean "your use case may be different than mine; I don't know everything, but this works for me in this situation."

Good, better, best, bad worse, worst are relative terms; they only have meaning in context.

Purely pedantic nitpicking, but the title feels poorly worded to me.

"Everything is not equally good": for all x and y in the universe of things, x is not equally good as y - seems nonsensical

"Not everything is equally good": there exists some x and y in the universe of things for which x is not equally good as y - seems to be the actual intent of the post

YMMV. :)

"Everything is not equally good." For example, Ruby on Rails is not equally as good as Scala[1].

[1] twitter dot com

Almost no one is facing the same issues as twitter. I find it extremely unlike that DHH thinks Rails is best for every possible scenario, and a single counterexample of an extreme outlier isn't an argument against this.

Twitter still uses Ruby on Rails to generate the website, as always. Their use of Scala is on the backend; the services that lay behind the Rails app. Facebook uses a similar architecture behind their PHP code, using many other languages, including Java.

If that's still true, it won't be for very long.

Which one? Twitter/Rails or Facebook/PHP and why not?

It's been the case for three years now. Why would you expect change very soon?

Reply on news.ycombinator.com