Best Practices Exist For A Reason
tomdale.net
tomdale.net
Between semicolon-free folks and JavaScript traditionalists, who gets to play the role of the expert? If Node.js is popular, does that mean that Isaac's comma-first style is "correct"? (https://gist.github.com/357981)
In the absence of a quantitative engineering method with which to evaluate either approach, isn't it a purely personal and political choice?
My advice? Don't work with people whose best practices are your anti-patterns. Or at least strive to work with people who agree with you on a core set of principles. Or hell, work with people who agree with you on nothing. Just don't complain when you can't work with any of them. :-)
And besides that, what's wrong with "purely personal and political choice[s]"? Being that I'm the person who's writing my code, I'm allowed to make a certain number of purely personal choices. That is, as long as I take into account the political consequences of doing so. Put another way, it's ok to have personal preferences, but you have to take other peoples' feelings into account too.
Edit: to use an analogy, imagine if all vehicle development was treated with the same terminology. Whether you're building helicopters, spacecraft, mass produced commuter cars, RC toys, nuclear powered submarines, formula 1 race cars, or tunnel boring machines: all of it is just vehicle manufacturing right? Just as anyone would find it ridiculous to put forward advice that could possibly apply across all of those distinctly different disciplines we should view advice about software with the same skepticism.
>"Best practices" for development of an iphone game may not be the same as for flight software for an orbital rocket or for an international banking backbone.
I disagree. There are a few things that should be common across all of these, like not storing passwords in plain text.
To me, consistency is important. All of my codebase is traditional comma postfix, so I'll keep using that; however, maybe with some new projects I'll try on the comma prefix pants and see if it helps.
Thanks for the link!
To me
, consistency is important
. All of my codebase is traditional comma postfix
, so I'll keep using that
; however
, maybe with some new projects I'll try on the comma prefix [...]
They're always correct, they copy without any editing, and the "var"s line up to show that it's a block of assignments the same as commas or whitespace.
var a = 2; var b = 4; var c = { blah: "whatever" };
Leaving out a semicolon has no effect. They can be copied without any fuss. It's easy.
x = {[
{ id: 1, data: 0 },
{ id: 2, data: 0 }
],
[
{ id: 3, data: 0 },
{ id: 4, data: 0 }
]} var foo = 1,
bar = 2
baz = 3
* There's a missing semicolon on line 2
* There's a missing semicolon on line 3
* baz was being used before it was defined.
That's just in a global context. When it's in a function you get a little more:
* the baz line is indented too far - because it's being interpreted as a new statement.
Sure, you could eyeball your code looking for this. Or just use an automated tool that'd do it for you
It's so ugly to me that the whole idea seems like it could be a troll.
And let me play the advocatus Dei
, because that actually looks quite readable
.
To me, consistency is important
. All of my codebase is traditional comma postfix
, so I'll keep using that; however
, maybe [...]
What convinced you? I read it and saw something wildly different, which is best reserved for a good reason. The reason appears to be visual recognition of delimiter mistakes that the parser will catch anyway.
The gist contains examples of various mistakes in both styles. Most of the errors are syntax error which will be immediately rejected by a parser. While the comma first errors are generally more obvious, they are silently bad by unexpectedly returning undefined. The standard "var" example is silently bad by leaking vars into the global scope when chaining initializers on a single "var". My conclusion from these samples is that comma prefix formatting only practically helps to prevent hidden mistakes in chained "var"s. Putting each variable declaration on a line of its own is far less distracting than reformatting every list and object.
Sample comma first style error. This seems like a strange thing to do, but I'm unfamiliar with this style.
return
{ a : "ape"
, b : "bat"
} // returns undefined,
// then creates a block with two named statements.
var a = "ape eat banana",
b = "bat, allowed to fly",
c = "cat toy",
d = "dog chasing the mailman,"
e = "elf lord",
....
// leaks "e" into global scope return
{ a : "ape",
b : "bat"
}
Omitting semi-colons, however, can cause novice JavaScript programmers serious grief. Nevermind novice JavaScript programmers – they cause me grief, and I put semi-colons everywhere! That's because I'm often called in to debug our customers websites, and on several such occasions I've had to fix bugs that never would have occurred if that customer used semi-colons properly.
Sure, if our customers understood JavaScript's ASI and how to use semi-colons "properly", I wouldn't have this problem either. Except, as Tom points out, the reality is that almost none of these people will invest in doing so.
Really?
The comma first style makes it less likely that your list will end with a trailing comma, which can cause IE-specific bugs. If you're trying to track down the IE breakage a couple of weeks after you wrote down the code, it can be fun to track down the extra comma. If you don't know that this bug is possible (as I'm guessing you don't), multiply the hair-pulling by 10.
See the cautionary story in http://www.enterprisedojo.com/2010/12/19/beware-the-trailing... for a real life example.
With the comma first style, that bug would never have been introduced. Neat, huh?
Lists usually grow on the end, not the beginning.
(I use this kind of formatting for SQL, for a similar reason.)
If you run any syntax checking or minification tools, this will be immediately obvious. If you test your site in IE at all, this will be immediately obvious. If you record JS errors on your site, this will soon be obvious. If you record any kind of usage statistic, this will eventually be obvious.
Take it from someone who has experimented with cutesy formatting in the past[1]: you don't need to use cutesy formatting to solve this problem.
[1 | You can end class declarations with _:0} instead of a bare } to avoid the trailing comma issue. ]
And it is important in our industry to write MAINTAINABLE code.
I freely remove, add, and move things to the bottom of my js classes without ever hitting trailing semis, or hitting a missing semi. I think that is more maintainable then the alternative
Having mechanisms in place to catch programmer errors is no excuse for adopting practices that make programmer errors more likely.
It is an IE bug that causes it to break. One of the many ways they do not follow the spec.
ObjectLiteral :
{ }
{ PropertyNameAndValueList }
ObjectLiteral :
{ }
{ PropertyNameAndValueList }
{ PropertyNameAndValueList , }
Not to mention, IE8 was released half a year before ES5 (according to a quick check on Wikipedia). So yes, pre-ES5 I'd say it's a normal thing to do not to allow trailing commas.
If you look at the language Tom objects to the most, it's language Isaac uses that promotes this tribal attitude that the other side doesn't have your best interests at heart, and questions their motives.
We're just terrible at this, and I'm not sure any human society has ever really cracked it.
The tribal attitude is the enemy of this process for sure, and humans are very prone to that attitude but that doesn't mean you can't still find the answer more likely to be right.
In the absence of a true quantitative engineering method, there is a middle ground other than personal choice: common anecdotes. Presumably, those shameful leaders have heard enough stories about the issue to make a reasonable recommendation, sort of like advising against shipping server software with an open default security strategy even though I don't have any peer-reviewed studies saying such is unwise.
Stylistically, I side with the no semicolon people. Grammatically they're redundant.
Pragmatically, if you ever find yourself saying "read this enormous, hard to read spec just so you can identify the three or four edge cases that will burn (yet remain statistically likely to commit)" you are on the losing side of history.
Basically, it's preposterous to have reasonably smart people seriously argue that the option that is least user friendly and most likely to cause pain down the road is the best one out of a sense of misguided spec puritanism.
You're framing this as a disagreement between equals. It is not. There are degrees of expertise. (Total lines of production code * amount of usage) is a reasonable approximation, and those who rank highest on this scale are overwhelmingly in favor of requiring semicolons.
If Carmack and Torvalds were in agreement on a C best practice while Odesker^W^W^W^W^W^W^W Topcoder user #12571 said to do something else, I wouldn't go with the latter.
Google, responsible for a JS engine, multiple compilers that target JS, and what is likely to be the single largest JS codebase of any company: http://google-styleguide.googlecode.com/svn/trunk/javascript...
Brendan Eich: "The moral of this story: ASI is (formally speaking) a syntactic error correction procedure."
Or you can go with the Node.js guys.
If you are doing a Topcoder round, then the random Topcoder user's advice is probably better than Carmack's. Carmack or Torvalds would probably tell you to use a variable name that gives a lot of information about its purpose, so that other people will find it easier to read your code. The Topcoder user might tell you to use a cryptic abbreviation that helps you remember a variable's purpose but clues little to the outside reader, since Topcoder users are penalized for allowing others to find their bugs. When implementing an algorithm, Carmack or Torvalds would code it as part of an extensible system with an API so that it can be used for many purposes. The Topcoder user might code it with no potential for reuse, but would do it in the way easiest to code quickly without making a mistake.
Right, because what could those experts possibly have to teach you? Fuck that intellectual crap, brogrammers are here to crank out some code yo! Just because those dinosaurs have spent "thousands of hours" studying and trying out different programming styles and concepts means I should listen to them? This is the internet age man. Some guy in a github comment told me different and github is cool.
It's like those functional programming professors who try and talk about how continuation passing style was a bad road to go down because of maintenance problems and how difficult refactoring and just reasoning about code becomes in large systems.
Fuck that egghead noise, node.js has callbacks and node is how you get webscale.
I'm right with you bro!
The point of your parent was "don't let the experts think for you", not "screw the experts" - that's a very important distinction. You should form your own opinions, not just uncritically adopt the opinion of some expert. It's unquestionable that expert knowledge and experience is important - but nothing beats forming your own (hopefully informed) opinion.
"You should form your own opinions, not just uncritically adopt the opinion of some expert."
False dichotomy again, there is a lot of ground between "uncritically adopt the opinion of some expert" and "forming your own opinion and hoping it's informed".
You should find that middle ground, it's where all the smart people are.
I didn't miss the point at all, I just completely disagree with you. I don't think it was needlessly rude either, it was sarcastic. And I feel no need to treat people who are promoting anti-intellectualism with kid gloves.
When discussing coding style, you can quote experts. Or you can understand their arguments, understand the impact and tradeoffs of coding decisions, and realize why they are true. And in the case of Topcoder, about as far as possible from building large systems, you can realize that the optimum style changes and understand why.
I think there's a broad trend to treat code as magic, something where anything can happen. Most people learn coding styles by being given suggestions, and then noticing from experience that they like it better. If you reject this mindset and believe that code is orderly, then this advice changes from Wisdom of the Elders (TM) into an engineering calculation. I specialize in automated restructuring of software architecture, so this mindset is something I try hard to fight against.
"Intellectualism is not about showing off how much you agree with authorities"
My point was that intellectualism is not about agreeing with authorities but giving proper weight to the opinions of intellectuals in the field when weighed against lay opinions or smaller groups of intellectuals.
You certainly don't seem to be arguing for the post-modern interpretation I was arguing against where expert opinion is considered "just another opinion, equally valid" so I'm confused why you think my stance is almost the opposite of what I was stating.
My first comment hit the advice to always follow Carmack over the random Topcoder user. It's important to understand what makes a best practice a best practice rather than just listening. A Topcoder round and Quake differ greatly enough that the random Topcoder user is sometimes the greater authority.
You seemed to map the call to critically evaluate expert opinion to a call to ignore it, something I'm sure you've seen too much of. But it's important to understand the difference between critical evaluation and anti-intellectualism. Indeed, the former is the higher form of respect.
I'm not sure how well it applies to a subject like best practices where it's either subjective or at least fuzzy. The arguments are "in my experience this causes problems when working with other people and is more prone to errors long term". There have been no counter examples or analysis I've seen around this issue, just a "that's just what the traditionalists say" type argument. Something along the lines of your topcoder example that applied to the semicolon argument (god i hate this issue) would be very welcome.
Any critical analysis here is hard to do objectively, and there is a consensus of experts (including Eich) on one side. If i'd seen any actual critical analysis rather than a jump to "don't let the experts tell you what to do! be a rockstar!" then I wouldn't been so fast to use the anti-intellectualism card. But it's so prevalent in the node community (more than Ruby's but less than PHP's) that it's depressing.
With semicolons, the search for the boundary between statements reduces to the search for that single character -- that means fast code skimming.
The boundary is also robust against changes; as long as that semicolon stays in place, nothing that happens within that statement can cause that boundary to become an interior.
There is an exception: Strings. This is no trouble for machines, as strings are processed by the lexer. Humans can't instantly take a string and process it as a chunk, but a UI innovation does this work for us: syntax highlighting.
As an example, consider the code:
x = a && b
foo()
x=a &&
foo()
LR parsers take advantage of this for error recovery, allowing them to find more than one parse error at a time.
A more sophisticated view is to think about syntax abstractions. Given a program like "x=y; return x;" we can create a program written in the abstract syntax "statement; statement". (The dual operation is to concretized that back into the set of all two-statement programs. This kind of relationship is known a Galois connection.) Having semicolons makes the abstract syntax unambiguous, allowing us to efficiently reason at the statement level.
Error correction simply means "add a semicolon if it's necessary."
He says it should be thought of as an error correction feature.
That is: do not in any case rely on ASI. Add your semicolons and only let ASI do it works when you do an actual ERROR and forget to add one.
And google is well known for terrible quality javascript. http://www.sitepoint.com/google-closure-how-not-to-write-jav...
And finally, Crockfords reasoning about things like this seems based on the presumption that js developers are drooling idiots. For example, he recommends against using the new keyword for object creation in javascript, even though object factories can often be a huge waste of memory, because dumb javascript developers may forget to use the new keyword at times when they want to instantiate objects.
I am not saying you shouldn't listen to smart people, but I am saying that you should actually learn the reasons they say the things they do, and use your own brain when making choices.
No.
Here's one idea: perhaps everyone's opinion is not equal on every subject. Perhaps by looking at the opinions of respected accomplished people who have garnered the respect of other respected accomplished people you can being to see who's opinion should get more weight. This is the process that came up with the very ideas of things like "engineering" and "quantitative".
I know we love the young upstarts disrupt the old way of thinking meme. That's why you framed it this in that way with the folks vs the traditionalists. You might want to remember that historically the young upstarts have been wrong 99% of the time. That's why it's news when they are right.
"One man's best practice is another man's anti-pattern"
And in the vast majority of cases, when you ask that question about a specific problem: One of those men is wrong. Even if they're both smart. But how to tell? If a huge number of other smart and experienced people have been convinced by one of them ... well I'd go with that guy. You won't be right all the time but it's the best system we've come up with so far for things that don't conform to double blind placebo controlled studies or mathematical proofs. It literally is the best way to be right most often.
This reminds me a lot of abstraction in software systems. For 95% of cases, the abstractions make things quicker in the end. For those 5% of edge cases which the abstractions don't cover properly, you spend inordinate amounts of time trying to get things to fit. The bigger point here is that trying to white-wash anything with a particular methodology is both the right and wrong thing to do. It's right because most of the time it will work, it's wrong because when it doesn't work, you will have a hell of a time dealing with it.
The best system we have come up with so far is actually to take things on a case-by-case basis instead of trying to solve everything in one fell swoop. The problem is that one must be mindful as consistently as possible to achieve this. Most people are not capable of keeping that line of focus for an hour, let alone a day or their whole lives. Funnily enough, the solution to this is meditation, not a set of best practices or expert opinions.
This will answer the question, "who is best at arguing their point". It does not reliably answer, "whose point is correct?"
That's why this system relies on convincing smart, educated people and not just people in general. The need for intellectual discourse to avoid this exact problem and what that looks like is taught as part of being educated.
No. Isaac doesn't define what style node.js is written in. And in an interview when he took on the lead developer role on he affirmed that he would abide by their current coding preference, which wasn't comma-first. That's Isaac in a nutshell, he works with the developers around him.
You may be mistaking Node.js with npm. npm is Isaac's code, and that is comma-first I believe. It's his project, and his rules, and his call as to what coding style he wants. People who participate in his projects either accept that, or walk away. Node.js is an existing project, so the existing coding preferences are in place -- unless, I guess, there's a general consensus that the coding preferences in place are causing an issue.
The only problem with this approach is the diffs between the repo and your working copy.
Second, the CouchDB thing isn't because of a lack of best practices, it's because the wrong best practices were misapplied. And in the end, its use of best practices prevented it from severely hurting CouchDB's reputation.
The best practice is that password data should be stored as a salted hash, so if the database is compromised, the original password can't be retrieved. It's rare that the salted hash is intended to be public information, and that's what the CouchDB people did. What they should have done is realized that since they're deliberately sharing the hashed passwords, the original best practice doesn't apply, and they need to break it down and reexamine it (probably a good idea anyway for a project of its size). And they should have reached the conclusion that SHA1 is much too weak and/or that the salted passwords shouldn't be shared. In the new version, CouchDB 1.2, where an effort to correct the problem was made, the passwords aren't public. But at least, aside from the encryption strength, they got the hashing right, in that they used salts so rainbow tables can't be used. SHA-1 is easy to brute force for simple passwords, but as passwords get longer and contain more than dictionary words, it gets harder to brute force. It's easy to communicate that they screwed up, and easy to communicate that it's no worse than Sony password databases that were compromised, but the truth lies somewhere between the two, and can be seen by carefully considering the details of the case.
The best practice is to use a purpose-built password algorithm like bcrypt. Algorithms like SHA1 and MD5 are designed to run fast; even with salts they can be cracked by massively parallel hardware like a room full of GPUz.
best practices and most practical practices are not always the same though.
Good advice comes with a rationale so you can tell when it becomes bad advice. If you don't understanding why something should be done, then you've fallen into the trap of cargo cult programming, and you'll keep doing it even when it's no longer necessary or even becomes deleterious.
source: http://blogs.msdn.com/b/oldnewthing/archive/2009/11/04/99170...
The reason people talk about omitting semi-colons in JavaScript is because they don't think that feature is unusable--they clearly think it's better than adding semi-colons everywhere! When reasonable people disagree like that, the situation is rarely simple and clear-cut.
Now, perhaps you could reasonably argue that the behavior is unfortunate. But you can also reasonably argue the converse. And, most importantly, it isn't immediately clear who is correct.
I guess I just don't get it because it has always seemed precisely a simple and clear-cut thing to me but somehow it is not.
Contrast this with Python where the overarching rule is there should be one, and preferably only one, obvious way to do things.
http://www.python.org/dev/peps/pep-0008/#tabs-or-spaces
The most popular way of indenting Python is with spaces only. The second-most popular way is with tabs only. For new projects, spaces-only are strongly recommended over tabs.
Function names:
...
mixedCase is allowed only in contexts where that's already the prevailing style (e.g. threading.py), to retain backwards compatibility.
However, I found it to be of little value. When fully reading Isaac's post (http://blog.izs.me/post/2353458699/an-open-letter-to-javascr...), it appears to not diverge much from what you've written.
I consider myself a beginner in Javascript and did not find Isaac's post the least condescending. I actually learned a few things and am very glad he took the time to explain how things work. Because honestly, I won't read the ECMAScript spec unless I need to.
I'm actually one of the people who never use semi-colons to terminate statements. That's for personal projects. Where I work, we (as a team) decided to use them. And that's fine, but I prefer not to use them.
I'm part of HN not for very long, but what strikes me are the similarities between supply chain management and programming in some aspects. The purely technical issues are out of scope here since I cannot judge them, it's more about the principles behind. Like in this case here.
Best pratices provide a guideline and are condensed experience. They are by no means the only or best solutions, rather ones that can be applied in a lot of cases withou screwing up. That said, one should also reach a point where understand them. One has to in order to understand when to use them, most people get at least to that point.
What you actually have to do is reach an understanding of not only when to use best practices but one of knowing why you use them. This is not true mastery, but it will get you far enough.
Ultimately, you need people who know when to use a best pratice, when to ignore them and when to actually brek conventional wisdom. What I see a lot lately is people doing the last to bits without having understood the best pratice's when and why in the first place.
For me, that's the key message of this post, and this point is valid in every industry, sector or task at hand.
That is some fairly brazen rhetorical sleight of hand. Schlueter is obviously not attacking the following of best practices. He is saying that semicolons-everywhere isn't a best practice and he gives an argument that the most common rationales given for the practice do not hold water. There are a few flaws with his argument, but what it certainly is not is a general condemnation of the following of actual, bonafide best practices.
Pretending otherwise is akin to accusing someone of "siding with the terrorists" because they argue that people should be allowed to take bottled water on airplanes.
Most arguments about style are almost certainly pointless, but if you are faced with a choice between two styles, one of which may cause a problem sometimes if you don't know what you are doing, and the other style, which will never ever cause you a problem, it seems the choice is kind of obvious, no? Particularly since the guy doesn't really make a case for what his objection to semicolons actually is to begin with. He doesn't like the look of them?
If you ever have to do network and systems administrations you will come across loads of best practices because needing to memorize every last rule and edge case of all the software on your systems would be truly maddening.
That said, it is always good to know the why behind best practices and not follow them blindly.
Best practice is about a practice that helps create maintainable code. Code is maintained by people. So a best practice is really about getting the best code delivered by the people you have in the most appropriate time possible, and code that can be maintained by those same group of people and others who may be brought into the project at a later date.
Insisting on semicolons means that the bar to entry for being able to develop and maintain a JavaScript code base is lower. Also, we can spend more time thinking about the purpose of the code rather than it's syntax. That helps when it comes to debugging, spending more time thinking about what the code is trying to do rather than losing time figuring out whether the syntax without semicolons is causing an unintended side effect.
The lack of semicolons argument reminds me of the Python versus Java argument. Choosing python gives you a much smaller range of people who could work on your project, but on average the quality of that developer would be better than if you opted for Java.
The advantage of the non-semicolon stance is that they are increasing the minimum level of competancy before someone can work on their project. The disadvantage of the non-semicolon stance is that they are increasing the minimum level of competancy before someone can work on their project.
Fuck Javascript. Good parts be damned.