Comment by ceol | Hacker News Reader

Comment by ceol | Hacker News Reader

Parent Full thread

> So your complaint is that when you use @ to suppress an error it... suppresses the error?

From your edit, you've only been working with PHP for about 6 months, so you may not have seen some of the code most others have. Can you imagine, then, diving into a website's back-end to see @ all over? It turns out, the previous developer realized all those nasty notices and errors stopped happening if he slapped a @ on everything. Now your client/boss/friend is pissed because when someone submits their form and they have a space in their name, the entire site crashes, and you have to figure out what's causing it through a tangled web of spaghetti code with no errors.

This is the norm in PHP because, like the author said, it makes it incredibly easy to write bad code.

> If this one thing that annoys me on HN it's the pervasive anti-PHP snobbery.

This isn't anti-PHP snobbery. To write it off as such is an attempt to dismiss most of the (quite valid) complaints. And complaining about someone complaining is really only highlighting your own hypocrisy.

As you said, you really don't understand his need to complain. That's fine, but many on HN do. This was a great write-up for all of PHP's downfalls. Some of it is really only complaining about how PHP isn't Python, but there are a lot of fantastic points about language design and how PHP sacrifices consistency for... really no reason at all. So don't assume every complaint about a language is "snobbery" or an attempt at "streed cred" (wut?) Instead, ask yourself why there are so many people writing so many complaints about this one language. Then, ask yourself why you're writing them off.

> This is the norm in PHP because, like the author said, it makes it incredibly easy to write bad code.

I do maintenance for a lot of PHP sites at my job, and the code is staggeringly difficult to debug or modify. It's punishment for my sins.

PHP was a tool for generating textual data structures (HTML) and it should've been leveraged by another tool for control flow and business logic, but it wasn't. The code I maintain is pre-frameworks; it's got no DRY, no separation of concerns, no design forethought at all. It's just HTML generation that satisfies its needs as they arise.

Now PHP is a quirky, mediocre, and capable language that's still really good at generating HTML.

> Can you imagine, then, diving into a website's back-end to see @ all over? It turns out, the previous developer realized all those nasty notices and errors stopped happening if he slapped a @ on everything.

You've turned something that's a person's fault into something that's the language's fault.

Other languages have some sort of warning suppression as well, like Java's @SuppressWarnings or C#'s #pragma warning disable. Although they won't suppress all errors like PHP does, it can still bite you if you don't fix them.

It IS snobbery. You're taking a look at other people's code, and judging the language from it. I've written PHP for about 3 years and I've never once used the @ to suppress errors.

> You've turned something that's a person's fault into something that's the language's fault.

> You're taking a look at other people's code, and judging the language from it.

At some point, you have to start blaming the language for fostering an environment where that code is acceptable.

Meanwhile, people still manage to write memory leaks in garbage collected languages specifically designed to not leak memory.

GC is not about preventing memory leaks, it's there to make managing memory easier not automatic. One of it's biggest advantages is the ability to deal with memory fragmentation which is ridiculously hard to do well in C++ style languages.

You can write bad code in any language. It's the programmers responsibility to make sure the stuff they are writing is good and that only comes from experience with the language.

Of course you can. But is it always an equal share of bad code for each language? If not, then you have to admit that the language itself will encourage or discourage bad code or bad coders.

It's easier to write bad code in certain languages, as well as vice versa.

The article tries to be a comprehensive list of problems with PHP and @ is a notorious one, even if you personally are disciplined enough to avoid it.

Not to mention that I have no idea why the original commenter picked on this. It was listed as one of the 7 or so things that can go wrong with that one single, not unusual line of code. It's not like he had a whole paragraph about why @ is bad.

> The article tries to be a comprehensive list of problems with PHP and @ is a notorious one

Nonsense. The @ error suppression is a tool, just like any other. It should be used sparingly, but it does have its uses. I have been writing in PHP for over a decade and I have used it exactly one time. And yes, it irritates me when I see it in other's code all over the place ... which is why I refactor all external PHP code before I place it inside of mine.

The amazing thing about PHP is that it has a plethora of tools available and the language doesn't force you to write code in some constrained manner according to what some snob perceives as the right way. The only right way is the way that works and works well.

You don't like a feature of PHP? Don't use that feature. Simple as that.

Ok, let me fix that: "The article tries to be a comprehensive list of problems with PHP and @ is an error suppression tool that is notorious for being misused throughout the community".

> I refactor all external PHP code before I place it inside of mine

> You don't like a feature of PHP? Don't use that feature. Simple as that.

I don't know anything about you, but judging from that attitude you haven't worked in many teams. Of course it's not as simple as that. If I had a penny for every time I had to fix someone not checking that strpos() === FALSE, well, I could fund my own startup. You may have the luxury of refactoring all over the place, but the reality out there is that horrible code like this is left to fester until it causes real business damage.

> I don't know anything about you, but judging from that attitude you haven't worked in many teams.

I haven't worked on any teams. I've always written software solo. What of it?

> You may have the luxury of refactoring all over the place, but the reality out there is that horrible code like this is left to fester until it causes real business damage.

Refactoring is not a luxury, but a necessity. Not only do I refactor other peoples' code, but I refactor my own. That's the only way to get to a quality code base.

> I haven't worked on any teams. I've always written software solo. What of it?

Well, that means you have zero experience with the majority of concerns expressed in this thread, and so are not qualified to opine on them. You work in a happy bubble and I envy you for it, but in the real world you very very rarely get the go-ahead to refactor old code. So in the real world, you very very rarely get to see a quality code base. In any language, really, but PHP compounds this problem with its idiosyncrasies. But you wouldn't know about that.

Isn't the fact that you have to refactor so much of other peoples code an indication that something might be wrong? I'm not familiar with the PHP ecosystem but I don't know of many people having to refactor Ruby gems.

> Other languages have some sort of warning suppression as well

But there is a huge difference here. PHP directly encourages it, making it so easy to do "just prefix it with @" and dedicating a part of the core language syntax to it (thus spending such a nice character for such a triviality). I believe (correct me if I'm wrong) that in Java, it is just another annotation, and in C# just another preprocessor directive.

And I think that's what the article is about - so many things are wrong in the very foundation of the language.

In Java annotations can be used to ignore compiler warnings, not runtime errors.

To get the same effect as @ you'd have to use try {} catch {} blocks all over the place and leave the catch blocks empty, as with any other language with runtime exceptions. Sadly this is done more that one would think...

But that was exactly my point. When you do it in Java, you are obviously doing something wrong (or at least not intended to be done). It just looks wrong at the first glance, with empty blocks and all. In PHP, it's just a single character prefix, no bother at all to add, and looks just like any other sigil. Its usage is definitely not discouraged by design, quite the contrary.

> You've turned something that's a person's fault into something that's the language's fault.

PHP programmers have only one single thing in common. PHP.

> I've written PHP for about 3 years and I've never once used the @ to suppress errors.

Is it so wrong to use @? I've always used (@$_REQUEST['foo'] === 'bar') as a shorter way of writing (isset($_REQUEST['foo']) && $_REQUEST['foo'] === 'bar') - I'm curious if there's a problem with that approach.

Actually yes because the error gets generated anyway (and that is slow) and then it's just suppressed at the last moment.

You'd probably be better off creating a function to do what you want succinctly rather than using the suppression operator.

You be using array_key_exists instead.

"You've turned something that's a person's fault into something that's the language's fault."

Out of curiosity, where do you side on the C-is-an-unsafe-evil-language-because-of-pointers debate?

"This is the norm in PHP because, like the author said, it makes it incredibly easy to write bad code."

PHP makes it extremely easy to write ANY code. But since article saying "I've noticed this piece of good code in PHP" is not going to get on HN anytime soon, you get much more attention to bad one. That's like saying Internet is the source of filth because it makes easy to post bad stuff. It makes easy to post all stuff, that's the point.

"This was a great write-up for all of PHP's downfalls."

These "downfalls" are known and discussed 1000 times. Really, who needs another article about "PHP sucks because I can find 3 functions with inconsistent parameters"? How exactly it is going to make anything better? What exactly does it contribute to the world?

> PHP makes it extremely easy to write ANY code.

No. PHP makes it easy to write bad code and between normal and challenging to write normal code. This is why Doctrine, the closest thing PHP developers have to SQLAlchemy, needs three ways to declare table and column metadata: a separate YAML file, a separate XML file, and in PHP comments. All three of those methods are awful, and only one allows you to keep the metadata and class in the same file.

> These "downfalls" are known and discussed 1000 times.

And, like a wiki, it's useful to have them all in one place.

And again you make the same mistake. You take one solution, you say it doesn't work like your favorite solution (while obviously all proper solutions should be done exactly like the one you prefer, since there's only one right way to do anything and it's, not surprisingly, exactly the way you prefer) and you make it the fault of the language used for it's implementation. How that even makes sense?

I write code in a number of languages, and there's nothing "challenging" in writing proper code in PHP. The fact that your example has nothing to do with code quality or language only underlines that.

> "And, like a wiki, it's useful to have them all in one place."

Maybe, but that article isn't it. PHP has a wiki: http://wiki.php.net/. You want to list things you want fixed in PHP and have it in one place? Ask for an account and maintain a page there. Rehashing old tired "strpos doesn't have underscore and str_rot13 has, oh noes!!! I can't use language with missing underscore!!!" is really stale by now and doesn't contribute anything new.

> How that even makes sense?

Maybe my point wasn't clear. It's not a difference in how those two ORMs accomplish their goal; it's that PHP doesn't let you do what SQLAlchemy does, even if you wanted to. Want to write an ORM that uses YAML, XML, or comment annotation in Python? Great! Want to write an ORM that uses classes or objects as column declaration in PHP? Too bad.

> Rehashing old tired "strpos doesn't have underscore and str_rot13 has, oh noes!!! I can't use language with missing underscore!!!" is really stale by now and doesn't contribute anything new.

Now I think you're just being petty. Obviously an underscore isn't what we're talking about here. Take your straw man elsewhere, please.

> Want to write an ORM that uses classes or objects as column declaration in PHP? Too bad.

I'm actually doing just that; it's not as clean as you can do in Python but it's not outrageously verbose either. Trying to map one languages features onto another is not a valid critique.

So PHP does not have the same power as Python, just like Java. Yet I don't see people saying Java encourage writing bad code.

Different language has different limitation. Just because you can not have SQLAlchemy in C, does not mean C encourage bad code.

> Want to write an ORM that uses classes or objects as column declaration in PHP?

What prevents you from doing that in PHP?

PHP doesn't allow you to pass functions and classes as arguments.

You can. Here's my code as an example:

https://github.com/gigablah/fsphinxphp/blob/master/src/FSphi...

Note the $getter parameter.

Unless I'm misunderstanding you, passing functions as arguments is definitely possible with anonymous functions/closures.

I have no idea what you mean by "what SQLAlchemy does" - I'm sure it does lots of things and most of these are perfectly doable in any language out there. You must be meaning some specific thing that is a favorite of yours - and indeed maybe it is hard to do it in PHP (or maybe not - since you didn't tell what this thing is nobody can have any idea). Who cares? I can name a bunch of things that is harder to do in Python that in some other language, but I don't run around shouting "Python sucks".

> Now I think you're just being petty. Obviously an underscore isn't what we're talking about here. Take your straw man elsewhere, please.

What he's talking about here is known and talked about for 10 years. It's just function names, it's not that big a problem. Yes, we know old function names are inconsistent. No, we can't change them because that will break code. If that's what makes or breaks language for you, PHP is not for you, and most languages are not for you, because you're obviously not interesting in solving the problem, you're interesting in boosting your self-esteem by boasting shiniest tool out there. Nothing wrong with it, but PHP is not created for that. If that means for you that PHP sucks, fine, but for 99.99% out there shininess is very low on the list of priorities. So if you want to discuss important stuff, leave old stale stuff like ranting about strpos alone and discuss important stuff.

> I have no idea what you mean by "what SQLAlchemy does"

You must have missed the rest of that paragraph. I specified what it does.

> What he's talking about here is known and talked about for 10 years. It's just function names, it's not that big a problem.

Again, missing the point. Python has these same minor inconsistencies in its stdlib. It's the combination of all these minor inconsistencies with a language that gives you more tools to write bad code than good that makes PHP torturous. It's everything that the author wrote about in one neat little package.

Then again, why can't the author not post it on HIS blog? No less.

Any old tired issues for you could be something new for someone else.

That's fine, anybody is free to rant on his blog. That's just a rant and not some revelation about something significant about PHP.

Reply on news.ycombinator.com