Slashdot Mirror
PHP Usage in the Enterprise
acostin writes "Some open survey results were published about PHP usage in the enterprise on the InterAKT site. An alternative survey on the PHP open source mouvement can be found on Zend site. See how we've evaluated the PHP market size($$$), what people think about PHP as an alternative to Java and .NET, and what should be done in order to have your large clients adopt open source solutions."
For both the Zend and InterAKT surveys, there are lots of raw numbers presented, but the interpretation is lacking. The commentaries on the InterAKT results are little more than "as you can see, such-and-such wedge of the pie is the largest," and there is no interpretation whatsoever on the Zend site.
.NET, and found both of them to be too far disconnected from the HTML that I'm trying to create. PHP provides an excellent blend of power, speed, simplicity, and directness.
Of course, this is a cheap and easy way to conduct a survey (multiple-choice), but the results are almost meaningless if they can't be put into context. I would have preferred to have seen a hundred randomly-selected PHP developers interviewed, essay-style, about why they are using PHP, their thoughts on PHP versus other technologies, etc., and then have the results compiled into a journal-quality article supported by graphs and raw numbers. The important information isn't in those graphs; it cannot be enumerated and broken down into clean categories.
Personally, I develop PHP sites because it's the fastest and simplest way I've found yet to publish dynamic web content. I've tried making sites with Java and
It's hard for thee to kick against the pricks.
What do we think about PHP as compared to Java and .net? What do we think about an grape as compared to a basketball and an egg? These are 3 quite different things. A HTML-generation targeted scripting language compared to a compiled general purpose language as compared to a distributed object and language framework is a pretty disparate set of things to compare.
The language as it has been in most of it's 4.x.x iteration has been just about fine. Good for quick slap em together websites and small applications. But...
.Net) where you can rely on the functionality of the objects without having to second guess the original developers.
I have seen huge cumbersome application servers built around PHP that are a nightmare to maintain without having intimate familiarity with the code of the application server, such as Ampoliros or Ariadne, something which defeats the purpose of using such a large system in the first place. Such things really do work better with a OO by design language such as Java or ASP.Net (I assume, don't know
My guess is that PHP needs a better OO design (and no, PHP5 is not it, yet) and better seperation of logic and presentation for larger systems.
But for smaller stuff, well it's hard to beat in terms of price and speed.
I know that you're somewhat joking, but I would have liked the market share questions from InterAKT to have included not just .NET, J2EE, or ColdFusion, but have also included other languages like Perl and Python (although Python is indeed the base language for Zope). There are still a very large amount of websites built in Perl these days, with Slashdot being one of the most famous. Zend's survey does a bit more to explore the languages that programmers "are familiar with," but does little to see how the competition for PHP is doing.
PHP's great to use for me because it's simple, powerful, and readily available in cheap hosting environments. If Zope, ColdFusion, or J2EE had more availability or less cost, then I would try those as well, but there's something to be said for being able to sign up for a $9/month account and downloading Apache, PHP, and MySQL all without paying for anything other than bandwidth costs. You still can't really compare PHP to the enterprise level of .NET or JSP at this point though, since many features like persistent objects in shared memory really can't be done well in PHP, and I haven't heard anything else about PHP 5 other than further enhancements to the objects and reference systems. PHP-Accelerator gives a great boost to the speed, but I'd really like to see native compilation built into the distribution rather than downloaded separatedly.
All in all though, PHP's a great language for quick development of small to medium sized websites. As the old caveat reads though, use the right tool for the job.
Yeah..
To write clean, well structured PHP, you really need to do some good design, use OOP, seperate content/formatting/logic/etc, and basically at that point, you're left with a half-ass OOP implementation, annoying pass by value, messy syntax, no exception handling, etc.
PHP was meant for making relatively minor webpage hacks, and it shows. This still remains PHP's strong point; building enterprise applications (that are designed well) does NOT.
Here's my list of bitches about PHP:
* No class attributes, only instance attributes
* No namespaces (and they were dropped from PHP 5). include()/require()'ing a file just dumps its namespace into the big happy global namespace, and it's a freaking nightmare.
* OOP sucks. You have to litter &'s everywhere to get references, and lots of other problems that I'll refrain from typing out. Ok here's one -- how do you pass an instance method for use as a callback? Something like array($object, "method_name_in_a_string"). Good god.
* No exception handling. Want to "handle" an error? Toss a @ in front of it, then you'll never see it. How helpful.
* Type handling is a nightmare, sure, they make it real easy for the newbies to use numbers-in-strings as numbers, but when you're not a newbie, you begin to run into issues where it's expecting you to be stupid and as a result ends up being stupid itself, and causing you to write disgusting checks just to make sure things are sane.
And that's the root of the whole issue, I think. Things that bend over backwards to cater to newbies end up doing a shitty job for people who have a clue.
Also, if anyone knows of any projects (with source available) written in PHP that are designed well, I would be interested to hear of them. I looked at a tiny bit of PHPMyAdmin's code just for kicks and was horrified. But that's not a valid judgement by any means. Again, I'm just interested to see if there are any out there (and have a look at them).
And a last thing, I might be biased by knowing Python (but I knew PHP first!). Python tends to flow very naturally for me, and even big complex things just end up being big and complex, instead of big and complex and A COMPLETE FREAKING NIGHTMARE like big things in PHP tend to wind up (for me).
Go Python! Death to PHP!
Cold Fusion is also a web scripting language, but costs quite a bit.
.net (a proprietary collection of patent encumbered programs and methods) even more so.
The last time I compared the two (admitedly a while ago) PHP had many more features and was the much better choice. (Even if they were priced the same, which they are not.)
Even more telling is the amount of books available for each. There are seven Cold Fusion books still in print according to Amazon. (Most the same book for different versions of Cold Fusion.) A search for PHP gets 112 hits. (I am not certain how many are still in print. Much more than seven.)
Comparing Java (a general purpose language) to PHP (a web scripting language) seems to be a bad comparison. Comparing it to
"Trademarks are the heraldry of the new feudalism."
The PHBs here view php as something "geeky" that isn't suited for business. I'm sure they'd lap it up in a second if it were called MS Visual php Studio, however.
One thing you should keep in mind about programs from MS is that although they are crap, they offer someone you can speak to on the phone 24/7 as opposed to us geeks chopping things up or finding a forum, or jumping on irc to fix things up or create something. Microsoft is pretty and CTO's, CEO's, CFP's etc., need to be able to understand a product somewhat. I've used PHP for some time, and from my perspective is, there are too many hands in the pot spoiling the food.
Seems like every other month some new and improved programming language comes along and becomes the standard or some future standard. From a business point of view, I would rather go with what is established as opposed to what is promising. Promising isn't going to speak to a CTO through some task should his IT department walk out. Aside from that, standards already around are accompanied by people who get certified to perform these tasks. Now we know not all certs mean squat, but it's easier to find people who follow standards than those who follow promise.
MoFscker
If you are interested in the scripting language comparison, see Server-Side Scripting Shootout.
I could not justify my existence if I were a turkey farmer. Would I terminate myself? Undoubtably, yes.
.NET has COM+ for backend systems, Java has EJB, but PHP has nothing :(
I can't use distributed transactions, transparent failover, declarative security and transaction demarcation in PHP.
IMHO the good parts about PHP are also the bad parts. ie, * you don't have to say what type a variable is, but that means you can't specify a type of parameter to a function. * you don't have to specify scope, but then you can't protect functions that should be private etc.
I looked at a lot of Java code for ideas on what I could do with PHP to clean it up . The main things that I did were: set up a 3 or 4 tier architecture.
- database abstraction layer
- business layer
- presentation layer (preferably using templates)
(I modeled a lot of this on Enhydra - www.enhydra.org)never use globals. Wrap up the HTTP_GET_VARS, HTTP_POST_VARS etc in a class (ie Request). Create classes to wrap the server vars and whatever else.
Use classes for everything. This gives you a reasonable amount of namespace control.
Never access variables directly in classes. Create accessor methods for them.
I think that if you are feeling the need to structure your PHP, you will probably need to move towards Java or some other more structured language. It can definitely be more challenging to write, but as your applications get bigger, the compiler-enforced type checking, programatticaly enforced/supported interfaces etc will save you a lot of time in the long run.