Why I Hate Frameworks
discuss.joelonsoftware.com
discuss.joelonsoftware.com
What's sad is that I have no idea if this is a joke or not.
When Polar Rose still used XML-RPC we rolled our own framework that integrates well with Spring. You basically expose your POJO and be done with it ...
EDIT: OK, so instead of Java, that should read "to know if enterprisey Java frameworks are a good fit for their problem." The irony is, that I just was IM'ing a friend about how this isn't the fault of Java the Language, but the "enterprise architecture" community that write Java frameworks like this.
After 8 years of Java, I personally found it really refreshing to build stuff in Python and Ruby.
Document doc = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse("test.xml");
I wonder if they ever thought of providing two interfaces.
1. The first is the common case.
Document doc = new DocumentBuilder().parse("test.xml");
Document doc = DocumentBuilderFactory.newInstance(opts).newDocumentBuilder(opts2).parse("test.xml"); Document doc = DocumentBuilder.parse("test.xml");
The purpose of "assisters" like frameworks and higher level languages is NOT to make good progammers more efficient.
It's to make mediocre programmers more likely to produce something of value and to make poor programmers capable of producing anything at all.
And if the bell curve tells us anything at all, it's that these "tools" target 90% of all programmers.
But think about it, my fellow top 10% (lol), do you really need all this stuff? If you're working alone or on a small team with a clear objective, haven't you always had everything you needed with low level tools? If you need any higher level tools or reuseable components, haven't you already been building these all along?
Sure it's fun to play with new things and learn from others, but when it comes time to really produce, don't we all know how to (and need to) roll with what we know?
The need for frameworks and high level languages only becomes apparent when we grow so large that we can't find enough senior hackers. Only when you dip down into the mediocre masses do you need this help.
True, but they can invent the wheel because there isn't one, and they can do so in a way that others will find it useful as well. It is a different point than edw519 is making, but what he said made me think about where I disagreed and why. The purpose of code libraries, is to make everyone more efficient, by solving a problem so others don't have to. If that means a programmer can produce twice as much functionality within a given time frame, or is able to produce any functionality at all where he couldn't otherwise, there is still a net benefit for everyone. This should not be scoffed at; it is the reason why we are able to produce software as featureful and useful as we do. Imagine each of the XFCE, KDE, and GNOME teams having to rewrite X.org all over again, and how much less each of those desktops would have to offer right now if they did. I'm willing to trade useful functionality sooner for less machismo.
But libraries don't always achieve this, and the degree to which it doesn't determines which people will be motivated to create a replacement. If a framework solves a common problem, but does so in a painful, unwieldy, or hard to understand way, some hacker out there will probably think it a better, more efficient choice to just write and use a new solution. It might be a fork of or series of patches to the original framework, or it may be an entirely new codebase -- which of these routes to follow is a judgement call. Nevertheless, it requires what I think is an important distinction, which is the ability to not just understand what problem the framework solves, but how it solves that problem. And not understanding it just in following what each line of code does, but why what steps it takes work. At that point, you end up leveraging a framework far more that relying on it. If something tomorrow were to happen that required you leave it behind, it would be a pain in the ass, but not a brick wall.
There is also the question, more important that the above in my mind, of understanding if you actually have the problem the framework solves, or if you are attempting to shoe-horn. But I worry that might be a bit off-topic of this particular thread, and would probably warrant another 3-4 paragraphs of musing.
And what is wrong with that? Some of us aren't the programming gods you are. If it helps people why get so high and mighty about it? I know with your superior intellect us philistines might seem dumb to you, but maybe you should stop writing about how smart you are when you can't even spell "programmers" right.
This was partially in jest (notice the lol), and I realize this is sometimes hard to convey in writing. Sorry.
But there's also a lot of truth in it, too.
No one is born a mediocre programmer. It's a choice.
Notice I didn't say a "junior" programmer. There's a difference. A junior programmer has less experience, but has a good attitude, is willing to learn, and practice his or her craft to get better. A mediocre programmer is just lazy and thinks he or she knows better.
This was directed at the thousands of lazy thoughtless programmers I've had to clean up after over the years. You know the ones. They refuse to follow accepted standards. They name their variables whatever they feel like at the moment. Good luck figuring out what "a" "aa" or "aaa" mean. Or locating them in a text editor. They use single entry, multiple exit. Good luck changing one line of code in the middle of that. They rewrite the same functions over and over, so I have to change all 14 versions. And spend the next 2 weeks regression testing.
I only wish these people were required to use a framework that forced them to conform to anything that would have made my life a little easier.
If "high and mighty" means "wants to get work done" or "cares about others", then thank you. Some of us choose to write clear, effective, easy to maintain code without a framework. If you ever have to maintain my code, perhaps you'll understand.
Yes. I am not a great programmer. I only started caring about it a couple of years ago and since I don't have a real education in it I am very behind in many areas. I am trying to get better, but it's not easy.
"No one is born a mediocre programmer. It's a choice."
I don't know if I agree with this. Some people are born with incredible talent in various areas. Some people are just born musicians, others athletes, others programmers. There is no substitute for hard work, but some people will never be able to grasp what others can do. No matter how long you play guitar, you're not going to be Jimi Hendrix.
I understand your sentiment, I really do. But I would be careful of calling misinformed people lazy. I would bet good money that there are "standards" I am breaking with my code that I don't even know exist. If the books I read or the sites I visit haven't told me about them, I wouldn't call myself lazy.
In the end, HN is a community of amazingly talented people who are, to me, beyond brilliant. I think it's easy to lose sight of the "real world" some times when there are so many smart people in one room. Just have a little bit more slack for people who are trying and not as good as you. I think you will find it explains a lot.
"Some of us choose to write clear, effective, easy to maintain code without a framework. If you ever have to maintain my code, perhaps you'll understand."
Do you "choose" to do that, or are you benefiting from the years of education and experience you have? It's not like I can wake up tomorrow and choose to be an expert programmer. I can only try to keep learning(which is why I hang around here). I completely understand your frustrations with work, it can be a pain in the ass to deal with people who make your work harder through their incompetence, but I don't think it's as deliberate as you are suggesting it is.
A nice thing about the programming world is that it's a sort of natural democracy. No one really knows how good they are.
I second that!
You've just put yourself ahead of 95% of the millenials here by saying that. The only way to become great at ANYTHING is to admit that you aren't, and work on improving yourself.
Being junior isn't a bad thing. Everyone starts at the beginning, even the prodigies. The bad ones are juniors that attempt to skip the learning process by claiming to be great, not the ones who do like you are and put in the effort to learn.
The difference between you and the typical millenial, at least the ones that I've had the misfortune of working with over the years, is that they think they're good, and don't listen to the folks around them who've actually DONE the work before.
"I don't know if I agree with this. Some people are born with incredible talent in various areas."
The only talent humans are born with is the ability to construct their own future.
Those who lack talent and experience often believe that success is predicated on luck. They see spectacular photographs taken in times of perfect light, and say, "Wow, you were so lucky to get such gorgeous light right when you were there with your camera!"
The smart ones eventually come to realize that you make your own luck.
I think if you hang out with "programmers" long enough, you'll learn that the worst ones spend the majority of their time complaining about the shortcomings of others' "programming" techniques or methodologies. In reality, there is no de facto standard modus operandi for the nirvana of any code, and if you want to learn, it's best to just keep an open mind.
Who isn't? Don't sell yourself short.
No matter how long you play guitar, you're not going to be Jimi Hendrix.
You don't have to code like Jimi Hendrix to be a great programmer. (Oddly, I suspect that if I had to maintain his code, I'd probably be going nuts.) You just have to get your job done, work well with others, and leave something maintainable behind. I have a feeling you're already there. I also have a feeling that everyone reading this has done something cool that neither one of us has ever even thought of. In our field, not knowing something isn't a problem, it's an opportunity.
I understand your sentiment, I really do. But I would be careful of calling misinformed people lazy. I would bet good money that there are "standards" I am breaking with my code that I don't even know exist. If the books I read or the sites I visit haven't told me about them, I wouldn't call myself lazy.
I agree completely. I'd give you an extra up vote for that response if I could.
...are you benefiting from the years of education and experience you have?
Of course. We all are. Not benefiting from any education or experience you have oughta be a crime.
It's not like I can wake up tomorrow and choose to be an expert programmer.
Right. Give it a week or two. If you code as well as you comment here, you're probably way better than you give yourself credit for.
In the back of my mind, I thought there was a chance my remarks would be controversial when I first entered them. Sometimes it's fun to push the envelope here. (Remember, I'm in the middle of Day 4 of regression testing someone else's crap.) You proved me right.
As far as OP is concerned, I still think of frameworks and 4th generation languages as "crutches" for a lot of us. So they serve their purpose. But at a cost many others have talked about here.
Thanks for the great discussion, chez17. I'll make a point to look for you in future posts.
So I guess I'm just an ambitious junior programmer? Which is to say, I want to get better, but I also want to make my project production-ready as quickly as possible, so I use a frame and fill it out.
So I think it's a mistake to say that frameworks are only useful for making bad workers produce work. Better to say that there's a time and place for everything, and that while frameworks are overused, they're useful for certain people doing certain tasks.
ex. for web apps it prevents people from putting everything on one page (DB calls, logic, HTML, ...)
It is the slow scream of death march.
I suspect that your large project was built by "Java Engineers" who tried to stuff way too much into the ORM layer. That's an abuse of the tool, not an indictment of the tool itself.
If the details of passing data over HTTP and interacting with a database aren't a critical part of my project, I'm not going to do it by hand. I'll find some existing library or framework to do it for me. It's not that I can't produce something without help. It's that I have a problem I want to solve, and it isn't how to communicate over HTTP or store and fetch my data efficiently.
If it makes sense to hack up a project in assembler, do it. It may give you a massive advantage.
My comment illustrates the fact that criticizing those two things in particular is beyond silly. Essentially, I'm saying: Where is the limit? Unless you code everything, from scratch, every time, in assembler... You're using a high level language or a framework.
And you should be... The main point of them is to save time. They don't solve all problems obviously, but they solve the kinds of problems that you don't want to solve over and over and over.
Why build your own framework when there's one already there? Already well-used, well-tested, well-extended?
Sure it may be easy enough to do, but is it really worth it?
I, for one, would much prefer to be spending my coding time writing application logic than Yet Another Validation Library.
Suppose someone is familiar with Java Servlets and JSP, basic JDBC and SQL, and the tomcat container (with connection pooling). If I have a simple database driven application, I can choose to hand code some of my application logic, or use a framework.
So I'm choosing between, say:
No framework: container managed security, hand coding my ORM layer, writing my own server side form validation, writing my own MVC (or just not bothering, if it is just a few pages).
Framework: Spring Security, Struts for form building and validation, Spring for the MVC tier, JPA and Hibernate for ORM.
These are slightly extreme - you don't have to use all those tools, of course.
But let's say we do get something out of the framework. The code really is easier to maintain and understand, provided someone has read a 300 page book on spring, another 300 page book on hibernate, and a 150 subsection of a book on Struts, along with doing all the exercises.
Even if the framework based code is better, that's not enough to justify the complexity. I need to justify taking an application out of the realm of someone who understands the basic servlet spec and putting it into the realm of only those who have read and digested an additional 600+ pages of Framework complexity.
Thing is, I like frameworks. You can read and digest 100 pages about Rails and really get going. So the bar for justifying the "added complexity" of a framework in the Ruby world is much lower than it is in the Java world. I don't use Python, but I'd imagine it's a similar situation - it's much easier to justify Django than the Java framework stack.
Good frameworks are the same way. I've already extended the hell out of Kohana to make it more useful to me. This is work I would have had to have done if I weren't using it, sure, but all the basic stuff is there all the same.
Would you rather build a porche, or try to convert a clapped out old volvo into a porche?
Tidying up rubbish, bad architecture choices, stupid coupling etc isn't fun. It's twice the work - fixing all the stuff they did stupidly, and then implementing it how it should be done.
Why learn woodworking when you can buy furniture at Walmart? Some people like craftsmanship and others like frozen dinners.
For many developers who toil in the same language all the time, or who have spent many hours in whatever language they're currently in, that's the case.
For me, I avoid Rails because it makes me feel like a failure. ActiveRecord, for one reason or another, hates me. I like Rails, and I like the Framework concept (however limiting it might be) -- because I feel that it gets you started more quickly. At the end of any significant project, I feel like you will have likely massaged enough elements of the framework to make it work for you that it's hardly recognizable, but it still got you something to show people that much faster.
That said, ActiveRecord is the showstopper for me in Rails. It seems like I've either never properly wrapped my head around it, or there are fairly low limits to what you can do, or I'm doing everything wrong. What that's always meant for me is that instead of using ActiveRecord to its fullest, I'm building SQL queries manually. And if I'm deviating from the framework's toolset, why am I using this framework in the first place.
This is NOT a knock on Rails, I'm probably just doing it wrong. Rails and I can still be friends... just not in that way.
I've always liked the 'middle ground' that Zend offers, in that it abstracts away the connection and handle, and gives me a convenient place to start building the model code. It's not 100%, but it doesn't make me feel guilty in the morning either.
I find it's just way, way easier to get up to speed with a framework as a base, and if your framework is any good, then it should be easy to shove aside later when you need to special-case things for performance or whatever.
You make this sound as if you are the great programmer who does not use frameworks but instead reinvents the wheel every time.
Are you really sure you do not use frameworks or libraries in your own work?
I also think Joel misses something: most programmers are bad programmers. They'll dig you into holes, write lots of spaghetti string code, etc. At least when they're using a framework there's some constraint on their madness. Frameworks (generally) make it a tad difficult to work outside of their constraints and so the default becomes to work in that standard way. So, you don't get people doing weird things as often - like building a string that is the page that they'll echo out like the piece of PHP I'm working on today does (guess they didn't realize that you could simply go out of PHP mode to do that).
:-)
Just FYI. This was post on the "Joel on Sofware" discussion group but the author is Benji Smith.
I like his blog. He used to post a lot more but he's been busy with a new project. (and a HD crash as of this morning.)
You should hang around better programmers.
Although it is less about frameworks and more about a disease which was largely contained to Java enterprise-y architecture stuff. (I say as a practitioner in that field... remind me why we are using the Factory pattern here, again?)
Alternately, you could use something like Google Guice, but that's (sorta) a factory-factory.
Basically, no one likes the situation, but they've chosen their tradeoffs, which they see as the lesser of N evils.
Because everything in J2EE, including design patterns, is driven by salivating greed, megalomania-level egos and lemming mentality (I have a similar comment in the original thread).
That may be the most unproductive I have felt in my entire life. I made a damn login form and it took like 12 pages of tutorial, tons of setup, generating a whole bunch of files that seemed like they could have been done algorithmically because they are always the same, editing like nine different XML files with little bits of glue, and eventually like one or maybe two lines of code.
Then, I got to wait for the WAR to compile each time I wanted to test a change, because the tutorial didn't seem to know about a developer mode which probably exists.
I hate to sound all ranty and partisan, but I could find no way in which that process could be efficient and couldn't imagine anyone doing it voluntarily unless they were ordered or just had no exposure to the rest of the world.
Also, I sort of object to having code stuffed in an eight-folder deep hierarchy where the preceding seven folders contain nothing at all. That's a lot of wasted navigation and clicking time and distraction.
Why do I need to have 5 different objects when one would do? Having 5 different objects did absolutely nothing for the client code except make it more complex. (And this actually complicates good coding for the user experience. When we notify the user of network problems, I have to add additional code to see if objects exist in the first place just to prevent exceptions.)
It would've made no difference if there was just one Object. (Except for shorter code.) I could use it to login to a server. I could use it to register interest in a topic then ask for messages on that topic. If I want to connect to a different server, then I could instantiate a second object. Why do I need 5 freaking instances to do the most basic thing possible?
Grails is basically Spring MVC, wrapped in dynamic language goodness. I'm using it for a project right now, and it's shockingly terse and clean. It's also a lot of fun, which is something I never thought I'd say about a Java "Enterprise Framework".
I think that is because a lot of people simply don't know what good design is. I've seen many projects where people think they don't need Spring but where it made 100% sense. And made the project simpler and easier to manage and test.
Even in small projects it makes a lot of sense to have a good seperation in layers or services and to use a mostly invisible helper to glue things together.
Modern Spring can configure your app with a very minimal amount of XML. If you don't like XML then use annotation based configuration.
I totally disagree with your statement that Spring only exists to manage the 'massive configurability of Java design patterns'. The only patterns that the foundation of Spring introduces are IoC and 'configuration by convention'. Those have been proven to be extremely powerful patterns that simply allow you to burn half of the 'J2EE Design Patterns' books.
It really does not help that all of our web projects are struts 1 and we have still have a fair amount of torque projects hanging around.
Yes, in a SPECIFIC area of the Java space.
There is a huge group of people who believe that J2EE and its complexities are a major pain the ass. That is why great frameworks like Spring, Guice or Stripes exist.
Sun loves you when you use J2EE/JSF/EJB/JMS/JWHATEVER. But there are serious alternatives that are more lightweight, easier and less intrusive.
Of course you will not add value (read: buy into) to the Sun Enterprise Ecosystem :-)
Staying away from (most) 'Enterprise' Java standards has worked very well for me. I really enjoy working in Java.
Someone comes up with a tool that's great for a specific job. Later, someone else comes along with a slightly different problem and the tool changes a little to accomodate them. Then a few others. Before you know it every feature under the sun is added to handle every possible scenario.
Now you're stuck with crappy middleware. It does a decent enough job after you configure it properly and strip it down to only the modules that apply to your specific problem. It's not so bad because you've been using it for a while and you know it inside and out. Plus, it can be used to solve everything. It can probably even cure cancer with the right module and proper configuration.
But sooner or later someone else comes along, decides this nifty universal tool is a steaming pile of shit and sets out to invent something simpler and less bloated. And it will only be better because it doesn't have everything, including the kitchen sink, bolted onto it. It's focused and lean. But then someone comes along with a slightly different problem than what it was intended to solve and the hot new framework with all the buzz starts going bad just like the last one.
Look at Stripes. It's just a better version of Struts2 which was based on Webwork which was intended to be a better Struts 1. Eventually, more and more cruft will be added to Stripes so it becomes as bad as what it was designed to replace and someone will come along and invent a better Stripes.
In my opinion, this is the biggest problem with frameworks. Many developers prefer to have a universal tool instead of having several that are good for different jobs so we end up with this cycle of cruft.
The biggest problem with the entire stack is that every piece would rather overengineer a solution to a small part of the problem, instead of giving us a simple solution for the whole thing.
Look, just have a stock web.xml, and don't touch interceptors. Instead, put together a single, coherent system for dispatching a URL to the right object, and the same for security requirements. If you need more, abstract once, there for the URL. That's it.
After months of reading through crap documentation (Hibernate, I'm looking at you), it's obvious that the frameworks are 90% of the problem, and your actual set of requirements are the last 10.
Why develop my own routing code when there's already a robust, well-tested version available? Or validation code? Or payment authorization code?
Though it needs to be said that the number one thing a good framework does is GET OUT OF YOUR WAY. I've seen some Java frameworks that take an hour to put together a simple Hello World and that's just nonsense.
Currently I'm using Kohana, which is nice because it's really very simple and easy to extend. So far I've been able to spend so much more of my time actually writing business logic instead of routing logic and the like.
Oh, and if you need them to be different from the default, you can modify them as you want, but we think the default works for most people.
Those bits just magically work!
:)
If you're making a spice rack:
Without framework: DIY shed filled with tools
Framework: IKEA
Frameworks are tools, not end products.
IKEA provides flat-packed products. You never buy a wardrobe at IKEA and then turn it into a couch instead. It will always be a wardrobe, and nothing else.
On the other hand, if you buy a set of tools and environment (aka framework) to build furniture, you can use it to build whatever furniture you want/can.
Any flat pack spice rack, will be shoddy quality, and probably break.
You don't get good furniture by mass producing it like that. Same with software.
What if I want a custom spice rack, with some secret compartment, in solid oak.
Next, you'll be suggesting that people write their own web server instead of using Apache? Or perhaps they can trust other people's code so long as they pay for it, so you'll send them towards IIS?
Come on, now.
If the webserver is a key component in your success, then definitely. Write a webserver. It was for me, which is why I wrote my own (After I found others lacking).
If it really matters, and the existing stuff isn't good enough for your use-case, write it yourself.
"And furthermore, for me, programming languages and their runtimes are tools. I think they are for a lot of us. You wouldn’t continue to use a hammer whose head keeps flying off when you take a swift backswing, would you?"
Good read: http://cbcg.net/2007/04/22/python-up-ruby-down-if-that-runti...
I am discounting Rails here: I think Rails is great for what it is and I’ve had nothing but good experiences with it precisely because it is so “opinionated”. Its a very productive environment and manages to successfully avoid the runtime problems that I’ve personally hit in the past.
So... are you trying to make a point about Ruby or about Rails? Dare I give out the clichéd "Rails != Ruby" response here?
Rails: Here's a kick-ass hammer, a kick-ass saw, a kick-ass measure tape, etc...
My point is that those kick-ass tools may be made of balsa. The editorial that I linked immediately came to mind because it too used tools as an analogy.
The bottom line is the Ruby runtime has some serious flaws, but they are not so pervasive as to make everything in Ruby total crap. There's a good chance you could write an application in half the time it would take in Java and experience no problems ever. It's also possible you'll hit a memory leak and have to face an ugly workaround, partial rewrite or even scrap the whole Ruby project. Finally, you may have a memory leak and find an easy workaround, and then one day the problem is magically solved by Rubinius.
Those are concrete scenarios to consider. I don't see how a physical materials analogy has any value in analyzing your options. The pros and cons of languages don't align AT ALL with physical properties. Thoughts like, "Java is more like cast iron, and Python is more like hardwood" is just one-dimensional mental masturbation. In the end the reason to choose a language like Java because you are willing to tolerate more overhead for less risk.
That sure sounds like personal conviction to me.
The rest is funnier. Give it a try.
I'd say a framework is more like a factory that produces a variety of fixed-size spice racks (among with other products). You have to learn what button to press to get the right ones, and the spice racks produced will never be exactly how you want them to be.
If you want to quickly generate lots of generic spice rack, then a spice rack factory is all you need. But if you want something really outstanding and unique, then a factory may be anti-productive, and you are better off building it with your own hands.
They're necessary, but they're hard to do right. The man who did Powerplant said he went through about a dozen different designs before getting one he liked. That's why it was often so useful. Most times, people say "hey, let's take these APIs and let people subclass! yay!"
And that's how you end up with crap like doGet(), doPost(), etc.
What made WO great were the tools. There was still complexity there, but WOBuilder and EOModeler made the whole process very productive. You could think of your data model, your interface, and your business logic independently of each other, and then use the tools to hook them together in a way that made sense. Because of the way the tools were designed, they actually made using good design for your application easier than using bad design.
That's the main thing missing from all the frameworks I see today, integrated tools that automate the tedious parts and encourage good design practices.
To me, the WO tools go on the heap with the Lisp Machine, the Xerox Park Smalltalk environment and HyperCard, as old development technologies that are still ahead of today's state of the art.
EDIT: spelling correction
I get it's a joke, but violence against women, when used in a semi-professional context in a community of software engineers is one of those things that keeps more women out of the field in the long run.
I'm not trying to be PC intentionally -- it just struck a raw nerve for me.
"What I'd really like to find are some appropriate libraries that I can use to provide several kinds of functionality for my project."
And then lists a bunch of high-level website functionality requirements (user accounts, content management, etc). I realize that the J2EE landscape was drowning in complexity and was pretty dismal in 2005, but this guy doesn't understand the issues involved in developing web applications.
First, to suggest that individual libraries can provide re-usable high level web application code is completely ludicrous. It can't work for so many reasons:
* All those libraries will have to interact, and the interface between them will be complex. If they do work together they will end up being tightly coupled and before you know it you have a framework.
* High level functionality that includes data models, persistence, and presentation layers is not generic enough to effectively be a library. The complexity and configuration would quickly balloon out of control before it met even the requirements of 5% of web applications.
* Such libraries would end up being far more constricting than a traditional framework. A good example is Drupal, which is a very powerful framework/CMS with an impressive hook system, and an extremely high code to functionality ratio. Anything you might want to do with a website can probably be done in Drupal using a combination of existing modules, and you can write additional modules that access almost any part of the system and modify its behavior. The problem is that fine-tuning of the interface or certain kinds of customizations become very difficult under the crushing weight of the underlying assumptions that Drupal makes to enable it's magnificent generic-ness. The result is that all the various sites made with Drupal end up feeling very Drupal-y and there's a disincentive to really design your UI from the ground up and create a very optimized application.
What J2EE has blinded the author to is the fact that web frameworks exist to solve the most common problems that come up over and over again in web development. He needs to go write a few PHP applications to get a feel for the issues, then try something lighter-weight like Django or Rails that confronts those issues directly and then basically gets out of the way. In the first half of the decade it was hard to find a good lightweight framework because people were still wrapping their heads around the actual issues that needed to be solved. Nowadays, not so much.