New-age Transactional Systems - Not Your Grandpa's OLTP
John Hugg discusses high volume transaction processing applications with high and low frequency profiles, and how VoltDB can be used for that purpose.
The content has been bookmarked!
There was an error bookmarking this content! Please retry.
Posted by Rob Thornton on Mar 08, 2007
Guice is a new open-source Dependency Injection framework for Java 5 that is closing in on a 1.0 release. Guice is a very annotation-driven, lightweight framework that provides an alternative to Spring, for a certain set of features.
Guice, which is the depenecy injection engine in XWork, focuses on being small, fast, and requiring minimal code to use. The simple example from the Users’ Guide is:
With Guice, you implement modules. Guice passes a binder to your module, and your module uses the binder to map interfaces to implementations. The following module tells Guice to map Service to ServiceImpl in singleton scope:
public class MyModule implements Module {
protected void configure(Binder binder) {
binder.bind(Service.class)
.to(ServiceImpl.class)
.in(Scopes.SINGLETON);
}
}A module tells Guice what we want to inject. Now, how do we tell Guice where we want it injected? With Guice, you annotate constructors, methods and fields with @Inject.
public class Client {
private final Service service;
@Inject
public Client(Service service) {
this.service = service;
}
public void go() {
service.go();
}
}The @Inject annotation makes it clear to a programmer editing your class which members are injected.For Guice to inject Client, we must either directly ask Guice to create an instance for us, or some other class must have Client injected into it.
Bob Lee, one of the project leads of Guice, has written a comparison piece of Spring and Guice. In it he notes that Guice and Spring look very different, as Guice is focused on annotations. Guice can be configured with strings, as in Spring, but they don't encourage it. Guice requires Java 5 and supports generic types.
Improve Java Garbage Collection, Runtime Execution, and JVM visibility with Zing
Using Drools? See what you're missing! Get the Power of Drools with the Assurance of Red Hat
Monitor your Production Java App - includes JMX! Low Overhead - Free download
Congrats on this work, there's some really good ideas there :-)
Hey guys,
This sounds interesting. Any links to where I can find any Guice+Struts2 integration docs/examples?
Thanks,
~Matt
i don't want to start a flame war, but how does Guice compare to Hivemind 2 (in beta) and Tapestry IOC (part of Tapestry 5). Both are fast and rely heavily on annotations too. Tapestry IOC for example is much more "lightweight" as it does not require to implement an interface for a module at all. Also Tapestry IOC configurations (returning new services) are much more natural to plain old java code :)
g,
kris
I think that there is enough room in this space still to have different implementations running around. It'll keep everything nice & honest in what is still arguably a fairly new programming paradigm for java. I for one am glad to see multiple versions that we can all learn from and evolve with. Er something like that anyways ;) Besides, we probably don't want to mess with Bob too much.heh...
+1
As much as i like Spring, i don't think it's good for innovation if there is no competition. So hopefully there will be 1 or 2 things for Spring to learn or give them new ideas.
Apart from that i think Spring is still very much useable and easy especially with tools like SpringIDE.
I have always felt that an DI container should be included in the JDK. I have thought about suggesting that Spring's DI container be "standardized" but feared that such a suggestion would be perceived as a sabotage against Interface21.
On a very practical level, I would feel a lot better to do animport java.lang.inject.Inject;
thanimport com.google.inject.Inject;
in my components.
On a reuse level, this could eliminate the duplications in some of the half serious dependency injection frameworks in many other frameworks.
There's a small example Struts 2 application in the repos and we talk about it in the user's guide.
Keep an eye on JSR 299. :)
Doesn't Hivemind use string identifiers? Also, "module" is the Guice equivalent of a configuration file. Guice does not require interfaces.
Regarding performance, I have a small benchmark here. If you submit a Hivemind implementation, I'll happily include it.
I think the same things in Spring as you and Guice look like a great alternative, well for DI anyway. Sometimes I think Spring is the new Struts in it's domain. It has similar timing and popularity, and as time progesses it sure seams like there must be something much, much better. I'll be curious to see if Guice will be able to foster a community and the like to make it a viable long term alternative for rest of us.
Congrats on 1.0!
looking forward to many more,
Steve
I (think that I) dislike the same things in Spring....
I have implemented a small avalon container called pulga (only 20k, small set of dependencies).
Pulga works like an object factory (like pico), using annotations (@Dependency) and conventions. Here is an example:
//Create a component
public class MyComponentImpl implements MyComponent
{
@Dependency
private SomeOtherComponent ref;
}
public class MyTest extends TestCase
{
@Dependency
private MyComponent ref;
public void testComponent()
{
assertTrue(ref instanceof MyComponentImpl);
}
}
More information at: black-beans.com.br:8088/xingu
JSR 299 looks interesting, and I for one would welcome the idea of a managed "web bean" for data driven web apps. What I fear based on the public material available on the JCP web site is that this will be too closely bound to JSF. Can we be sure that any implementation of JSR 299 does not require the use of JSF on the web tier?
I like the programmatic approach used by Guice to configure modules. We have been using programmatic configuration since the start in Mentawai.
Check how you can do IoC and Auto-wiring with Mentawai very easily, with no need for an external IOC framework. You are always free to use a external framework, but check this first:
recipes.mentaframework.org/posts/list/8.page
recipes.mentaframework.org/posts/list/14.page
John Hugg discusses high volume transaction processing applications with high and low frequency profiles, and how VoltDB can be used for that purpose.
Kevlin Henney examines code samples to see what can be learned from them starting from the premise that one won’t write great code unless he knows how to read it.
Jason Ayers share the observations he made watching a team of developers collaborating in real time on the same code base, pushing XP, pair programming and continuous integration to their extremes.
Michael Snoyman presents Yesod, a web framework written in Haskell and containing a web server, templating, ORM, libraries (templating, gravatar, etc.).
Richard Kreuter and Kyle Banker on how to avoid classical RDBMS transactional systems by using compensation mechanisms, transactional messaging or transactional procedures.
Attila Szegedi talks about performance tuning Java and Scala programs at Twitter: how to approach GC problems, the importance of asynchronous I/O, when to use MySQL/Cassandra/Redis, and much more.
One category of risk that project teams need to ensure they address is business value failure – delivering a product that fails to provide value for the business investor.
InfoQ spoke to the authors of Software Systems Architecture on a couple of new topics, the System Context viewpoint and Agile, which have been added to the second edition.
14 comments
Watch Thread Reply