Tuesday, January 23, 2007

Can Someone Tell Steve Jobs About OSGi?

Steve Jobs in an interview with Jon Markoff: “Java’s not worth building in. Nobody uses Java anymore. It’s this big heavyweight ball and chain.”. Mr. Jobs is known to use the hyperbole, but I am afraid he has a point. I remember having a fiery discussion with Mr. Smalltalk, Ralph Johnson, in ’97 about how nice Java was. One of my key arguments was that Java was so lean and mean in comparison to Smalltalk. He shook his wise head and told me to wait and see. Oh boy, how right he was. Since then Java’s library growth has consistently beaten the growth of even the best hedge fund. Time to reflect and refactor are words one can not easily associate with the Java world. The JCP has given us a hodgepodge of inconsistent APIs split between worlds that seem to diverge farther every day.

Yes, I know. Four billion phones or more run Java, generating billions of dollars of revenue. However, everybody in family has a Java phone and non uses Java applications. When you look at what people run on their telephone then it is mostly games. Where are the killer apps? These apps would appear in the market when the operators and manufacturers would allow third parties to access all the resources. Of course, JSR 232 provides a very interesting architecture for interesting applications that will enable these types of applications, but phones are only just now appearing in the market.

Sadly, Java largely failed on the desktop. How many pleasant applications do you use that are written in Java? Eclipse is still the exception. Java seems promise so much for the embedded market but it is still a hard to sell Java in a consumer device due to the extra cost of the VM and increase in resource consumption. Java only shines in the server market, where resources are plenty and inefficiency is a minor inconvenience.

Why is the situation so bleak? The “Write Once, Run Anywhere” mantra (WORA) of Java is highly attractive. The software industry spends countless billions every year porting applications from one environment to the other. Application developers limit themselves to fewer markets because the cost to develop for additional platforms is just not worth it. There have been significant problems in the past with the WORA (write once, debug anywhere israther tired joked repeated ad nauseum by Microsoft) the joke), but today it works amazingly well across platforms, note again the Eclipse success story.

Personally, I think the biggest problem is that the Java architecture is not suited to what it desires to be: a ubiquitous computing environment. A key aspect of a ubiquitous computing environment is that it should be lean and mean at the core. Few want to carry the dead weight of someone else’s requirements. The key mistake Java/JCP made was splitting Java up in different domains: JavaCard, J2ME, JSE, and JEE. Then in different configurations: CLDC and CDC. Then further refined with profiles like Foundation, Personal, Basic, etc. The intentions were good; the result is a disastrous fragmentation.

Lean and mean means you need to be able to mix and match. One of the most interesting Java projects at this moment is therefore Apache Harmony. Harmony is an open source implementation of the Java SE VM and class library. Instead of taking a short-cut, cloning the Sun VM, they use a novel architecture of modularizing the Java class library. With Apache Harmony, you can start really small and add what is needed. Such an architecture permits the use of the same basic environment across industry and platforms. Developers no longer target a specific pre-defined profile but only express their requirements on the platform. When the application gets deployed and certain packages are missing, the platform can be dynamically extend the platform by adding the missing packages.

Modularization creates a market for solutions. It is no longer required to come up with a catch-all API in the JCP, which is too often designed under political pressure. For example, J2EE would never have been a success in a free market. It received its life because it was the “ordained” API for enterprise applications by virtue of its JCP origin. It was not until an external party like Interface21 showed with Spring that Enterprise applications can actually be lean and mean as well.

I think that Java is too big and too important to our industry to be shepherded by a single company like Sun. By trying to shield the environment from fragmenting they sadly have achieved the opposite. By exerting too much control they have enabled some pretty bad APIs. Why not set firm ground rules, that is a small VM and rules for modularization, then let different parties compete with solutions. The market will quickly pick the best practices and create de-facto as well as de-jura standards.

Time to reflect. What do Eclipse, Apache Harmony, and Spring, actually have in common? Hmm, maybe Steve Jobs should invite an OSGi evangelist to give a presentation?

Peter Kriens

P.S. You did register for EclipseCon I hope?

Tuesday, January 16, 2007

OSGi Vehicle Workshop in Detroit

Last Thursday (11/01/2006) we met for an automotive workshop at the Delphi headquarters in Troy with participants from Intel, Wind River Systems, Delphi, Renesas, Vetronix, Booz Allen, Ford, GM, Aonix, Siemens VDO, ProSyst, and Telcordia. This workshop was initiated by Kai Hackbarth (ProSyst). Kai is working with a number of companies in the VII project and noticed that there is a lot of interest in OSGi from the participants in this project.

The VII project has a very ambitious goal: it wants to reduce the number of deaths on the American roads and decrease congestion (this is similar to the CVIS project in Europe). To achieve this goal, the American Department of Transport (DOT) wants to build an extensive communication infrastructure that supports vehicle to vehicle communication as well as vehicle to centers communications based on the DSRC communications service. Vehicles will get an on board unit and the roads will get a roadside unit that manages the communication of nearby vehicles. This infrastructure can be used in many different ways. Some examples:

  • Vehicles will transmit roadside information to central management centers which will be digested and dispatched as traffic information.

  • When a vehicle breaks, it can send out this information to other vehicles

  • Stop lights can send their state to oncoming vehicles

The number of scenarios that can be supported is of course only limited by your imagination. It is therefore that OSGi plays an important role in this project. The life cycle of a car is around 10 years while the life cycle of these types of services is significantly smaller. It is therefore obvious that some form of software downloading and management is required. This is of course the archetypical application area of the OSGi specifications. Maybe the most interesting aspect is that the on board equipment (OBE) can be used for other, commercial, applications as well. This is not a primary goal, but it enables a lot of possibilities!

We had modeled the meeting after the successful Enterprise meeting in August 2006. This meeting resulted in the OSGi Enterprise Expert Group, which will meet in Dublin at the end of this month. The format is simple; some of the participants present their ideas in a 5 minute presentation. This kicks off a discussion where we gather all the different areas that could be relevant for OSGi standardization. This list is then prioritized and further discussed.

We found the following areas of highest interest:

Resource management and hard real time issues
The traditional embedded concerns of resource management and hard real time issues came far ahead of all other items. There exists an OSGi RFC addressing some of these concerns but the focus on server side and desktop applications have moved these concerns to the background. It seems clear that this should be a number one concern for the vehicle group. Interestingly, Aonix presented a very interesting idea. They had create a VM where you could run hard real time applications written in a subset and annotated Java. The code would run in a special memory space that did not require garbage collection. This concept meshes obviously very nicely with the OSGi model of deployment. Special metadata could be used to indicate that specific classes should run on the hard real-time VM.

Vehicle Interface
A vehicle interface provides access to the vehicle configuration and status. There has been work in GST, AMI-C, and the OSGi VEG in this area. Both GST and OSGi use the OSGi Device Management based on OMA DM as the foundation. The complex aspect of the vehicle interface is to model the car. How many chairs, how many axis has each chair, how many windows, where are the seats located. Such an ontology requires vetting by a lot of knowledgeable parties.

Human-Machine Interface. We have been talking about the HMI since day one but we have never been able to standardize a single HMI. There are developments that seem to make the development of a generic HMI more feasible. Normally, the car manufacturers regard the look and feel of the vehicle as their prerogative but it is becoming clearer that the largest number of applications requires some form of a HMI. Therefore, an OSGi standard should address some way that applications can interact with a user while at the same time allowing as much as possible freedom to the car manufacturer.

Non-Java code integration
The demand to expand from Java to native code becomes larger and larger. In the embedded world it is rarely that there are green-field applications. Embedded systems must work with legacy code that is not easy ported to a Java environment. From the participants it was clear that we need a story how to integrate non-java environments with OSGi. As I discussed elsewhere, I strongly believe that our service registry is a very good mechanism to let different subsystems communicate, regardless of the language that is spoken.

There were several other areas discussed, like: Integration with AutoSAR, handling of safety information, vehicle specifics around provisioning and management, a comprehensive driver model, diagnostics, provisioning of data, navigation model, security, a lightweight form of OSGi, privacy issues, reprogramming ECUs (other processors in the car), distribution, and some none technical areas.

A great deal of time was spent on discussing the OSGi vision of universal middleware and how to get large car companies to buy into the technology. Though the basic OSGi technology has proven itself in a large number of applications, it is not well known in higher echelons. Marketing the technology is hard because there is no product that needs to be sold, what we are selling is a promise. To fulfill this promise we need the participation of the key players in this market: the car manufacturers. Without strong backing from the car manufacturers, few suppliers will invest in pushing a technology like OSGi. We discussed how to get support from the car manufacturers (several were around the table) but that is a difficult issue due the state of the current car industry and lack of knowledge about OSGi in the higher echelons. I guess we have to continue as we do today and hope that the combined results of this work will convince the car manufacturer’s management to invest in our technology.

However, it was clear that OSGi plays an important role in the thinking of many participants of the VII project, just like the European GST and the CVIS project. For these architectures the only option is the OSGi service platform, or otherwise limit the standards to protocols only. Though protocols only can reduce the cost of the OBE, it adds cost in other places because the diversity is increased on the management level. It is therefore not unlikely that in the long run the OSGi service platform becomes a standard component in the car.

Overall, this was very good meeting where both technical and non-technical issues were openly discussed with representatives of the key player; there was a lot of energy in the room. We also intend to follow up this meeting with an OSGi Vehicle Expert Group meeting, further detailing the requirements. It looks like the OSGi technology will happen in our industry, now we only have the small detail of making the vehicle specific specifications left.

     Peter Kriens

Tuesday, January 9, 2007

Spring and OSGi: Jumping Beans

This week there was an interesting discussion going on in the Spring-OSGi mailing list. In this mailing list, a number of people discuss their fantastic work marrying Spring with OSGi. This project has created a lot of interest; Even JBoss seems to have been inspired by this work and is moving to OSGi, putting the majority of application servers on firm OSGi footing!

Anyway, the discussion was about the dynamics of services. Hal Hildebrand (Oracle) started a discussion about the cardinality of services. This opened a floodgate of mails that I tried to ignore due to after-vacation overload. However, Costin Leau (Interface21) asked me to jump in so I was forced to analyze the thread. I got myself a cup of strong coffee and wrestled through the mail flood.

The issue was revolving around the cardinality. The cardinality indicates how many services of a given type you require and how many of them you are interested in. The OSGi-Spring specification treats OSGi services similar to beans. This allows you to inject beans into other beans, following the concept of inversion of control and dependency injection. For unary services like an OSGi Log Service this works fairly well. However, the devil is usually in the details, so let us take a closer look.

The key differences between beans and OSGi services are:

  • Beans are passive, beans are non-existent until someone creates them with Class.newInstance(). The whole idea of Inversion of Control (IoC) is that this creation is done by an outsider and not the POJO itself.

  • You rarely require two beans of the same class. Why would you use two different database objects or log objects?

Most Java programmers will recognize this model. Many developers use the Class.forName to find a class and then use Class.newInstance to create an instance for extendibility. Virtually any (non-OSGi) Java application uses this model to extend or configure itself this way. Notice that there is a strong 1:1 relation between the Class and the instance in this model.

In contrast, OSGi services are more like jumping beans; they have a life of their own. When beans start jumping all over the place (spring, as they say in Dutch!) then most programmers will be shocked to the bones. Fortunately, once they grasp the power of a dynamic model they usually fall in love it. However, this model implies that the IoC is not limited to configuration and activation, it means that the IoC container must handle life cycle issues during the active life of the application.

Obviously, many Spring applications will not be able to handle these dynamics, or even want to handle these dynamics. The Spring-OSGi group therefore chose to hide these dynamics behind a proxy. Instead of directly injecting the service object, a proxy to this object is injected into the configured bean. This gives us a choice to control the behavior later when the underlying service arrives or goes away. However, it also gives us a choice of optionality and multiplicity. The proxy mechanism of Spring gives us full control, the tricky thing is to define what to do with this control.

These are exactly the issue we spent a lot of time discussing during the development of declarative services, inspired by Richard Hall’s service binder and designed by BJ Hargrave (IBM). Declarative services do not use proxies but instead bind and unbind services to the application using injection. After lots of deep discussions we came up with the following policies for binding and unbinding services to the application.

A mandatory service means that the service must be present before the application is activated and the application will be deactivated when there is no such service available anymore. An optional service does not influence the life cycle of the application. The declarative runtime can unbind a service and bind another service at will when the service is optional. Normal Spring beans are similar to mandatory services.

Mandatory is normally indicated in the cardinality as 1.., and optionality as 0.. .

The application can be interested in a single service (unary) or it can track a set of services. The set may require some explanation. The dynamic service registry enables some very interesting software patterns. For example, one bundle manages a Bluetooth hardware interface. The whiteboard pattern shows that the simplest solution for this bundle is to register a service per discovered device, and unregister this service when the device is no longer reachable. Other bundles can then just track these services from the registry and use when required. This pattern nicely decouples the lifecycle of the bundles that participate, making clients easier to develop.

The idea that you can add and remove beans dynamically from a configured bean is new to Spring but it follows naturally from the OSGi model.

This cardinality is indicated with ..1 for unary and ..n for multiple.

Once an application is activated because its dependencies are met, the lifecycle state of a referred service can change. The traditional choice is to deactivate the application and restart it when the conditions are right again; this is called the static policy. Some people are repelled by this idea but in an OSGi system this is quite natural, and well under the control of the operator. An update can cause a ripple effect of many bundles restarting but well designed systems should be robust enough to handle this. Hmm, let me rephrase, if your system can’t handle this model, it will fail in the end anyway.

In certain cases, it would do no harm to replace a service on the fly. If an application uses the Log Service, it is likely not interested which Log Service it uses. If this service is unregistered, it could easily be unbound and then bound again with a new instance. Most stateless services can be replaced this way. Binding and unbinding during the active life of the application is called the dynamic model.

Let us now take a look at the possible combinations and see what they mean, and where they are applicable.

  • static, 1..1 – The simplest case. The service must be present before the application is activated and the application is deactivated when the service disappears. The world looks perfectly simple and static for the application. The typical use case for this policy is the Http Service. Many applications should only be started when the Http Service is present and when it goes away the application has no purpose anymore. A once chosen Http Service can not just be replaced because the application must register the servlet with the Http Service; the Http Service is not stateless with respect to the application.

  • static, 0..1 – The common case when the service is not really required to operate the application and the programmer does not want to bother when the service shows up later. The application is therefore unconditionally activated, regardless if the service is present. However, if a service is bound to the application before activation, the application is deactivated when that specific service goes away, even if another becomes available. This cardinality could for example be used for the Configuration Admin service because it is likely to be there or not.

  • static, 1..n – This case is not very useful, tracking multiple services require a dynamic policy to make sense.

  • static, 0..n – See previous.

  • dynamic, 1..1 – The service must be present before the application is activated, however, the service can be unbound and bound again if a replacement can be found. This mode is useful when you have a stateless service but you require its presence to operate. For example, if the application needs to send messages then it can not function without a message service. However, it does not really care which message service is used.

  • dynamic, 0..1 – An optional service that can dynamically be replaced. The application is unconditionally activated and never deactivated on behalf of this service. This policy is typically used for a Log Service. Logging is nice, but not generating a fire alarm because you can not log it is bound to make some unhappy customers.

  • dynamic, 1..n – At least one service is needed, but many are handled. The application is only activated when at least one service is present and is deactivated when no more services can be found. This can be useful if you perform an expensive function that is only relevant when there is at least one consumer. For example, assume you run a GPS device in a phone so battery power is at premium. If there are no positioning listeners, it is no use to run the GPS chip thereby reducing power.

  • dynamic, 0..n – The typical tracking case. The application is unconditionally activated and never deactivated on behalf of this service. Services are bound and unbound as they come and go. For example, if you are tracking a set of Bluetooth devices in the service registry then this is your policy to use.

The Spring-OSGi spec can copy the concepts from the declarative services because they have exactly the same issues. The discussion in the mail group got confused because the interaction between the policies for proxying (static/dynamic) were confused with optionality.

Spring adds the concept of proxies. It is my advice that the rules are as close as possible to the declarative services model which means that in most cases the application must participate in the binding and unbinding. It is tempting to automatically replace the service reference in a proxy but most of the time services are stateful and then it is easy to confuse them. Associating proxying with dynamic does not make sense because normally the application must do some action during binding. Maybe Spring needs an extra policy to indicate that replacing the service in the proxy is allowed or not, for example statefull and stateless. Then again, I am starting to have serious doubts how useful proxy rebinds are in reality.

Anyway, this year will be an interesting year for declarative services. It is clear from this issue that declarative services did a good job in this area but the possibilities of Spring will likely encompass all the important features of declarative and then add more. I am not sure if it is a good idea to have declarative services compete with Spring. My personal preference is that the Spring-OSGi specification becomes an official OSGi R5 spec … Let me know if you have ideas how we should handle this.

Peter Kriens