A brief post where I will simply encourage you to visit Ken’s blog and read about this new Samsung cell phone that has some of our software in it.  It seems Engadget covered this same phone just last week, but they didn’t pick up on the fact that IBM’s J9 is inside as Linux for Devices managed to do.

The teams are composed of 3 students focused on technical work, and 1 MBA who handles the business side of things.  Their goal is to take a concept and create a compelling business case for further investment (by IBM).  Think of this as a Start-Up on steroids – they’ve got less than 4 months to do it all.

This week all of the extreme blue students are gathered in Armonk, N.Y. where they will deliver their finely tuned pitches.  The audience will be composed of their fellow students, IBM mentors and most importantly IBM Executives.  Sam Palmisano is usually able to find time to take a brief tour of the booths that the students have setup, with a special stop at one of them for a deeper dive.

The Ottawa J9 team hosted an extreme blue project this year.  We used this to help take one of our prototype concepts further towards a complete business case.  I have to confess this is a bit of a teaser post – since I can’t disclose very much about the project.  You’ll have to speculate wildly based on the following image:

You may want to check out the video clip from the local news station that talks about the IBM Ottawa Extreme Blue expo.

Our team has a long history of hiring our COOP students, its a great deal both ways.  The students know what they are walking into, and we’re got a better idea of what they are capable of.  Even I was a COOP student for the VM team way back when.  Each of the names involved with this project are top notch students -whoever manages to hire them will be lucky (and yes, we’ll be trying to).

Good luck in N.Y. this week guys!

Here is the summary:

The Java™ Native Interface (JNI) is a standard Java API that enables Java code to integrate with code written in other programming languages. JNI can be a key element in your toolkit if you want to leverage existing code assets — for example, in a service-oriented architecture (SOA) or a cloud-based system. But when used without due care, JNI can quickly lead to poorly performing and unstable applications. This article identifies the top 10 JNI programming pitfalls, provides best practices for avoiding them, and introduces the tools available for implementing these practices.

Many kudos to Michael Dawson for his wordsmithing and persistence to make this article happen, he really deserves the lion’s share of the credit.

Ron calls out his favorite pitfall as using the wrong JNIEnv.  I’ll pick the one that makes me laugh because it is sad but true:  Choosing the wrong boundary between native and Java code – believe it or not there was a real world example that motivated the sample code.  The customer had decided to go for a “pure java” solution for their embedded application, so the minimal amount of native code was written – two functions that could read or write individual bits (not bytes).  While the example is outlandish, many times the integration of legacy code via JNI is done without thinking of the cost of the boundary.

Let’s discuss the picking the right side of the JNI boundary for a minute.  The NIO implementation is a good example of trying to fix file performance by keeping the data on the “right” side of the JNI boundary.  While NIO provides other functions allowing for tighter OS integration, one key performance win is due to the file data is kept in native buffers.   If you’re trying to get the data into java objects to be manipulated, you are likely defeating the performance win of NIO by forcing all that file data across the JNI boundary.  Where NIO will excel is needing to pull a bunch of data from a file and put it in another file – if you don’t need to hoist that data up into the java object space, you’ll get close to pure native performance.

Please check out the article, or at least bookmark it and send it to people who write JNI code.   Maybe we’ll see fewer false JVM bugs that turn out to be JNI related problems.

I won’t be attending this year, but my friend and colleague Trent Gray-Donald will be there presenting a session titled “Under the Hood: Inside a High-Performance JVM™ Machine” on Friday June 5th, 10:50am – 11:50am.  While you won’t find him on the Rock Star Wall of Fame, he was did earn that title back in 2007 – his sessions are usually well attended so if you want to see him in action go early for a good seat.  This year his presentation discloses more details about the IBM JVM technology than have been previously made available, expect to see assembly code snippets on screen during the presentation along with lots of technical detail.  His presentation should be a reasonably good primer in what it takes to be a “state of the art” JVM in today’s competitive market – and it will be presented by someone who has hands on experience creating the technology.

If you can’t make it in person, JavaOne has a pretty good track record of making past sessions available online. Of course you can also follow along on Twitter, and there appear to be two hashtags (#JavaOne and #JavaOneConf) being used to help track tweets.

As for what to expect from JavaOne this year, I’ll shy away from trying to make any firm predictions – but it has been announced there will be more details on the Java App Store.  The Oracle/Sun deal will certainly be one of the topics that can’t be avoided.  Java7 should get some sort of announcement along with some benchmarking fun I’m sure.  One camp to keep an eye on is the Microsoft keynote – the 1st time they’ve done one for JavaOne.  Last, but not least – IBM is a general session cosponsor thus Craig Hayman will be giving a talk as well (Thursday, June 4th, 5:30 – 6:15 p.m), one I’d recommend people attend if they can.

A quick NSLookup query turned up the problem – it seems the DNS server at work had been reset and the DHCP lease for my workstation had not expired yet.  This means that the DNS server had no idea what IP my machine is using, but my workstation is under the assumption it has the rights to that IP address still.

Jim happened to be one of the few online, so I chatted with him a bit to see if he had any ideas how to resolve the problem.  He mentioned that this had happened to him a couple of times and his solution ended up being to drive back into work, I figured that would be my backup plan.

Jim then suggested using arp on a machine I had recently ssh‘d into from my workstation.  While I hadn’t actually used ssh to connect to any of our lab machines, it  got me thinking about our CVS server – as I was running Eclipse on my workstation and did have in theory an active connection.  From my laptop I was able to ssh into the CVS server, then using netstat I got a full list of the active connections.  Looking through the list turned up a few which were numeric (clearly where DNS had failed to provide a reverse name mapping).  From there it was a simple matter to remote desktop to the “right” numeric address to reach my workstation – I got lucky on my 1st try.

Unfortunately I didn’t attend that Eclipse Summit Europe, so I didn’t hear what Dave had to say.  Talking to some folk that I do know who went – their impression was “Vintage Dave“.  I have made reference to this in the past – but I did work for Object Technology International and my university education was heavily influenced by Dave Thomas – myself and my classmates were some of the first to work our way through learning datastructures etc., in Smalltalk.  So a lot of my opinions of what Dave says are going to be coloured by my past experience (and his influence on the start of my career).

Almost any talk given by Dave will get people to sit back and think – he’s also not afraid to stand in front of an audience full of Java programmers and tell them that they’ve picked the wrong language.  In my experience he always has some good points and is worth listening to, but to interpret his talks as a set of black and white statements (or out of context) is dangerous.  There were two other bloggers that gave some in person accounts of his talk – I’d encourage you to read those.  Another reasonable reference is to read about Dave’s thoughts on JavaScript – and if you go back to what got Java started (the browser plug-in) and where Java is today, it is clear that the language didn’t fulfill the original promise.

There are others who are predicting that embedded java is dead, even James Gosling has suggested that J2ME is done.  However, if we take a look at embedded Java today – you’ll see that the Blu-ray format is a new niche for this technology.  As well, even low cost phones such as the Nokia 2760 support the J2ME profile.  This means that as a platform, there are actually billions of devices out there – how is this not success?

I think it depends on what the original goals were, and the future of the platform.  Among the various J2ME enabled devices, there are differences.  This isn’t really a lot better in the browser / javascript space, but Web2.0 holds the spotlight at the moment.  Probably the biggest threat to the J2ME space is the new generation of smart phones:  iPhone, G1 and Palm Pre – none of these have embraced Java as the path forward.

Looking at embedded hardware, we’re starting to see more and more capable hardware in the palm of your hand.  For example – iPhone is a 400MHz ARM cpu compare this to a hot desktop circa 1998 (PentiumII 400MHz), the very definition of what embedded means is changing.  In the near term J2ME will continue to be an expected feature in cell phones, so there is still a market which will continue for several (5?) years.  As the smart phone market sorts itself out we’ll see which new (or old) technology rises to the top.  It is probably safe to say that J2ME will have an impact on the outcome,  but to bet on it as the winner is probably long odds.

Disclaimer: The postings on this site are my own and don’t necessarily represent IBM’s positions, strategies or opinions.