Chris Aniszczyk's (zx) diatribe

I attended and had the honor to moderate Eclipse Platform Day at JAX 2010. It was another nice day in Mainz, especially along the jogging trail…

Boris Bokowski kicked off the day by talking about adopting e4. Boris spoke to a packed room! There was a lot of interest in understanding how Eclipse is evolving and what it means to reach 4.0 and how it may affect downstream folks. There’s still some confusion, but people are starting to understand why e4 was necessary and having concrete examples definitely helps.

Later in the day, Steffen Pingel discussed the new p2 discovery user interface that is shipping with the Eclipse Helios release. If you’re looking for a simplified and customizable user interface for your Eclipse-based application, I highly recommend looking at the discovery code.

Christian Campo spoke about the SWT Qt port he’s been working on for awhile. He managed to bring up a version of the Eclipse SDK that ran on top of Qt… which opened the door to easy theming and some other features that Qt offers out of the box.

In the end, I hope people enjoyed the day and learned something new.

Today I had the opportunity to attend part of Eclipse Modeling Day organized by Eike Stepper and Sven Efftinge at JAX 2010. On top of that, it was a nice day in Mainz!

Ed Merks started off the day introducing the audience to the Eclipse Modeling Framework (EMF) which is the foundation of all things modeling at Eclipse…

After Ed’s talk, I heard a modeling case study by Stéphane Barbey from Paranor AG. They managed to create a sophisticated workbench from banking software using a slew of Eclipse modeling technologies. Personally, I think the modeling community at Eclipse needs to start formalizing case studies like this as it really shows what people do with the technology in the real world. Maybe the Eclipse Foundation can help create some case studies similar to what we did with RCP for the Eclipse Modeling community?

I also attended Tom Schindl’s presentation about EMF Databinding…

I always forget about the great example that Tom put together that demonstrates the power of EMF, CDO and Databinding. After Tom’s presentation I succumbed to jetlag and went for a nap, but I’m sure Ed Merks will have a nice write up of the event. I saw him taking pictures throughout the day. On a side note, the Modeling Day was so well attended at the conference that we had to move rooms. For some reason I don’t completely understand yet, Europeans love modeling technologies. In the end, it’s exciting to see the Eclipse modeling community growing each year.

So I’m setting up a fresh Eclipse install for 3.6 M7 test pass week this morning and found that I was unable to install anything via p2. Any p2-related action that I performed ended up with a strange “Profile id _SELF_ is not registered” error in the log followed by many profanities. So the next obvious step was to debug the problem and about half way through debugging… I realized I hit a similar problem before on Mac and it was related to a bug in The Unarchiver application. I was right… again… for some reason, The Unarchiver will append “0000664” to p2 profile data while decompressing.

This is obviously a problem and will cause failures. The workaround is to strip the “0000664” suffix from the p2 profile data or simply use another unzipping tool. So for any folks who use Eclipse and hit this on Mac, I hope you find this post and save some pain.

The Eclipse Planning Council is holding a run off vote to name the Eclipse Helios +1 release. The last vote held helped narrow things down between two names…

Isaac

Indigo

Rock the vote, we’re almost there.