Sources available!

The Hibachi sources are now available on the Eclipse server. There have already been a few folks who have downloaded them and built Hibachi. Getting the sources through the parallel IP process was not a piece of cake. We had to replace the parser generator, and leave out the pretty printer. Even though the authors of the pretty printer gave their written consent for us to use it, because it had been released elsewhere under GPL (which is not Eclipse compliant), we couldn't use it. The IP process also brought up such issues as comments that suggested use of a possibly non-compliant library (not the actual use of the library, mind you).

I suppose it's a good thing that the legal team at Eclipse is so thorough. In the end it will help insure widespread acceptance, but in the short run, it adds a non-trivial delay to the moment when a piece of code becomes available for use.

On our way

Hibachi is officially an Eclipse Open Source Project. Since it was approved by the Eclipse Management Organization, we've gotten our newsgroup, mailing list and website set up. We've also gone over the code and adapted it so that it will pass Eclipse IP inspection, including swapping our parser generator for another one.

Since June, I've been trying to build a diverse community (vendors, open source projects, universities, third parties, individuals), and while it took a bit longer than anticipated, I think the effort will be worth it in the long run.

I've presented Hibachi at a couple of conferences (Ada Europe, Ada UK), and this week I'm giving a Hibachi workshop at SIGAda with a couple of the committers helping out.

I also attended the Eclipse Summit in Ludwigsberg, Germany. It was smaller in size than EclipseCon, which definitely has it's advantages in terms of interactions and access.

So now we're just waiting for the Eclipse legal department to give us the ok to put the sources on our server during the parallel IP process. Then we're on our way.

Next phase

I think the initial interest is known now. We've got participation from Ada vendors, the open source Ada community and Academic institutions. I will be presenting Hibachi at Ada Europe next week, and SigAda has agreed to a Hibachi workshop at the annual meeting in November in DC this year.

I've had some contacts and interest from industrial partners, but so far more as users than contributors.

Now we need to define each contributor's role and methods of working. Then verify/revise the project plan based on our roles and have the project review.

Growing interest

I've been encouraged by the response from the Ada vendors to the project proposal. The major active Ada vendors are on board and the combined level of commitment leads me to believe that we will have enough community involvement to create a viable project. It is interesting to see how the principles of cooperation and competition play out within the Eclipse open development environment.

I think the Ada vendors have all come to the same conclusion, that there's more to be gained by pooling resources and cooperating, then by each one going their own way. What I've been seeing is that our plugin (AonixADT) is more and more being compared with CDT and JDT, rather than Ada IDEs from other vendors. And with 10+ committers on a project like CDT, and even more on a project like JDT, it's pretty obvious that any single Ada company's technology, no matter how good it currently is, will fall further and further behind customer's expectations.

Going Live

The Hibachi (ADT) proposal is now online and open for comments at http://www.eclipse.org/proposals/adt/ and the newsgroup news://news.eclipse.org/eclipse.tools.adt .

People will have surely noticed that the project is known by two names, ADT and Hibachi. ADT is the acronym for the official project name (Ada Development Tools). Hibachi is a codename. I noticed that Eclipse allows projects to be known by codenames (like Mylar). I thought that a codename would be a good way to pay tribute to Jean Ichbiah, the chief designer of Ada who had passed away just the month before. Also, although we (Aonix) tried to always refer to our plugin as AonixADT, to avoid confusion with an eventual Eclipse ADT project that we had envisioned, there was some namespace creep which I didn't think would be helpful when it came time to build a community. So while the project will be known as ADT, which is unavoidable for a sibling of CDT and JDT, I will try to use Hibachi whenever possible in memory of Jean Ichbiah and to avoid confusion with the Aonix product.

As far as the name Hibachi, until I hit on it, the leading candidate was Beaujolais, after the Beaujolais Effect. Others suggested "Green", Jedi, Kermit, and various more obscure references that have meaning for the old-timers here at Aonix, the company that Jean started to turn the Ada language he designed into a technology and products.

I only came up with one complete anagram for Ichbiah- Hibachi.