| Re: [eclipse-incubator-e4-dev] E4: shouldn't we define the goals first? |
Let's be clear about the terminology first:
"Eclipse 4.0" simply means "the first release of 'Eclipse' in the 4.x stream". The problem is that it's *singularly* unclear what the word 'Eclipse' denotes, in this case. Let me show you what I mean...
Given the team that you and I are both on, and your other comments, I might assume that you really meant the "Eclipse SDK". If that's true, then the answer is simple: "Eclipse SDK 4.0" is simply the same as it has always been: "the SDK that you would want, in order to be able to build itself". (Isn't recusion always described as "simple". <g>)
However, it's important to understand that the things that happen at Eclipse.org are *much* bigger than that. It's definitely not just about IDEs any more. There are runtimes, widget toolkits, RCP and SOA frameworks, you name it. Even in the IDE space, the Eclipse SDK, is only a small piece of the many, many IDE-ish things that are built with Eclipse. (Some of those things are captured by other IDEs like Netbeans or IDEA, but many are not.) The thing is though, work on all those things proceeds *outside* the scope of the _Eclipse_Project_ (i.e. the team you and I are on, not Eclipse.org in general), and as such are not things that we will implement.
So how did we end up with initial set of directions that are being discussed? We started with the germ of an idea that one way to look at what the Eclipse Project produces is that we make the "plumbing" that most other things at Eclipse.org are made from. [That position was easier to defend before the Equinox team moved to the RT project. ;-) ] Essentially, we build the "platform". So, the obvious starting point for the work is to identify the things that the platform needs to do to be a better basis for the other things that are done at Eclipse.org. Several of us spent a significant amount of time during the 3.4 release cycle trying to understand what that list might look like by doing two things:
1) looking inward at our current platform and based on bugs, mailing lists and newsgroup postings, discussions with others at Eclipse.org, etc., trying to see where the pain points were (like the huge learning curve for new developers, and a Resource layer that was really only useful for Java plug-in developers)
2) looking outward at the larger software development community to see what would be expected from a modern platform. That depends, of course, on how loosely you define a platform, but there were some obvious trends:
- Applications are *expected* to have a polished, stylish look and feel now -- we had better be able to build applications that look like that on our platform
- The web is *very* important. Even for applications that are not moving to "pure" web UIs, it is expected that some of the application's functionality could be provided via the web -- we had better be able to do that too.
Unfortunately, it also became very clear to us is that a) we simply didn't have the scope to understand whether we had captured all of the platform requirements, and b) even for the set we did know about, it was obvious that we would need more resources than we could provide ourselves to implement them. That, together with some "gentle chastisement" (tm) from the community about how closed we had been in the past, made it very clear that the only way this effort would succeed would be to open up the development so that others at Eclipse.org, who had *specific* needs for improvements in the plumbing _for_purposes_appropriate_to_their_work_, could participate and drive the development in ways that were useful to them. In the world we live in, that's as close as you can get to the "purpose" you were asking about in your note.
So, for all things "4.0", I think first about a new platform that will be "the plumbing for the next generation of things done at Eclipse.org". *That* is what we have taken to calling "e4", and that's so far been the focus of most of the discussion in this list. We also build an "Eclipse SDK" that is the thing that we need to build that platform, and that is what we would call "Eclipse SDK 4.0".
In reality though, there is a much bigger/harder question which may be the crux of what you are really trying to figure out (although you may not have realized it): Who is responsible at Eclipse.org for the "big picture"? The team that is ensuring that the Eclipse SDK + WTP + TPTP (+ all the other pieces that are required) really is a credible competitor for NetBeans or IDEA or whatever else is a threat to some parts (or all) of our community?
That's a good question.
McQ.
Oleg Besedin---05/02/2008 03:10:42 PM---In all the discussions that went around E4, I haven't seen a "definition" of what Eclipse 4.0 is go
From: | Oleg Besedin/Ottawa/IBM@IBMCA |
To: | eclipse-incubator-e4-dev@xxxxxxxxxxx |
Date: | 05/02/08 03:10 PM |
Subject: | [eclipse-incubator-e4-dev] E4: shouldn't we define the goals first? |
