There are a lot of questions, but don’t worry, you can skip those that don’t make sense to you and you can stop whenever you want. However, I encourage you to fill the whole survey so as to get a feeling of what the developers community think of widely available languages.

Some interesting results are shown in this comment to the post in LtU, here are some others I have shown when I finished the survey (do the survey first if you are planning to do so, before continuing reading to avoid biasing the results):

VB stands high in “The thought that I may still be using this language in twenty years time fills me with dread” question and also in “I am reluctant to admit to knowing this language”. Surprisingly, php is second on this last category, being a so common language for web applications. Delphi ranks third.

Java ranks high on “There are many good tools for this language” and “Third-party libraries are readily available, well-documented, and of high quality”.

In question “This language is likely to be around for a very long time”, the four high ranked are C, C++, assembler, and Java.

This week, I’ve read that the Eclipse Foundation and Google has set up Eclipse Labs, a forge for Eclipse related projects backed by Google Code. Thanks to every one involved in making this possible.

Personally, I think this will spread the use of Eclipse and it can be a central repository where sharing new ideas. Even more interesting: it seems that there will be an Eclipse Labs API, that, among other things, will make possible for the Eclipse Marketplace to discover projects hosted at Eclipse Labs.

Having a look at it, one thing I miss is some listing of projects in the front page. You have to click on one of the tags or type your search. For instance, I’d like to have a listing of projects ordered by activity.

Eclipse Labs doesn’t impose IP restrictions, and you are given a set of licenses from which you have to choose one for your projects (these are EPL compatible licenses, remeber that GPL is discouraged). Projects hosted at Eclipse Labs are not official Eclipse projects, so “org.eclipse” namespaces should be avoided. One thing to take into account: you have to choose between hosting your code in svn or Mercurial… so some of us who had started to think of git as the svn replacement will have to re-evaluate our choice.

By the way, Wascana is already hosted at Eclipse Labs, so what are you waiting for? As Doug Schaefer says, I’m impatient to see what people do with Eclipse.

I’m from academia and, contrary to what Andrew Eisenber pointed out in a comment to Chris post, I think some projects can only be done (and make sense) within the academic world. I cannot imagine myself being paid to build EclipseGavab, and EclipseGavab itself makes sense only in the classroom. However, this is just my opinion, and it may be caused by a lack of research outside universities in my country.

Anyway, as Stephan Herrmann stated in a previous post, there are usually two kind of research projects. He talks about individual projects and fund projects. The first ones are usually ideas that are explored and sometimes abandoned as soon as a paper is written or a PhD is finished. These ideas may be more challenging. The second ones are usually projects that involve a group of researchers. Fund projects usually achieve their objectives, getting something “polishing” implemented.

I think both kind of projects may have interest for the Eclipse community. Individual projects may be more challenging, and require bigger efforts to be consumable by interested Eclipse projects, whereas fund projects may produce code that is directly usable by others.

Sadly, from the different posts I’ve read about this topic I can make a conclusion: there is little support from Eclipse ecosystem to academic projects apart from GSoC. Let me sketch here some toughts about making Eclipse more “academic-friendly”.

• I think that having some infrastructure for Eclipse-related projects could help. I agree with Marcel, on that some tools are developed just to show an idea, a paper is written and the tool is forgotten. I have some of these in old workspaces (including my PhD). But, could these tools have been interesting to others? Could they have created a community? If Eclipse-related research projects had a forge (my apologize to webmasters) with an agile way of adding new projects (no IP logs, fast approval, …), these projects could have a chance to foster community. This even would be a first step towards becoming an official Eclipse project. The Eclipse marketplace is a first step towards this, but a real forge is needed IMHO. In fact, I think I heard something about this last year, but I’m not sure.
• Having a repository (it may be just a wiki page for now) listing Eclipse-related academic works may also help. People may find interesting research projects/papers, and this may help others to get involved. Some ideas may even be of interest for some Eclipse projects. Keeping an eye on academic works could give an idea on how is Eclipse being used.

Last, but not least, these ideas may be applicable to non-research projects also. What do you think?

I was the responsible to detail the contents of the subject, which consisted on teaching three different modules of one paradigm each: functional programming, concurrent programming, and (well it is not really a paradigm, or is it?) dynamic languages. I had to decide which language to use within each paradigm. Obviously, given that we had one semester, I didn’t like to spend one session per module to teach a new, different, development environment. So, as an Eclipse user, I started thinking of using Eclipse as the single IDE for all languages.

This could have been all, if I had some degrees of freedom. I chose Haskell for functional programming, Ruby for dynamic languages and Java for concurrent programming. All of them have great support within Eclipse. Haskell is supported by means of the EclipseFP project, Java has the great JDT community, and there was also the great job from the DLTK project team.

However, I had very few sessions to teach each one, and my students didn’t knew Java (neither Haskell nor Ruby). Ok, let’s see if we can use at least some language that is closer to what they are used to so that they don’t need to learn all three languages. And I came over PascalFC. PascalFC is a Pascal-like language focused on learning concurrent programming. And my students already knew Pascal. Nice.

Wait. There’s no PascalFC support in the Eclipse ecosystem. Oh… Even worst, some colleagues were supposed to teach a similar subject on other degrees. But their subjects were focused on object orientation with Pascal (due to same constraints) rather than dynamic languages. And they were asking also for a Pascal plug-in.

When looking for a Pascal Eclipse plug-in I came over two options: Pasclipse and EclipseColorer. The Pasclipse project seemed to be abandoned (and looking back to its page, it is). And the EclipseColorer was just a colorer. And obviously, there weren’t any PascalFC Eclipse tools.

It seemed that it was a lost battle… But we wanted to have a single IDE for all these languages, and that IDE had to be Eclipse. So we take on building these two development tools. Me and another colleage worked hard during part of the summer and the first semester of 2006 in order to release a first version of two tools for Eclipse: Pascaline and PascalFC Development Tools.

When we finished we were half way where we want to be. Next step was to release an Eclipse distribution that contained these two tools, among the others (JDT, DLTK Ruby, and EclipseFP). We called it EclipseGavab 0.5 (starting numbers for versions are a funny thing to play with). We were finished by February 2007. We did it. We even included in the Windows version all the tools (compilers, JRE, interpreters) needed by the different plug-ins.

Nowadays, EclipseGavab is in its third version (EclipseGavab 2.0) which is based on Eclipse Ganymede (3.4), and now contains more tools, like CDT (which is used in a subject about C/C++ programming), and Subversive, Mylyn and ECF. These three last ones were included to enable collaborative development. I’m looking for some screenshots of the older versions, but I can’t find them, I have to look in the backups. So here you have some of the latest version:

The splash

The Pascaline tool supports Pascal development with the aid of the CDT language extensions, which is very helpfull for instance to provide debug capabilities.

There is also help available.

The PascalFC Development Tool links the deadlocks shown in the console to take you to your “dead” code.

Shared editing may allow students to share ideas and resolve problems.

We now also detect installation changes and re-configure your workspace configuration so that compilers are found by tools.

Micael Gallego and Francisco Gortázar are the two people behind this blog. We’ve been working for the last five years with Java, Eclipse and whatever technology that’s worth the effort, sometimes side by side, others not, but always having interesting talks about software, IDEs and development. So, why not share all the leasons learnt, and all the software developed, under a common umbrella? This is sidelab.

When it comes to Eclipse-related interests some things come to my mind. Let’s remember some of them just to give an overview of who we are…

I remember the LDT days, when IDEs were going multi-language, and people was trying to factor out what all these language had in common in a single framework. We spent hours reading the news from the LDT project. Grammarware has always been one of our interests.

It was also nice when the people at CDT refactored the project allowing others to benefit from their framework to implement their tools. This is how Pascaline could be developed.

In the early days of Eclipse, I recall developing a design pattern detector implemented on top of ASTView. Recently, I’ve tried ASTView just to make sure it is still there.

More recently, we’ve been giving a seminar on Mylyn, and another one on JNA/JNI in Eclipse. And some students are developing plug-ins such as veex, gcj4e, and jmh4e.

And last, but not least, Eclipse allows you to do things like EclipseGavab, the Eclipse distribution we’ve been using to teach programming: JDT, CDT, EclipseFP, Subversive, ECF, Pascaline, PFCDT, lots of plug-ins doing their best to ease students learn several paradigms, languages and tools. Eclipse has been used for almost 4 years as the unified IDE in Computer Science…

Great job by great people, let’s see if we can put our two cents in for the community.