Every once in a while, I review the SWT snippets. These are available both through the SWT project website and directly from the CVS server. Today, I updated the project from CVS and started looking through the additions. My usual process is to start with the highest numbered snippet and work my way backwards until I run into stuff I’ve already seen.

Today, I made it as far back as snippet 250. Snippets 250 and 251 show off the new DateTime widget (currently part of Eclipse 3.3M5eh). This widget can be used to create a calendar, date spinner, or time spinner. Snippet 251 is shown in action below.

I found what I believe is a bug (178372) in the SWT.DATE variant of the widget (I worked around the bug for the screenshot) and I have a problem with the calendar control (I’d like it to let me deselect the date; i.e. allow null to be a valid value so that I can unset dates) that’s really more of a feature request (178362).

The real problem that I have with this new widget has nothing to do with SWT. At least I’m pretty sure that it doesn’t. My problem is that this widget looks nothing like any other calendar widget on Windows XP.

I’ve read a lot of blogs and comments on blogs lately that suggest that other applications are “more native” looking than Eclipse. When I read this sort of charge, I’m left with what should be a pretty simple question: just what the heck does “looks native” even mean? The DateTime widget uses the native DATETIMEPICK control on the win32 platform (an emulated one is currently used on Motif, Photon, and Windows WPF), so this widget definitely “looks native”. However, a quick look at the “Date and Time” properties dialog on Windows XP reveals a completely different control with a completely different look. And in Microsoft Outlook, the drop-down calendar on the Tasks view is again different (I’m not posting screenshots because I don’t want Microsoft’s lawyers after me).

After poking around for while, I did find an instance of what at least looks like the native control on the Windows XP “Scheduled Task Wizard”. So, I guess that it’s probably fair to say that the “Scheduled Task Wizard” looks native. That’s one.

To be completely honest, I’m not sure that I even know what a “native” Windows XP application is supposed to look like. As I survey the various applications I have running, it occurs to me that they all have at least slightly different looks. Curiously, Microsoft itself seems to be the worst offender. The Microsoft applications that I run are the ones that look the least “native”. I managed to yank the menu bar off the top of the search window in Outlook and now it’s attached to the left side of the window (with text flowing bottom to top). I thought that menu bars are supposed to be fixed (I could be wrong, I’m not expert on Windows User Interface Guidelines). The new version of the Windows Media Player looks anything but native. The new version of their Office suite—which I’ve only seen from a distance—seems to throw any notion of a “native” look (and feel) out the window (failed pun attempt intended).

So what makes Eclipse look less than “native”? Is it those fancy curvy tabs? That’s a pretty easy thing to change if it really bothers you. Is it resizeable views and perspectives? How about we make Eclipse look like Outlook? There… isn’t that better?

I wonder if the real problem is that Eclipse just looks too native…

I have to admit that I’m darned excited (though apparently not quite excited enough to unleash a stronger expletive) about the Eclipse Java Persistence Platform (EclipseLink) project proposal. The name “EclipseLink” is an obvious play on the name “TopLink” which is the name of the Oracle Product from which the initial code contribution will come from.

Why am I so excited about this project? There’s a couple of good reasons. A big reason is that TopLink is a darned good product that’s used by a lot of companies (small-, medium-, and large-sized). It’s a product that has evolved in a very “Eclipsey” (Eclipsish?) way: customer feedback is continually driven back into the product; it’s a platform that helps developers do what they really need to do based on actual field experience.

Another reason why I’m excited about this project is that it pretty much brings me full circle. TopLink was the first commercial product that I ever worked on. Granted, I worked on TopLink Smalltalk way back in the early 90s, but the TopLink for Java developers did give me an opportunity to rant here and there about what I thought the Java version should do.

So… this project is exciting for me because it adds high-value, high-quality functionality to the Eclipse stack; it adds top-quality developers to the Eclipse committer list; and I get to hang out with a bunch of old friends.

Check out the project proposal here.

Tod Creasey is a long time (possibly the longest time) committer on the Eclipse platform UI project. In this podcast, Tod discusses the changes that are coming in the platform UI with Eclipse 3.3.

You can find the podcast here.

On a side note, the music that I include at the start and end of the podcasts is an original work by my best friend, Shawn Kilpatrick.

In this Eclipse Developer’s Journal article titled “Ship Happens! Insights From the Eclipse SWT Community”, Joe Winchester interviews SWT developers and a host of others—including SWT committers and contributors—about SWT. It’s an interesting read that gives insight into the design goals of SWT and addresses some persistent myths.

You can find the article here.

It’s about time that I continued that series I started discussing a rather old blog entry extolling the virtues of Eclipse 2.0 titled “Ten Reasons to Use Eclipse“. I’m up to reason #4 (If you need to catch up, see reasons 1, 2, 3).

#4 - Eclipse is Easy To Extend

The Plug-in Development Environment, or PDE, is an integrated development environment for developing extensions to the Eclipse platform. Using the PDE is like using Eclipse to develop Eclipse, allowing developers to leverage the excellent Java tools to extend the platform itself. I discuss the JDT (Java Development Tooling) more in detail below.

Thanks to the PDE, and to the attractiveness of the platform itself, Eclipse has succeeded in generating tremendous traction in the software development community. Plug-in projects aiming at integration more development tools into the Eclipse platform, are already too numerous to be listed here. Let’s only mention a few:

• the unescapable JUnit and Ant are part of the JDT (the base Java IDE).
• EASIE by Genuitec is a set of tools to manage J2EE application servers from within Eclipse. They’re precious for J2EE development.
• a ClearCase team provider to access ClearCase repositories.
• AntlrSupport integrates the Antlr parser into Eclipse.
• the AJDT provides support for Aspect Oriented Programming in Java.

The list is growing fast! As Eclipse gathers more market mass with plug-ins addressing all facets of the software lifecycle, it will soon become a de facto standard workbench for Java developers.

The list is indeed growing fast. It was growing fast in the 2.0 timeframe, and it continues to grow fast. In the past there were several portals leading to an indescribably huge number of plug-ins for Eclipse, both commercial and open source. Numerous portals still exist, but now there is an über portal in Eclipse Plug-in Central (EPIC).

As expected, the list above has changed a little… EASIE by Genuitec has morphed into MyEclipse which by all accounts is a wildly successful IDE based on Eclipse technology. I wonder too if the original author might have included the Eclipse Web Tools Platform on the list since he seems to have interest in J2EE. I guess that’s changed as well; J2EE has been replaced by Java EE. AntlrSupport seems to still be around, but I’m not sure what it’s current status is. There certainly is support for Antlr development in the form of Antlr Studio by Placid Systems.

The other plug-ins mentioned still exist and have been joined by thousands more.

To be honest, I’m not sure how easy Eclipse is to extend. In my experience, it’s relatively easy to extend Eclipse to do relatively simple things. My first few plug-ins were little utilities that did things like parse through XML files, and add methods to Java classes. This sort of thing is generally easy. I fairly quickly evolved to building fairly capable and feature-rich plug-ins. Doing harder things takes a little more effort as is naturally harder. However, given that the PDE makes it relatively easy to get started, it’s certainly the case that a step-by-step approach into the wilderness that is the Eclipse API is a good way to get your feet wet. But, like any other platform or framework, trying to jump into the middle is really hard.

So, I’m not sure that it’s easy. It’s not all that hard. However, it certainly seems worthwhile to extend Eclipse. And extending Eclipse is certainly easier than the alternatives.

Instantiations is making their software available free-of-charge for individuals developing open source software. According to the website, the software can only be used on open source projects; other uses would require that a license be purchased.

Instantiations has been part of the Eclipse eco-system from the beginning. I’ve been using their WindowBuilder product for years. In fact—if I remember correctly—I’ve been using it since 1991 (maybe 1992). Of course that was the first Smalltalk version (they still do Smalltalk). It’s a great product built on an incredible legacy and a wonderful showcase of the power of the Eclipse platform. They have several other products, including RCP Developer, GWT Designer, CodePro, EclipsePro and WindowTester.

Go here if you need to use any of these great products on your open source project.

Naci Dai, Web Tools PMC Member, J2EE Standard Tools (JST) Project lead, and SOA Tools Platform (STP) committer, spoke with me about the Web Tools project at EclipseCon 2007.

Among other topics, Naci—pictured right with Oisin Hurley of the STP project at EclipseCon 2007—discusses the current state of adoption of web tools, development goals for the 2.0 Europa release, and the EclipseIST conference that he’s putting together in Istanbul in April.

You can listen to the podcast here. Add this podcast (and other Eclipse podcasts) to your favourite reader here (or here for iTunes).

I spoke recently with Alex Blewitt, OSGi Expert and Editor-in-Chief of EclipseZone, and Neil Barlett, OSGi Enthusiast, about OSGi. Neil has recently authored a series of articles on the topic. In this podcast, Neil and Alex talk about how OSGi is being used today and its potential for the future.

The podcast is here.

We recorded the talk in the registration area at EclipseCon. For effect, I thought it’d be cool to have a background hum of the excitement that pervades EclipseCon. I’m not sure the effect is quite what I was hoping for (especially since—other than the guy with the carpet cleaner—there really wasn’t a lot of buzz in the registration area).

Note that I’ve changed Neil’s title above per his request that he be labeled an OSGi “enthusiast” rather than “expert”. I question his choice given the quality of the articles he’s produced and his obvious depth of knowledge in the subject, but I respect his position.

On a related note, I’ve been working on the podcast feed. It now loads into iTunes, but for some reason it doesn’t show the recent podcasts. I’m still trying to sort out why it’s not working quite right. If anybody has experience with Atom feeds and iTunes, please feel free to impose your skills on me….

I’m up to #3 in the “Ten Reasons to Use Eclipse” posted in the Eclipse 2.0 timeframe by Renaud Waldura. If you’re curious, part one is here, and part two is here.

#3 - Eclipse Has Momentum

Although Eclipse is an open-source platform, it isn’t the only one. A most notable competitor is Sun’s own NetBeans, commercially distributed under the name Forté. NetBeans has lots of features, is as free as Eclipse, and surely must be taken into account in a discussion of Java IDEs.

But while Sun’s efforts in opening NetBeans to other software providers have been met with mixed results, the Eclipse Consortium seems to have succeeded in creating a group of very strong backers actively supporting and extending the Eclipse Platform. First among them is of course IBM, who funds the Eclipse Project, and is building its commercial J2EE IDE product, WebSphere Studio, on top of Eclipse. But many others can already be mentioned: Rational Software is using Eclipse for its Rational XDE modeling tool; Fujitsu for its application server product line; and JProbe, Sitraka, and others are porting their respective products to Eclipse.

I’m going to start with a minor correction. It is true that most of the committers on the Eclipse Project are IBM employees, but it’s not technically correct to say that IBM funds the project itself. To avoid potential for confusion, it’s important to note that the “Eclipse Project” is just one of the projects under the Eclipse umbrella (it’s the one that produces the Eclipse SDK). It’s also worth pointing out that there are non-IBM employees who commit and contribute to the Eclipse Project.

Some things have changed: I’m quite sure (though not 100%) that Sun no longer uses the name Forté. This posting is interesting in that it confirms the NetBeans folks assertion that they’ve been building a platform all along. I’m not qualified to comment on success of NetBeans as a platform; I’m sure that folks are using it, but I really don’t hear too many success stories (consider this statement FUD-lite).

I do hear frequently about new success stories with Eclipse as a platform. We’ve documented some Eclipse Rich Client Platform (RCP) applications (both commerical and open source) here. There are numerous other products based on Eclipse, including the recent release of JBuilder 2007 and Apache Tomcat (which includes the Eclipse JDT). Eclipse Plug-in Central lists numerous other products that are built-on, based-on, or complement Eclipse technologies including a whole page of IDEs.

Does Eclipse still have momentum? EclipseCon attendance is up (I’ll let somebody else report the final numbers), it seems that new projects are proposed weekly, big companies are joining the foundation, newsgroups are active, and we keep having to increase our bandwidth. It sure feels like Eclipse has momentum.