3 - Tim Wagner of the WTP Project

Riyad Kalla: All right, welcome to episode three of ten in the Eclipse Zone Callisto podcast series. I'm you're host Riyad Kalla. In this series we are interviewing key members of the Eclipse projects that will be included in the Callisto release. Today we'll be interviewing Tim Wagner who is a part of the WTP project. Thanks for joining us today, Tim. Why don't start by giving us a description of what WTP is about and what role it plays in Callisto?

Tim Wagner: Thanks Riyad, great to be here. So, WTP stands for the Web Tools Platform and that really kind of suggests the scope of our project. Our goal is to create a platform of APIs and also the exemplary tools that go with them that support the development of web applications and also development against standard middleware frameworks. So, J2EE being the sort of the most obvious one of those.

Riyad Kalla: Right.

Tim Wagner: The Callisto release contains version 1.5 of WTP and we're really excited to be apart of the Callisto release for a couple of reasons. First I think Callisto's fantastic for community as a whole. By bringing so many projects together in one place at one time, it's really going to simplify the mechanics of getting people access to those projects.

Riyad Kalla: Right.

Tim Wagner: I also think that it helps people discover them, and that's fantastic for WTP because I think it's really going to help get our project in front of a much wider audience. The other thing that the 1.5 release represents is a real step forward in WTP in terms of the stability both for our tools and the platforms. So we think that users will appreciate improvements in the tools and adopting companies, companies like BEA (my company), IBM, Genuatech, and others are really going to benefit from the improvements in the stability of the platform.

Riyad Kalla: Right.

Tim Wagner: We also like to think that with Callisto, WTP has now kind of reached this kind of important conceptual milestone in that it's starting to become part of what we think of as the standard Eclipse package. So that since a lot of people need web and application tooling, we think this is going to help WTP become part of that fundamental download that every developer's experience includes.

Riyad Kalla: Well, that's right. When the first drop, when the 0.7 drop came, I remember it was already right off the bat one of the most active newsgroups people were downloading. At the time, there's was still some growing pains to be worked out and things like that, but that demand has absolutely been there right from the get go.

Tim Wagner: Absolutely. And it definitely has been a growth process for us. Getting that initial release out the door was an important milestone, as I'm sure you're aware, in any project's lifecycle to start to get software out into people's hand. And one of the reasons we gave it that pre-1.0 numbering was precisely because we wanted people to understand that it was definitely a work in progress and we weren't ready in that point in time to start declaring our APIs yet.

Riyad Kalla: Right. Walk me through this. When the first release of WTP came out it seemed like this enormous drop of code. Conceptually it showed a lot of functionality but usability wasn't quite there and there was a lot work to be done on it. How did the WTP project get started?

Tim Wagner: Sure. What really started this project rolling was a great contribution from IBM of a bunch of their code and also from a company called Iteration, a company lead by Naji Die, who's still actively involved in the WTP project today. Object Web, which is a European consortium, was also heavily involved especially in the beginning in a lot of the planning and formulative stages of WTP. So what we've been engaged in is a process of trying to absorb that code, rewrite it to make it sort of work in the Eclipse platform sense. And you still see some of those important pieces going on today. For example, the project navigator work that's gone into the platform is largely a work of some of the prototypes that occurred in earlier releases of WTP and now has become part of that fundamental platform architecture.

Riyad Kalla: Right, right.

Tim Wagner: There's definitely a thread there of migrating code to the right places, continuing to refine the pieces we inherited from the beginning to make them work as best as possible in the Eclipse environment. And at the same time, we've started grow in new areas that are completely unique to Eclipse and WTP within Eclipse. For example, some of the work we've done on project facets, which not everyone may be familiar with, but it's a technology for doing project specific functionality that is unique to a particular deployment environment. Things like that that have no relationship, or didn't come with the original code drop that evolved and developed subsequently. So, it's really a combination of those two evolutionary threads.

Riyad Kalla: Okay, so IBM contributed the original drop. How did BEA and how did you'd get involved in that process?

Tim Wagner: So my involvement really started in -- around February of 2005, which is when BEA joined the Eclipse community as a strategic developer. So at that time Kristov Nie of ObjectWeb and Bjorn Freemen Benson, who's now with the Eclipse Foundation, had been jointly managing the WTP project through those earlier conception stages. And I had a few different goals. I mean obviously BEA has a real interest in WTP. It's something we use and adopted to build our commercial products. My goals when I joined the project as lead were to start shipping software, obviously, you see that with the.7 release, eventually to align us with the project's release schedule; so what has now become Callisto, and to advance our API plans. So you can see with every release have now been declaring more APIs and that's sort of a gradual maturation process as different components reach that stage of their lifecycle. I've been very pleased with how the project's been moving. It's also been a real fun time for me. It's been great to learn all about Eclipse and certainly to become of that WTP community, which continues to grow.

Riyad Kalla: So everyone knows WTP is a sizeable project. And there is a lot there. There's the editors, there's the project support, and all that. How are the decisions made to create the three top-level project you have now under WTP, which is the WST, JST and JSF projects? How did you guys avoid, say, going down the Tapestry road, the PHP road which is now another top level project. And then we have Struts, Hibernate, Ibetas Spring, and so on. I mean, it seems like the list of top level projects you could create under web development is endless. So how do you guys stay focused?

Tim Wagner: What do you do about the framework de jour problem there, huh?

Riyad Kalla: Absolutely, yeah.

Tim Wagner: So, let me sort of start answering that by taking you through our project structure and then I'll kind of circle around to how we handle the scope issues a little more.

Riyad Kalla: Great. Okay.

Tim Wagner: So, first of all, as a top-level project WTP really have these two classes of sub-projects within it: incubating and graduated. And those names kind of suggest where they are in their lifecycles. Those incubating projects are ones that are still gathering community or maybe working on their infrastructure or their developing their tools. At the moment we've got three of those. I think you mentioned some of them. There's the Java Server Faces or JSF sub-project. There's the one we call Dolly, which is really the persistence portion, or the JPA APIs of the JB3 spec. And there's the newest one, which is the AJAX Tool Framework, or ATF for short. So JSF and Dolly have just released their first technology previews coincident with the Callisto release. So we're really excited to have them get to that point and have people start to be able to pick up those technologies and give them a spin. We have two graduated subprojects, two that represent our more mature components that you touched on. Those are WST, which stands for Web Standard Tools and JST for J2EE Standard Tools.

Riyad Kalla: I see.

Tim Wagner: And these really kind of divide up the world into, the easiest way to think of that is the Java and the non-Java. WST has a lot of our language infrastructure, for example, both the language frameworks and then a bunch of the specific editors for HTML and CSS and so forth. And also it has some of the underlying infrastructure for such things as deployment and server management but nothing that's J2EE specific. All those J2EE specific APIs and tools are really over in the JST side.

Riyad Kalla: I see.

Tim Wagner: And you can kind of think of that as the J2EE 1.4 tooling, more or less.

Riyad Kalla: Okay.

Tim Wagner: Now you also asked this important question about how do we avoid feature creeping? How do we keep WTP from embracing the entire universe, as it were?

Riyad Kalla: Right, right.

Tim Wagner: So, part of that answer actually lies in our project charter which draws this interesting line that says WTP can only adopt technologies that have been standardized -- that have some kind of specification associated with a recognized standards body.

Riyad Kalla: Ah, Okay.

Tim Wagner: So this gives us a great litmus test. And what happens in practice, of course, it's really the more mature technologies that tend to fall into that camp. So, those are the ones that we work on. That's why, for example, J2EE is part of the centerpiece of WTP. But, for example, PHP is something that goes on elsewhere in Eclipse today and so there's kind of a nice dividing line there. But, of course, what's in scope for us is still really a big challenge.

Riyad Kalla: Sure. Absolutely. What's in scope is bigger than most of these projects combined under the Eclipse umbrella. So has the size and the breathe of managing WTP, all the way from the first drop to what you guys foresee being in the 2.0 release and so on -- is that becoming a problem to manage? Not only all those people, but all those teams to come together and collaborate and make sure everyone's on the same page? What's that like?

Tim Wagner: So, there are definitely challenges there. From a planning perspective I think we have a very clear idea of where we want to get to in the next year. And that really helps create some structure around both the process and the specific activities we do. For example, one of the things we have coming up for next year is sort of this general arc of working on the Java EE5 specs. We see that already in the technology previews for Dolly and JSF. You see it in our planning documents -- everything from deployment to scripters, some of the web service technologies that are unique to Java EE5 that we're going to be working on for WTP 2.0 that will come out next year as a part of Europa.

Tim Wagner: So, some of it is having those general themes we can work against that help provide some kind of clarity in that everybody can execute to over the course of the various years. It's also the case that we have a number of different companies involved: Sybase and Oracle for example, have just gotten actively involved in WTP in the last year. So, that helps us handle the scope in the sense that we scale out to more people and more companies. And, of course, they bring with them very talented developers and managers and so forth, who can help do part of that workload. Some of our growth is kind of a natural horizontal extension in that sense. But, it's certainly true that the larger the project gets, the bigger the management challenges become, and so it's certainly become a full-time job for me, which is a little bit challenging sometimes given that I also have a full-time job with BEA.

Riyad Kalla: Absolutely. Now what is your time at BEA and working with WTP? And BEA, for those of you who don't know has the Workshop Studio project that's winning awards -- quite an accomplished product. How does that work? Working part-time on the Workshop Studio and bringing all the knowledge you guys have developed internally from your tools -- how does that work with the synergy with WTP? The team's knowledge between the two platforms?

Tim Wagner: Sure. So, while it's a lot of work for me personally I think that as a business model for BEA there's an amazing amount of synergy here. Our model for doing this is a really simple one that a lot of other companies are doing very similar things. Our engineers contribute to open source technologies like WTP. We also have people working on the platform in the form of the Java annotation processing. And then as a company we then turn around and adopt the Eclipse platform, and obviously WTP as well, as the base for commercial products. At the moment, we have two products in the Workshop family. We've got Workshop Studio, which just released its first WTP based version last month. We also have Workshop for WebLogic 9.2, which we just released last week. That's the successor to the award-winning WebLogic 8.1. And eventually, of course, we're planning to merge these two codebases and produce a single family of products out of that merged codebase.

Riyad Kalla: Is there a lot of ideas getting put back into WTP as you guys are working on these much bigger commercial projects? Are you saying, "Oh, we really need this functionality in the platform, so let's put that into WTP, " and so on and so forth?

Tim Wagner: Oh, absolutely. There's kind of a general litmus test there. If there's a feature that is -- that benefits the entire community because it helps to grow a shared practice or facilitates a standard or represents an evolution to components that are already in open source -- those are things we always contribute directly down to WTP. And you see that with some of the examples. Project facets, for example, is something that we contributed out of some work that we had started at BEA but it benefits all of the adopting companies and will show up in IBM products and Oracle products and really any company that's adopting it, not just BEA. And then, of course, there are things were we think we're adding, we're differentiating or we're adding real value to our specific customer demographic. For example, some of the views and graphical editing experience inside Workshop Studio we still think has a unique characteristic to it that has a lot of value. And so, some of those obviously stay on the commercial side. But it's really a sliding window. We've got what we call a blended development strategy. We kind of think the leading edge of that window being the place where we innovate and the trailing edge of that window is something where we're moving things actively into the commodity arena, which means either we're donating them to WTP or someone else has already done that and then we're adopting that as the framework on top of which we build.

Riyad Kalla: That can be tricky, trying to balance that because you want to stay competitive, you want to stay ahead of the game, but at the same time you might have some really ideas that you want to kick down to WTP.

Tim Wagner: Oh, absolutely. So, as you imagined, that's a big part of our product management lifecycle now is understanding what our relationship is and should be to the various open source projects that we play in. That includes both Eclipse on the tooling side, Apache on the run-time side with our Beehive project, and a number of other places where BEA is actively engaged in the open source community as well. I like to tell people it's a little bit like jogging in front of the bulldozer. You can stay in front of it but you don't want to take a nap on the road.

Riyad Kalla: [laughs] Yeah, you're exactly right. So we can expect to see the Nightrox and accessory stuff in WTP 2, right?

Tim Wagner: [laughes]

Riyad Kalla: Probably not?

Tim Wagner: Probably not in this version.

Riyad Kalla: Oh, Okay. We get another year before that's out. All right, you about innovating and collaborating on a lot of things. Eclipse Zone recently spoke with the creator of the Common Navigator Framework. And it seemed to me, both from a usability standpoint and from all the examples from the Navigator Framework that all the usability and the motivation for this really came out of the WTP project. Can you explain how that happened or what the role of the WTP project was in that new framework?

Tim Wagner: Sure. That's something that we really had really done a lot of work on. If you look at our 1.0 code line, the 1.0 and 1.1 releases, you see a lot of that functionality and really the precursor to what's down in the Eclipse platform today. That was something where WTP needed that additional resource flexibility -- resource to visualization capability. And so, we had a done a bunch of that prototyping work at the same time being well aware that WTP really wasn't the best final resting home for it. And so, a little over a year ago, started conversations with the platform team and over the 1.5, really the whole Callisto development cycle, that work transitioned down to the platform where of course many other very bright and capable engineers added their own ideas for the APIs into implementation. So, what I think we've ended up with now is a fantastic piece of technology for community as a whole. And from WTP's perspective we took something that really from our perspective was much better as a piece of infrastructure, and we got that down into the platform. It's a nice way in which we've continued to evolve our codebase and moved some of those things we've had along for some time down into the right parts of Eclipse.

Riyad Kalla: That brings up a good point. I notice a lot of users, when they install Eclipse for the first time, they immediately want something like -- well, I see a lot of people using the Ant editor as a fundamental XML editor or other folks using just the text editor to edit their CSS or HTML. And so a lot of folks that maybe don't do web development need editors for these different types of files that WTP seems to have in spades. WTP ships with six or eight different file type editors. Has there been talk about some of those editors being pushed down into platform?

Tim Wagner: So that isn't a conversation that we've had to date. I think that that will happen conceptually as WTP becomes associated in more people's minds with the fundamental packaging of Eclipse technologies that they get.

Riyad Kalla: I see.

Tim Wagner: You see that both with Callisto, it's very easy, for example, now to add WTP on to the platform base so that you have all of that functionality. And I think you also see it reflected in adopting companies. So, for example, we talked about Workshop Studio, which is now built on top of WTP so all the infrastructure, all the language support in WTP is also present in that commercial product. And certainly it's something that if there were interest in moving some of the language and editing infrastructure -- WTP calls that the Structured Source Editor or SSE, you'll sometimes hear it referred to -- that's certainly something that we have had some conversations about moving that particular piece into either the platform base or into it's own top-level project. But it represents a fairly major investment in energy in terms of that whole language arena for Eclipse. And I think that's something that a lot of people are interested in. There's definitely been some discussions around it, but it's a big step so it's sort of getting the due care that it needs in order to make sure that is in fact the right thing to do for the Eclipse architecture as a whole.

Riyad Kalla: I see, Okay. Changing gears a little bit: for people who don't know, why don't you describe what Project Antoine is and how that started, cause I think caught some people off guard as far as, "wait a second, what's this thing?" when it was first announced.

Tim Wagner: Sure. I think this is really fantastic and interesting research project. As you mentioned, it's called Project Antoine, and since I don't speak French I'm sure I'm slaughtering the pronunciation here. But it's sort of named for a French writer whose memorable quote was, "Perfection is achieved not when there is nothing left to add, but when there is nothing left to take away." And so, the idea behind this -- and it's really a UI and usability study -- is to figure out what actions are people doing and where is the UI not really providing them with the fastest, easiest, most direct path to their end goal. So this is being done by a researcher by the name of Leaha Finlaor and she working out of the University of British Columbia, Computer Science Department. She's doing this with the, certainly with the support of WTP, and the intent at the end of that research study is to share her results both with us and, of course, therefore with the larger Eclipse community. I think everybody will end up benefiting. And Leaha is actively recruiting volunteers for her study. She's still in the first part, which is really the data gathering part where she's looking at the user -- what users are actually doing and gathering some data. She has a very non-intrusive and non-performance degrading plug-in that you can just drop into your Eclipse WTP installation and it'll do that recording and help her collect that data. So, if anybody is interested in doing that and participating in her study I think that would really help the entire Eclipse community and certainly WTP project going forward.

Riyad Kalla: Why WTP? Did she communicate with you guys ahead of time that this was something she wanted to do, or did it catch you guys off guard as much as it did users?

Tim Wagner: Oh, no. We had several discussions. Leaha came and did a presentation to the deputy PPMC and both asked us to support her research project both in terms of helping her find volunteers and helping our community understand what was going on with this research project and why it was ultimately in their best interest to help participate. So we think of this at the PMC level as a win-win deal. We love the idea of researchers using Eclipse as a platform for study, for research, to advance both their own understanding and certainly to advance ours in terms of, particularly in this case, of the user experience. So, we are very much helping Leaha out with this in a cooperative sense. But all of the research is being done by her and the other grad students who are working with her project. It's not a WTP research project per se.

Riyad Kalla: Has there been talk about maybe, if this is a success, continuing this type of work with other big projects like Bert and TPTP?

Tim Wagner: Oh I think that be very obvious extension of some of this work. Hopefully some of the discoveries that Leah would make, since of course WTP doesn't occur in a vacuum; we're part of Eclipse. We rest on top of much of the Eclipse platform. So I think some of the things she discovers be applicable directly to more than just WTP itself.

Riyad Kalla: I see.

Tim Wagner: And certainly she's also building her data-gathering on top of Mylar, so I think there's some interesting...

Riyad Kalla: Oh, Okay.

Tim Wagner: ...interesting synergy with that technology project. As this goes forward, I think it's also a very interesting case study in how you do these experiments. I know that certainly my company BEA and others would be very excited and interested to see additional tools in terms of studying the user experience and really learning how we can make Eclipse more useable, more functional, and certainly more productive for users. Particularly with things like Callisto where as people now include a much wider array of projects and technologies in what their downloading and installing. I think it's even more important that we have a consistent and really streamlined user-experience. Hopefully, this an early step in that direction.

Riyad Kalla: You bring up Mylar. I always found it interesting there was something about Eclipse and the Eclipse platform that attracted these much more colorful uses of Eclipse then just, "Oh here's a new editor, here's a new project style, or here's a deployment tool, " that other IDEs have never done so far. So, that's interesting. I've very curious where this is going to take us and if it is going to lead us, at some point, say rolling Mylar into Eclipse four or something along those lines. That would be a first, I think, for any IDE out there.

Tim Wagner: I think that's very true. In fact, I'm now committer representative on the Eclipse board. And one of the things we talked about when we met in Chicago last week was exactly that. Really had this interesting discussion that in many ways Eclipse is defined as much by what we call the larger ecosystem, the breadth of the projects and the number of different people who can come to this and use it as a platform, than it is some particular feature set of an editor, for example.

Riyad Kalla: Right.

Tim Wagner: And I think that will really be one of the enduring qualities and legacies of Eclipse as a community, that it's been successful for so many people and so many disparate technologies finding that common base and that common infrastructure helpful.

Riyad Kalla: Okay, so switching gears here for a second and talking about 1.5 release that just came out in Callisto and I'm sure folks are using right now. I noticed just right off the bat, one of the first things that made me sit back and go "wow" is that it seems like a lot of effort was made to multi-thread a lot of the WTP's behavior. I'm noticing my projects building magnitudes times faster, things like that. Was that a huge focus point, or did that come out as a side effect of something else?

Tim Wagner: I'm glad to you hear you say that because certainly a lot of effort went into improving performance this time around. Really the whole 1.5 release, our focus on this was to try to mature the tools and platforms. So that meant, stability, reliability, performance, all these things which maybe aren't as fun necessarily as putting new features there just absolutely vital for making it really work for end-users. You picked a perfect case-in-point. We've done a lot of work around both the stability and the performance of the multi-threading for the 1.5 release. And also invested a lot of energy in improving the validation in general. So, more of those jobs run in the background now. We've made some performance improvements, particularly around editing large XML files and some of the threading issues around complex models with a lot of different editors sort of looking at the model simultaneously. So, we're hoping that the end-user editing experience is better and also for adopting companies they're seeing the APIs really behave in a much more mature and preferment fashion now. Very happy to see that part of the code base getting the attention that it did.

Riyad Kalla: Now, what are users requesting the most from you guys consistently since 1.0? Are they asking for improved editors and better auto-complete or are there things that they consistently asking for? Big items?

Tim Wagner: Well, probably the single biggest request that we get has been a visual JSP editor. When you sort of poll users that's the one that's always been at the top of the list. And it was very exciting, just I believe around two weeks ago, Sybase announced that they are planning to donate their visual editor for JSP and JSF to the WTP project.

Riyad Kalla: Oh, fantastic.

Tim Wagner: Yeah, it was just really fantastic news for users that there is a plan to get probably the single most requested feature into the project. So Sybase has done some internal discussions with the PMC and with the larger WTP community about where that code should go. They've begun their internal processes with it. So it's not quite there yet for anybody to go download, certainly not anything that's part of the 1.5 release stream. But something that as move forward next year into 2.0 and the Europa release with the rest of the Eclipse projects that I think we're going to have a really exciting new feature area and be able to address that long standing request.

Riyad Kalla: You guys are going -- as commercial companies you guys are really pushing the bar up on yourselves to innovate. I mean, now that you're going to have the visual designer in WTP, that's excellent for users. But now I'm excited to see what's going to be coming in the commercial tools to keep innovating. So, now there's kind of an unsung hero here that people don't realize until they start moving between IDEs or trying to do different things. And that's the project model. As I understand it, that's has been kind of a sticking point for a lot of people for a long time. Now WTP has, I would say, the problem of needing to figure out a project model that is compatible with everyone so all these commercial projects can build on top of it, keep their compatibility between themselves, between each other, because they all use the same project model, but still allowing all the functionality that all these individual commercial have independently developed. Has that been a big focus internally for you guys? Is that a big effort there, to get the project model down?

Tim Wagner: Certainly, on the commercial side at BEA that was one of the larger pieces of migration of the studio product on to the WTP release. We invested a lot of time and energy making sure that project model was both compatible with WTP but also for the studio end-users that there was no change or loss in functionality as they experienced it.

Riyad Kalla: Right.

Tim Wagner: That was definitely something we spent a lot of time on, insuring that was a smooth transition. And within WTP the project model is probably one of those things that has garnered the most conversation and discussion and technical planning over the course of the last year. Not just because of our desire to align more directly with what was going on in the Eclipse platform but also because it's just a very complicated area. Obviously there's a lot of J2EE considerations to consider there. Different companies do things different ways...

Riyad Kalla: Absolutely.

Tim Wagner: ...have different deployment requirements. And WTP, the project model for us, has been the nexus where so many of those concerns and considerations kind of came together. Yeah, that's definitely something we spent a lot of time and energy on. And part of what we've been doing is also trying to, where possible, follow that maxim of "Simplify as much as possible but no further." Because I do think we probably had too complex a project model previously, something that was actually made it very it very hard for adopters to work with. So, we've tried to streamline that a little bit in the 1.5 release.

Riyad Kalla: Do you think there's going to be a lot of work moving forward for the project models and deployments for 2.0 or do you feel like you've got pretty much 80% of what you want, it's just polish time.

Tim Wagner: I think we're most of the way there now and it is, as you said, more about the spit and polish at this point. I think one of the things that's also happening is that if you take a look at some of the areas of Eclipse, for example STP, the Solo Tooling Project, that because they're at an earlier part of their life-cycle and they're also trying to embrace some different deployment and run-time models that some of the discussion around what makes for a good project model and so forth has also now come to embrace projects like STP. And so, they're sort of helping to take up the reins of some of those discussions in figuring out maybe what comes next or what areas need to be generalized. Obviously, that's a project that we've been working with and we certainly imagine over the course of the next that they'll be doing some very interesting things in helping to drive, sort of push further on the project model. Maybe we'll see the next stage of evolution coming from that direction.

Riyad Kalla: Very cool. Okay, so what is your favorite feature or piece of work that made it into the 1.5 release?

Tim Wagner: Oh wow. Let's see. I think I'm going to have cheat and give you two here.

Riyad Kalla: Sure, go ahead.

Tim Wagner: One thing that's just very fun, because anything that's visual of course is always kind of a joy to go and experience. So, the XML schema and WSDL editors, if you use the graphical views for those...

Riyad Kalla: Oh, yeah.

Tim Wagner: They've been improved. We've made them both not only more performent and skill and scaling better to large files but just sort of tuned up the graphics a little bit. If you just try those out, give them a spin. I think they're now easier to use; they look great, and they're just really closer to the commercial kind of quality of editors of that sort. And I'm really excited by the new experience that you get there in 1.5. The other thing that's had a lot of work in 1.5 that I think people will notice right off the bat is the web-servers. Even just on the UI side, we've made a lot of enhancements to the wizards there. I think it's much easier to understand where you are in the life cycle of the wizard. And also just making the common path simple so that more defaults, a much more streamlined experience, and then of course support for things like Access 1.3 really helped to make to make the whole web-services part of WTP distinctly improved in 1.5.

Riyad Kalla: How much of this work is coming from user feedback? And I ask that because for most of these other projects it seems like -- take VE for example, the visual editor project -- all of it is, I think, based on user feedback and just the individual's experience. But for you guys, since the WTP is so big and consists of so many strategic partners that have commercial projects that compete in the web-space, the web development space, is there a lot of feedback coming from users on the newsgroups and stuff that you didn't already know or you didn't know needed to get done or you haven't already heard a hundred times from your commercial customers?

Tim Wagner: [laughs] It is certainly the case that sometimes we have individual end-users who have very specific -- they find bugs obviously, they have specific feature requests. As you say, a lot of the larger companies, the strategic developers and other projects and commercial entities who have adopted WTP have helped us by identifying the bugs early, finding the gaps in the features, or performance or whatever they are. And one of the ways we have of recognizing that is what we call the adopter portion of our website. That's a place where we provide tools, processes, feedback and so forth to help whoever those are -- they can be commercial companies, they can be other Eclipse projects, they can be open-source projects -- where they can come and we'll give them special attention because they represent, each one of them, multiple end-users. So we something called our hot-bug list, where we let them register bugs that are blocking their adoption or really block a whole set of users from getting something done by virtue of fixing that.

Riyad Kalla: Oh, that's cool.

Tim Wagner: And then we also have a usage tracking tool. So, because we haven't declared API in all of our areas, we still areas, components maturing in that respect, it's important for us to know which provisional APIs are in use by the various adopters. So we have a tool that they can run and give us feedback and say, "Hey, this is an area we are using. We found this API helpful. We didn't find that one helpful." And that enables them to give us feedback about that, so that we're able to both support them by being careful on how we transition from provisional APIs to final APIs, but also know what's useful and what isn't. So, it's definitely the case that I think reaching out to adopters in WTP has had a lot of reward. But, you know, we still get the bug and feature requests coming in from the end-users and I think probably with Callisto we'll certainly see plenty of that. At least, I'm keeping my fingers crossed.

Riyad Kalla: Right. Well, now, what is -- you mentioned, talked a little bit about WTP 2.0 what's due with the next major release of Eclipse next year. What are some of the features that either were sidelined because of time constraints or just the amount of work required that you are really looking forward to having in the next release of WTP that just have you excited?

Tim Wagner: Certainly the progress towards Java EE5 is a big one there. And you can see that in the sub-projects that are beginning to mature, JSF and Dolly being the two cases in point there. And then just also in terms of JST, our standard J2EE tooling, of really getting more of the functionally from the next release out there. We think that it's early in the adoption curve, obviously, but our goal to start to have tools ready really as the Fortune 1000 IT shops start to adopt into the EE5 space. We're really exciting to move in that direction. I think that's something all the developers see that as an important feature and something they're very exciting on. You can really sort of measure the progress as we advance on that score. There are a couple of infrastructure pieces that are perhaps not as glamorous but I think are important. For example, we're planning to move off of our relational database tools and onto the DTP, the Data Tools Project, within Eclipse. So I think that's, from an architectural perspective for the Eclipse community as a whole, that's a very important transition for WTP of the scale and magnitude of the project navigator evolution that we went through in the Callisto release. And probably the one that has a lot of people excited and has been generating a lot of buzz is watching the ATF project grow.

Riyad Kalla: Right.

Tim Wagner: That's a very -- obviously a new technology area in general and certainly a new project for WTP and for Eclipse. So it's really at that early, nascent stage of its life, but it's starting to grow its community. I think there's a lot of interest in it. And I think over the course of the next year you're going to see probably getting to technology previews on that and really watching a lot of people looking at that, at what's coming out of ATF, and getting excited about the tool there.

Riyad Kalla: Yeah, I almost feel bad for them, because they're at that stage where the tool's in it's infancy but the technology's in it's infancy too. So every week there's a new AJAX framework. So I feel bad that they've got to weed through all that stuff and figure out what to support and right after they support it you've got ten users telling you you've supported the wrong ones.

Tim Wagner: Yeah, it's both exciting but with the excitement come a lot of challenges there. So...

Riyad Kalla: Yeah, that's true.

Tim Wagner: But it's definitely at a place where the technology is changing rapidly. There's certainly a lot of interest in it. I think some of their pieces really compliment, for example, just better JavaScript editing support is something that WTP users have been asking for for a long time. And finally, we're going to see that coming in via ATF.

Riyad Kalla: Okay.

Tim Wagner: I'm excited to see some of those complimentary pieces of technology even just in the pure editing sense come to pass.

Riyad Kalla: Right. So, backing up to JEE5 real quick. Talking about support moving forward and things like that, are we talking about possibly incremental updates to WTP like maybe in the middle or the end of this year, a like 1.7 update that would add some of that functionality or is all that going to drop in the WTP 2.0 release, you think?

Tim Wagner: Our current plan is to stay aligned with the platform, and what is now being called the Europa train for next year.

Riyad Kalla: Okay.

Tim Wagner: So, we'll definitely do maintenance releases along with Callisto. So there will definitely be a WTP 1.5.1 and 1.5.2.

Riyad Kalla: I see.

Tim Wagner: But I think for the next feature release we're probably looking at coming out next year, in part just because it's difficult to produce a lot of releases during the year and it's also I think there's some big pieces there. I want to give them a chance to be developed to a reasonable level of maturity before they go out the door.

Riyad Kalla: Absolutely.

Tim Wagner: But what you will see is, on the sub-projects, I think you will continue to see work by them as they move through their lifecycle. So JSF and Dolly, for example, will continue to do off-cycle releases as that technology matures. And the intent and the hope is they join the normal WTP 2.0 train and sort of become full fledged components when we go out with Europa next year.

Riyad Kalla: I see. Okay.

Tim Wagner: So definitely keep an eye out for updates to those tech previews.

Riyad Kalla: You guys are covering a lot of bases there, and I know a lot of developers, including myself, appreciate that work because it makes our life easier. So we're getting to the end of this podcast. Is there anything else you want to bring up before we head out, Tim?

Tim Wagner: We'll I'd just like to express my thanks for the opportunity to talk to you today about WTP.

Riyad Kalla: Absolutely.

Tim Wagner: And really say thank you to all those committers and contributors to the project who've worked so hard to make it a success. We've done six releases now in less than a year, culminating obviously with Callisto. So, it's really a testament to their commitment to the project that we've made so much progress in such a short period of time.

Riyad Kalla: Absolutely. We'll we wanted to thank our guest, Tim Wagner, from the WTP project for taking the time to sit down with us and talk with us with his role in the WTP project as well as the WTP project's role in the Callisto release. We appreciate everyone tuning in and hope you enjoyed yourselves. This concludes episode three of ten in Eclipse Zone Callisto podcast series.


[music]

Transcription by CastingWords