Dev Thoughts

Amazing…

I tried logging in today to TSS (TheServerSide)… obviously since I rarely go there anymore, I forgot my password… but I really really really wanted to login and reply this ridiculous post be Bill Burke. Anyway… I went and did the unthinkable and actually re-registered again (stupid… but did I mention I really really really wanted to reply to that post :-)). OK… so I registered… never mind the stupid information I needed to fill in just to be able to reply a post (my Address? dah!!!), but after all this relatively tedious form, I finally received the so called activation email. Why “so called”, because it appears that even when you click the activation link you’re still not activated. After a while I got another email confirming the activation. Ah… naively I quickly tried to log in again… but nope… not possible. So I found myself clicking the “Forgot Password” link again (only a few minutes after I registered… I must have really bad short term memory). Got another email, with my username (email) and password. Obviously that was the exact same info with which I tried to log in. Now, I guess I’m really stupid for even taking it this far by trying all these steps. Point taken… I’ll stick to InfoQ and forget about TSS for good… I guess, after all, it is an appropriate place for posts like this one.

Oh… Yeah… Bill… here’s my comment:

“BTW, beyond the functionality this framework introduces, I think this is a great general example/usecase for stupid over-/mis-use of annotations in general. Thank you!!!”

(Thank god for this blog… I just needed to get it of my chest)

In recent blog post Marc Palmer lists 10 misconceptions about Grails. The blog also engaged quite a discussion in InfoQ from some proponents and disbelievers. In JTeam we’ve been trying to use grails in several small projects. In general, we really liked it. As we’re a long time Spring and Hibernate users, the learning curve was mainly the learning cure of moving from Java to Groovy (which by itself was quite smooth). That said, we did encounter two major difficulties that in a way prevented us from pushing it further to larger projects:

1. Lack for GOOD IDE support. In general we’re all Intellij users and the only groovy plug-in available then was GroovyJ - which is far too simple for any grails devleopment. Eclipse come with a better support for groovy, but still, debugging was practically useless as it was not possible to evaluate variables and expression. I haven’t looked at the Eclipse plug-in for quite some time now so they might have improved it by now. What is more interesting for us though is the new Groovy & Grails plug-in developed by the JetBrains guys.
2. We use maven2 as our build system. It was REALLY quite a hassle to integrate a grails application with maven. Definitely as a separate module in a bigger multi-module project. Eventually we found ourselves developing a project in two different environments and synchronizing libraries all the time. This is a great obstacle as we see it. And as the user forums indicated back then, they had no plans of integrating with maven (As I understood Greame Rocher, the founder of Grails, was/is a big anti-maven guy). The worse part was that back then they moved from using Ant to Gant (A groovy based ant) which was 10 times slower than the original ant scripts.

All in all, I think Grails (and Groovy) has great potential. they did really good job of bringing all this set of successful and useful technologies into the enterprise groovy world. I’m quite enthusiastic to see how the new intellij plug-in works… and perhaps we’ll even manage to integrate it as a maven2 module. I’m sure when these issues are fixed, we’ll look at Grails again and see how we can leverage it in our project development once more.

It’s been quite some time since I last blogged about practically anything. Been quite busy in the last year but I guess it’s not the best excuse to cover up my laziness. Anyway… I’ll try to do better from now on.

Back from Javapolis last week. I had a great time there although it was unfortunate that I didn’t get to attend many sessions. Still it was sure nice (as always) to meet so many people who I share my passion with.

This year Bram and I gave a quickie on facetsearch - a new open source initiative that aims to bring facet based search functionality to enterprise java. Personally I was quite surprised that quite a lot people actually showed up considering it was during the lunch break. It’s quite exciting to see more people recognize the importance of search in general and faceted classification in particular. We also had very interesting talks with some of them.

The other big thing I’m taking with me from this conference is Grails. I’ve known about this project for quite some time now (mostly from Steven) but I was definitely not aware of the progress they’ve made during the last year. Lately this project got quite some attention and so I made sure to at least attend Graeme’s presentation. Well… I was quite stunned… it was probably the shortest presentation in the conference (exluding the BOF’s and Quickies) yet it seems to have made the most impact. Let’s just say that later that day I wanted to buy the new Grails book, and I was quite surprised to hear from the seller that they were all sold out directly after the presentation. Waw!!!

The presentation was very much to the point. I liked the fact that the one of the first things Graeme said was that easy CRUD application is NOT in any way the goal of Grails for that’s a common believe among many developers (if anything I guess it’s just a side affect). The main goal of this project is to provide a solution that benefits from both the dynamic languages and the enterprise java worlds. Graeme showed really cool examples of how easy and intuitive grails makes working with ORM solutions like Hibernate (JPA is supposed to be supported as well). The web tier is based on SpringMVC and looks very easy to work with. It was even more exciting to see how grails doesn’t lock you in… you can actually use any “real” java class in all tiers. One of the main doubts we always had with Ruby was scalability, with this integration we have nothing to worry about… by the end of the day it’s just java. Anyway, I’m definitely going to further check it out and I would recommend you to do so too (the only thing that bothers/worries me now is tool support as I’m a committed intellij user).

Spring 2.0 was released yesterday. I’ve been using spring 2.0 since its RC2 release and it’s a joy. So far the 2 best features that stood out above the others are the namepsaces support in the configuration files and the new annotation driven transaction support. Nonetheless, I can definitely see the scoping and the new AOP support (with aspecj pointcut expression) added to this list. Congrads to all the spring developers!

Other news that were supposed to be exciting is the Intellij IDEA 6.0 release. I’m a huge IDEA fan, and I think the creators of this wonderful IDE are geniuses. This is why I was very much surprised and disappointed to experience 2 crushes of the application in one day (today)… On top of that CVS support doesn’t work - in fact any attempt to connect to net fails. Guys, again… there’s no way I’m switching to eclipse… ever… but please, don’t fall down from the standards you have set - this release should not have been released yet - I’ll stick to IDEA 5.x until version 6.0 will work properly.