Talk about backfiring.

Lindows, which was fighting with MS over its name, had setup a page to tell MS "see, there are other companies who also use 'Windows' and are not sued by you." MS's reaction apparently has been "You're right, we have to be consequent!" And so they did...

Well, I've found my RSS aggregator on Windows: FeedReader seems to be a nice little tool. Only 4 things left that I need:

1. An RSS aggregator for Linux
2. A weblog client for Windows
3. A weblog client for Linux
4. The exact configuration parameters to connect the 2 previous ones to FreeRoller

Off to Google it is :-)

Well, I'm being hyperlinked at last (although I have to admit I had to do some PR for this).

With the risk of this becoming a pong game, I want to make myself clear on one point: I didn't mean that the MS IDE's were the pioneers for Eclipse. What I did want to say is that I've been drooling at some collegues who were working with VB and could use auto-completion (some years ago now), while we were stuck using UltraEdit or something alike (btw, verifying this link reminded me of some scary looking stuff. I'm still not sure if this is ironical or not). And to be honest, jEdit now has a plugin that does autocompletion. But (and this was the point I wanted to make) there had to be yet another commercial entity developing refactoring support before the OSS community jumped on that bandwagon. (And yes, I agree that "OSS" is even stretched a bit here, because IBM is the driving force for Eclipse).

All in all, I think that we are agreeing movingly with each other without knowing or wanting to admit it. It's not the first time that that happens...

Marc has been writing some bashing on Eclipse. I seem to be on the other side of the fence: I like Eclipse. I have been using jEdit for quite some time, so you could say that every IDE that gives me autocompletion is a step forward for me :-) Don't get me wrong: I still like jEdit as an allround editor, but for programming Java I'm leaning over to Eclipse 2 right now.

Several people have told me to use IDEA, but noone has really convinced me yet. For a start, IDEA is not free. There are several solutions to this problem:

• Get a crack. This is not fair to the developers of IDEA, and so I won't do it. In fact, once you start using Linux, you'll find the idea of "cracking" things less and less interesting.
• Get somebody to pay for it. Like your employer. Yeah, like my employer cares. "We already have UltraEdit and WSAD, you don't need an IDE that allows you to be productive."
• Pay for it yourself. Good idea, but the last few months I've been burning most of my money on books and a son, so if I can save € 200, I will gladly do it.

Besides the financial story, there are other reasons to use Eclipse instead of IDEA:

• Support the OSS philosophy. Free the source. I am willing to accept a little missing functionality if I get access to the source.
• I like the perspectives. I remember those days of opening 5 files to adapt your build process, open 2 more for building everything, open 10 Java source files to look at some stuff, ... In the end, you had tens of files open, which were used in groups in time. While looking at build.xml, you wanted to have a look at build.bat, and you lost several seconds searching in the myriad of open files. Hence the perspectives. Do one job at a time, with the files you need at that time.

And Marc, if you want to use team stuff (i.e. CVS), it might be a good idea to look for "Team".

And only now I've seen his conclusion:

The OSS believer in me would definitelly like a free alternative to the IDEA.

That will teach me to read before write :-)

One final tought though: why did it take so long for the OSS community to come up with an IDE that can do auto-completion? Do we really need a commercial entity to show us how to do things, so we can copy it (that is, everything but its code)? It's not that auto-completion is new: the VB IDE has been around for a long time. But apparently we need a commercial Java IDE before we can make an OSS alternative. Cocoon seems to be one of the few exceptions to this rule: it is unprecedented, and still unchallenged.

It seems like there is a lot of stirring in the mud-pool going on at Apache. They're re-examining their project structure, as well as funding and conferences. All this is done by a few people, who have too little time already as it is, so some things are misunderstood due to lack of time. And people start biting each other's heads off. I hope it'll come together in the end.
In fact, it's this kind of overhead that has worried me ever since I heard about OSS. GPL against LGPL against APL against BSD, RedHat against Mandrake against Debian, Tomcat against JServ, etc etc. Julius Caesar already knew the "divide et impera" adagio. Bill will like to see it. Nuff said.

Well, I'm now giving it a go with mock objects (which I started to think of this morning), but I'm still a bit uncertain whether this is the way to go. I have a gutt feeling that everything will become extremely complicated over time :-(

One of the main problems I've encountered some times now, is: how can I unit test programs that make use of a DB? The testing is very likely to read (duh!) and change the data in the DB, thus making the tests dependent on (a) the scrap tests that you do manually while coding, and (b) each other.
Of course, we have dbUnit (with its academic/philosophical base), but this approach requires 4 (four!) databases. In the company I work in now, it appears to be unfeasible to have one database (which happens to be Oracle here) per developer. So now I'm looking for the silver bullet.
I already encountered one interesting read, but it seems to be a discussion without a real conclusion.
(Remember you can always send me your thoughts.)

I was just referenced to the Null Object design pattern. Nice read, and nice to store in the back of your head for future reference.

I forgot to tell you who I am. I think this page should give you a tiny idea already.

Anybody knows about a good blog client (reader)? I started with newsisfree.com, but one way are the other my account is all messed up now. I'll have to do some Googleing...

A lot to do about spam these days. Some of the solutions that are proposed are whitelists ("I know you, you can send me mail") and paid e-mail ("For every mail you send me, you pay me � 0.01). I think a combination of these can be a healthy solution: you pay me � 0.01 to send me a mail, unless you're on my whitelist. Mails like "Hi, I'm tomK, we were on college together, please put me on your whitelist" will become common.
It goes without saying that we need to adapt the e-mail clients also. "Put sender on whitelist" will be an indispensable feature.

Well, the GetTogether is over (for me at least, I guess the evening event attendees are still eating and drinking right now). It was quite a full day. Driving to Nazareth was nice (I'm used to drive to Brussels nowadays), but it still took me about an hour. A quick coffee and a roll, and we were driven into the auditorium by Steven. Matthew introduced us to the things we already knew, but I was amazed to hear questions that clearly stated the opposite -- the GetTogether seems to have even reached other people than "hard-core fans".
Sylvain has given a nice presentation about embedded Cocoon. Although it's not my business area, I think it would be very refreshing to work in a space-restrained environment one time. The greatest works of art are created due to lack of ressources. Lack of ressources forces you to use your creativity, and this gives creative works. Q.E.D. :-) In any case, Sylvain succeeded in giving a different view on things, and that's always interesting.
Carsten gave a nice overview of the portal framework in Cocoon. I've never looked at it, but I now have the feeling that (a) I should have, and (b) it is cleanly done.
Ovidiu, the HP guy with the MacOS X laptop, gave a nice presentation of the flow control in Cocoon. I finally see what's the intention here, and I hope to play around with it once. The only drawback is that I don't like XSP :-s. Anyway, according to Steven, Ovidiu put this presentation together just for the GetTogether, so all other things even, the GetTogether managed to generate documentation for Cocoon, which makes the effort worthwile already :-) Which reminds me to buy the Cocoon book from Carsten and Matthew (I waited till today in case I should've won it ;-))
Torsten seemed a very intelligent guy to me. Alas, I don't think he has given many presentations yet. You could feel his nerves in the back of the room. It's strange that smart people get nervous to talk about something they know inside out to people who have travelled several countries to hear them. I think this is some sort of misplaced modesty.
Last on stage was Marc (the guy who once hired me, and thus is one of the many causes I know the word "blogging" now, but that's another story) to tell the obliged "marketing speak of the organization". Sad that this wasn't perceived as such by most people, until he presented it as being so. As always, Marc tended to think that all people share his view on the world, and forgot to tell what the purpose of xReporter truly was about. We learnt a lot about its architecture though. He also made an effort to make some advertising for wings, which I thank him for :-) It already got me one reaction: "why do you use jCharts, and not JFreeChart?" Well.
So the day is over. I enjoyed it very much, not in the least for finally being able to put faces on email addresses. Lessons learnt:

• Not all highways are traffic jammed in Belgium in the morning.
• It is possible to organize a seminar that doesn't run out of sandwiches at lunch time
• Smart people are modest. Bloggers aren't.
• If all speakers keep up to their promises, we can expect tons of documentation on the short term.
• When in doubt, always take your laptop with wi-fi with you. It might be that a geek seminar provides wi-fi access. Saves you an hour of typing in the evening.
• One stubborn guy can get 100 people together

Last Friday, I had to add some extra features on a project that was finished some months ago. I wasn't too keen on doing this: trashing around in code that has a fairly complex structure is usually asking for trouble. But guess what: I liked it. And I liked it for 2 reasons:

• I spent time on refactoring that code before I left it. It paid of last Friday. The only thing I had to do was adapt an interface. XP, you rule!
• The second one, and this is actually the point I want to make: I've been working on the same project for some months now, and it was very refreshing to put my mind on something else again (and for longer than half an hour in the evening). Now, I don't know if it is a general accepted feeling among programmers, but this certainly goes for me: variation keeps the mind fresh. The pain is, that most programmers have to work for months on a stretch on the same thing, after which they're reassigned to the following project, which again is the same stuff for months. Programmers are humans, not machines, and should be allowed variation in their work.

Of course, not all programmers are like me (I hope), and few things are more frustrating than a failed project that keeps haunting you, but getting to work on several interesting projects at the same time would surely improve my efficiency.

I've been looking forward to the Cocoon GetTogether for months now. I'm really curious how it will end. In fact, the expectations are so high, that it can only be a disappointment. On the other hand, if it's going to be a big success, Outerthought will be obliged to make it a yearly event :-)

BTW, I have to say I'm amazed. When I first heard about the GetTogether, I expected to have a meeting with 20-some Cocoon people from Belgium, of which 15 or so would be ex-collegues of mine. Apparently, I've underestimated Steven. It's even a real event

The oddest thing happened yesterday. Some blogger wrote some stuff that I thought was common knowledge, but was received as a complete new insight by a lot of people.

This of course has given me the impression that I'm full of complete new insights, but that I don't express them enough (I guess starting a blog gives you the right to put aside your modesty). And a long-time acquintance of mine pushed me over the edge to start writing things down. So here goes.

Expect Java. Expect XML. Expect OSS. Expect social engineering, or "people skills for geeks" (in the two directions).